INTRODUCTION
The process of incrustation involves the colonization of living and non-living surfaces by sessile organisms, and when establishment occurs on a living substrate it is called epibiosis. Epibiosis frequency and intensity tend to be directly related to substrate limitations, thus this association can provide a series of advantages to the epibiont: hosts provide protection against predators (Wahl, Reference Wahl1989; Abelló et al., Reference Abelló, Villanueva and Gili1990); host's movements can optimize dispersal and genetic flow of the symbionts (Wahl, Reference Wahl1989), expanding their geographical distribution through larval dispersal (Key et al., Reference Key, Jeffries, Voris, Yang, Gordon, Smith and Grant-Mackie1996a); and, water currents generated by the movement, breathing, or feeding of the host can improve access to food and remove metabolic residues produced by the symbionts (Wahl, Reference Wahl1989; Key et al., Reference Key, Jeffries, Voris and Yang1996b).
The available space on arthropod exoskeletons and the continuous renewal of their carapaces due to moulting are two of the factors that regulate epibiont colonization patterns in communities where the occupation of new substrates is vital for survival (Levin & Paine, Reference Levin and Paine1974; Connell & Keough, Reference Connell, Keough, Pickett and White1985). When crabs undergo ecdysis the sessile symbionts remain fixed to the exuviae (old carapace). As a consequence, these epibionts must complete their entire life cycle—settlement, metamorphosis, growth, and reproduction—during the intermoult period of their host (Abelló et al., Reference Abelló, Villanueva and Gili1990; Jeffries et al., Reference Jeffries, Voris and Poovachiranon1992; Shields, Reference Shields1992; Negreiros-Fransozo et al., Reference Negreiros-Fransozo, Costa and Fransozo1995).
Studies of epibiont occurrence patterns can help confirm the presence or absence of terminal moult in hosts, provide data about the impact of these epibionts on host population, and help elucidate biological, behavioural and ecological characteristics of the host populations (Abelló et al., Reference Abelló, Villanueva and Gili1990; Shields, Reference Shields1992; Negreiros-Fransozo et al., Reference Negreiros-Fransozo, Costa and Fransozo1995; Messik, Reference Messik1998). Nevertheless, few studies have been performed on hosts that have terminal moult (Fernández et al., Reference Fernández, Paparar, González-Gurriarán and Muíño1998; Patil & Anil, Reference Patil and Anil2000; McGaw, Reference McGaw2006).
Previous studies indicate a predominantly positive relationship between the presence or number of epibionts and the size of the host (Jeffries et al., Reference Jeffries, Voris and Poovachiranon1992; Voris et al., Reference Voris, Jeffries and Poovachiranon2000; Santos & Bueno, Reference Santos and Bueno2001; Lovrich et al., Reference Lovrich, Calcagno and Smith2003; Yan et al., Reference Yan, Huang and Miao2004), but others registered a negative correlation (Santos et al., Reference Santos, Bueno and Norenburg2006) or lack of correlation (Fernandez-Leborans & Gabilondo, Reference Fernandez-Leborans and Gabilondo2008). Differences on infestation levels between sexes were also explored in the literature (Abelló et al., Reference Abelló, Villanueva and Gili1990; Gili et al., Reference Gili, Abelló and Villanueva1993; McGaw, Reference McGaw2006; Winter & Masunari, Reference Winter and Masunari2006; Fernandez-Leborans & Gabilondo, Reference Fernandez-Leborans and Gabilondo2008), and the results were usually related to behavioural differences.
Cirripedia is one of the most representative groups of epibionts, especially the family Poecilasmatidae, as it comprises a large number of species of pedunculated symbiotic barnacles, including the genus Octolasmis (Cólon-Urban et al., Reference Colón-Urban, Cheung, Ruggieri and Nigrelli1979; Jeffries & Voris, Reference Jeffries and Voris1979, Reference Jeffries and Voris1996; Jeffries et al., Reference Jeffries, Voris and Yang1982; Key et al., Reference Key, Jeffries and Voris1995, Reference Key, Jeffries, Voris, Yang, Gordon, Smith and Grant-Mackie1996a,Reference Key, Jeffries, Voris and Yangb). Most of species of the genus Octolasmis are found on living substrates, and Decapod crustaceans show the highest known rates of infestation by these epibionts (Jeffries et al., Reference Jeffries, Voris and Yang1982).
In Brazil, Octolasmis hoeki (Stebbing, 1894) was reported in the branchial chambers of Libinia spinosa H. Milne-Edwards, 1834; while O. lowei Darwin, 1851 was found adhering to the branchial chambers of Arenaeus cribrarius (Lamarck, 1818), L. spinosa, Libinia ferreirae Brito Capello, 1871, Portunus spinimanus Latreille, 1819, Portunus spinicarpus (Stimpson, 1871), Hepatus pudibundus (Herbst, 1785), Callinectes danae Smith, 1869, Callinectes ornatus Ordway, 1863, Callinectes spp., and an unidentified majidean crab (Young, Reference Young1990; Santos et al., Reference Santos, Bueno and Shimizu2000; Santos, Reference Santos, Escobar-Briones and Alvarez2002; Santos & Bueno, Reference Santos and Bueno2002; Mantelatto et al., Reference Mantelatto, O'Brien and Biagi2003).
Bertini & Fransozo (Reference Bertini and Fransozo2004) reported the presence of L. spinosa along the Brazilian coast during the entire year at depths from 5–45 m, as well as in great number at depths of 35–40 m. Pires (Reference Pires1992), Bertini & Fransozo (Reference Bertini and Fransozo2004) and Braga et al. (Reference Braga, Fransozo, Bertini and Fumis2005) suggest that L. spinosa is one of the most abundant species on the northern coast of São Paulo State. Despite of its abundance, only few studies were conducted in Brazil regarding the distribution (Braga et al., Reference Braga, Fransozo, Bertini and Fumis2007), food habits (Barros et al., Reference Barros, Cobo and Fransozo2008) and infestation by parasites (Santos et al., Reference Santos, Bueno and Norenburg2006) and epibionts (Winter & Masunari, Reference Winter and Masunari2006).
The present study evaluated the prevalence and mean intensity of infestation by Octolasmis lowei according to the sex, size, and moult stage of the host Libinia spinosa along the northern coast of São Paulo State, Brazil. The main goal of this study was to analyse how these variables behave in a terminal moult crab.
MATERIALS AND METHODS
Samples were taken at Caraguatatuba, São Sebastião, and Ubatuba bays along the northern coast of São Paulo State, Brazil (Figure 1), during the period from September 2001 to June 2002, as part of the BIOTA/FAPESP project, marine benthic invertebrates, sub-project non-consolidated infra-littoral.
Map of sampled areas on north-east coast of São Paulo, Brazil.
Sampling was performed using transects running parallel to the coastline using a commercial fishing boat equipped with double-rigged trawl nets with 4 m of mouth width when expanded. Along each transect the nets were towed for 30 minutes, corresponding to a distance of about 2 km, at depths from 5–45 m (interval = 5 m).
The specimens of L. spinosa (host) captured were conserved in ice and transported to the laboratory where the following data were recorded: sex, size (CW = carapace width in mm), moult stage (Skinner, Reference Skinner1962, Reference Skinner1985), and gonad development (Costa & Negreiros-Fransozo, Reference Costa and Negreiros-Fransozo1998). The sizes of the smallest non-infested ovigerous female (34.2 mm CW) and the smallest non-infested male with mature gonads (38.0 mm CW) were used to define the juvenile phase.
In order to observe O. lowei in the branchial chambers, the dorsal portion of the crab was cut off and the branchial filaments removed and preserved in 70% alcohol. The infested gills were examined in posterior analyses using an optical stereomicroscope for evaluation of epibiont populations. Specimens of O. lowei that were found were removed and counted, and their development stages were noted (cyprid larvae, non-ovigerous, or ovigerous). Barnacles that had eggs inside the capitulum were considered ovigerous, and barnacles without eggs inside the capitulum were classified as non-ovigerous. Terms like sexually mature or immature were avoided due to their relation to gonad development, which was not evaluated for O. lowei in this study.
Prevalence of infestation was evaluated by calculating the proportion of infested crabs (hosts) in the sample population. Intensity of infestation was considered to be the number of epibionts present on each infested host (Margolis et al., Reference Margolis, Esch, Holmes, Kuris and Shad1982; Bush et al., Reference Bush, Lafferty, Lotz and Shostak1997). Due to the non-normality and heteroscedasticity of the data, the median was used as the measure of central tendency, with quartiles as a measure of dispersion.
Prevalence of infestation according to host sex and sampled area was tested using the Chi-square test. Log transformation of the data did not generate a normal distribution, so the non-parametric Kruskal–Wallis test was applied in order to evaluate variables related to host size and intensity of infestation, and Dunn's method (a posteriori) was used to identify differences among classes.
The intensity of infestation data from the different epibionts categories (cyprid larvae, non-ovigerous and ovigerous) were correlated among themselves and with host size using the Spearman's correlation index (Sokal & Rohlf, Reference Sokal and Rohlf1995).
RESULTS
A total of 1049 specimens of L. spinosa (439 males and 610 females) were sampled, showing a 63% prevalence of infestation and median intensity of 21 epibionts/host (range: 1–644; Q3 = 81).
The prevalence of infestation values as total are presented in Figure 2 and discriminated in Table 1 according to the sex and size of the host. The absence of infested crabs can be observed in the first three size-classes, and these correspond to juvenile population categories. The smallest infested ovigerous female of L. spinosa was 30.7 mm (CW) and the smallest infested male with developed gonads was 31.1 mm (CW). The infested hosts were always larger on the average than non-infested animals among individuals measuring more than 34.6 mm CW (P < 0.05) for both sexes.
Abundance and prevalence of infestation by Octolasmis lowei according to the size of the host Libinia spinosa.
Prevalence of infestation by Octolasmis lowei on Libinia spinosa sampled along the northern coast of São Paulo State, Brazil, according to the size and sex of the host. Different letters indicate significant differences (P > 0.05).
There were significant differences (P < 0.05) between prevalence of infestation values among the demographic categories (males, non-ovigerous females and ovigerous females), with the highest values being recorded for ovigerous females, followed by the males, and then by non-ovigerous females (Table 1).
The mean sizes of the hosts according to their sex and moult stage are listed in Table 2. It can be noted that the largest number of individuals were found in the inter-moult stage and that infested crabs had larger mean sizes than non-infested crabs (P < 0.05). Males demonstrated larger mean sizes than females for both infested and non-infested individuals (Table 2).
Mean size (CW in mm) and abundance according to moult condition and host sex (Libinia spinosa). The upper case letters beside values refer to differences between non-infested hosts, and lower case letters refer to differences between infested hosts. Different letters indicate significant differences (P < 0.05). Italicized values refer to the size and abundance of the ovigerous females.
*, represented by a single individual.
A total of 30734 individuals of O. lowei found in the gills of L. spinosa were analysed, being: 24618 non-ovigerous, 4898 ovigerous, and 1115 cyprid larvae. As the values of median intensity of infestation (Table 3) did not differ between the ovigerous and non-ovigerous females categories (P > 0.05), these data were pooled for statistical analyses as ‘females’, and the pooled females were then found to differ from male values (P < 0.05) for this variable.
Median intensity of infestation (epi/host) according to host (Libinia spinosa) sex for all epibiont (Octolasmis lowei) categories. Upper case letters refer to differences between males and females, and lower case letters refer to differences between ovigerous and non-ovigerous females, in each column. Different letters indicate significant differences (P < 0.05).
There were significant differences among the median intensities of infestation for all epibionts categories (cyprid larvae, non-ovigerous and ovigerous) for both host sexes (P< 0.05). The median intensity of infestation on L. spinosa was always higher with males than females (P < 0.05) for non-ovigerous and ovigerous epibionts (Table 3). Cyprid larvae did not show significant differences between the two host sexes (P > 0.05) (Table 3).
As can be observed in Figure 3A & B, increasing host size is followed by an increase in median intensity of infestation by ovigerous and non-ovigerous epibionts (the first three host size-classes were not included due to the absence of infested individuals), and this positive correlation can be corroborated by the results presented in Table 4. However, there was no evident relationship between host size and intensity of infestation by cyprid larvae for either host sexes (Figure 3C; Table 4).
Intensity of infestation by non-ovigerous (A), ovigerous (B) and cyprid larvae (C) of Octolasmis lowei in Libinia spinosa according to sex and mean size of the host. Median; Box: median±25–75%; Whisker: non-outlier range.
Spearman's correlation index between host size (Libinia spinosa) and intensity of infestation by Octolasmis lowei, and between epibiont categories according to the host's sex.
*, P < 0.05; ns, P > 0.05.
Table 4 demonstrates correlation indices between host size (CW in mm) and intensity of infestation among epibionts categories (cyprid larvae, non-ovigerous and ovigerous) for both host sexes. It can be seen that the correlations between epibiont categories are stronger than the correlations between host size and intensity of infestation, and that only the correlations between intensity of infestation values for cyprid larvae and host size were not observed to be significant.
Differences concerning the mean size of hosts (Table 5), percentage of juveniles (CW < 34.5 mm), prevalence (Table 5) and intensity of infestation (Table 6) were found among sampled areas. These results suggest that L. spinosa populations from Ubatuba and Caraguatatuba are similar, and the host population from São Sebastião is composed of larger individuals, with higher indices of infestation and predominance of females.
Total and juvenile host (Libinia spinosa) abundances, mean size (CW in mm) and prevalence of infestation by Octolasmis lowei according to host sex and sampled area. Upper case letters refer to differences between areas (columns), and lower case letters refer to differences between sexes (lines). Different letters indicate significant differences (P < 0.05).
Median intensity of infestation by Octolasmis lowe i according to host sex (Libinia spinos a) and sampled areas. Upper case letters refer to differences between areas (columns), and lower case letters refer to differences between sexes (lines). Different letters indicate significant differences (P < 0.05).
Regarding the distribution of L. spinosa according to depth, 76% of crabs were sampled between 20 and 35 m. At Ubatuba, 63% of hosts were found between 15 and 20 m. At Caraguatatuba, 60% of sampled crabs were distributed between 20 and 25 m. And, at São Sebastião, 58% of crabs were found between 35 and 40 m. The low number of crabs at each depth of the sampled areas did not allow statistical analysis.
The prevalence and intensity of infestation according to sex of the host followed the same pattern in all sampled areas and when analysed together. This indicates that the major influence on these indices may be associated with size and sex of the host.
DISCUSSION
The prevalence of infestation by Octolasmis lowei (63%) on Libinia spinosa that was observed in the present study can be considered high when compared with other studies, although inferior to the 90% and 83% prevalence reported for the lobster Thenus orientalis (Lund, 1793) by Jeffries et al. (Reference Jeffries, Voris and Yang1982, Reference Jeffries, Voris and Yang1984). In a study conducted by Negreiros-Fransozo et al. (Reference Negreiros-Fransozo, Costa and Fransozo1995), prevalence of infestation in Callinectes sp. did not exceed 30%—apparently due to constant host moulting. A similar situation was noted among the smallest individuals examined in the present study, and was also observed by Santos & Bueno (Reference Santos and Bueno2002) in Callinectes danae and C. ornatus, which demonstrated 22.4% and 12.1% prevalence of infestation respectively. As such, prevalence indices have been found to be low in Brazil, in spite of the large variations in prevalence of infestation in portunid crabs described by other authors: 92% in Portunus pelagicus (Gaddes & Sumptom, Reference Gaddes and Sumptom2004); 63.5% in Charybdis callianassa (Herbst, 1789) (Walker, Reference Walker2001); and 85.7% in Charybdis feriatus (Linnaeus, 1758) (Yan et al., Reference Yan, Huang and Miao2004).
As the areas sampled by Negreiros-Fransozo et al. (Reference Negreiros-Fransozo, Costa and Fransozo1995) and in the present study were very similar, the influence of environmental conditioning is attenuated and host behaviour may be the main factor influencing infestation. The low activity of L. spinosa would seem to facilitate the establishment of cyprid larvae, resulting in higher prevalence of infestation values than seen in more agile crabs (Connell & Keough, Reference Connell, Keough, Pickett and White1985; Abelló et al., Reference Abelló, Villanueva and Gili1990).
In the present study, only a small portion of individuals were observed in the pre-moult phase, and these had a prevalence of infestation higher than 30%, and their mean size was close to that of mature crabs. Only a few hosts found in other moult stages (post-moult and recent post-moult) had no established epibionts. Jeffries et al. (Reference Jeffries, Voris and Poovachiranon1992) observed that these moult stages are usually represented by young individuals and that epibionts are rarely found on these young crabs due to their short intermoult period—reflecting the connection between sexual maturity and the initiation of infestation. All of these arguments corroborate the observation that the presence of epibionts is a good indicator of mature size in L. spinosa.
The correlation between host size and intensity of infestation and the 100% infestation rate in the largest host size-categories seen in the present study had been predicted in earlier studies for crabs which have terminal moult (Abelló et al., Reference Abelló, Villanueva and Gili1990; Jeffries et al., Reference Jeffries, Voris and Poovachiranon1992; Voris et al., Reference Voris, Jeffries and Poovachiranon1994). These observations indicate that the length of intermoult period is the main factor facilitating epibiont establishment; however the influence of other factors on infestation control (morphological, physiological or behavioural mechanisms) cannot be ignored.
The significant differences in prevalence of infestation in the different host demographic categories (males, non-ovigerous females and ovigerous females) agree with the proposal of Shields (Reference Shields1992) and Wahl (Reference Wahl2008) that behavioural, morphological, ontogenetic and physiological differences among those categories influence infestation levels. Differences between demographic categories were also examined by other authors (Abelló et al., Reference Abelló, Villanueva and Gili1990; Jeffries, et al., Reference Jeffries, Voris and Poovachiranon1992; Shields, Reference Shields1992; Gili et al., Reference Gili, Abelló and Villanueva1993; Voris & Jeffries, Reference Voris and Jeffries2001), and their main observation was higher infestation in ovigerous females due to their differentiated behaviour.
Similarly, as found by Walker (Reference Walker2001) for Charybdis callianassa, the burrowing behaviour of L. spinosa (Braga et al., Reference Braga, Fransozo, Bertini and Fumis2007) could have a strong influence on infestation indices. The exposed surface area of burrowing crabs is minimized, thus reducing the surface on which cyprid larvae could first settle. As ovigerous females do not have this burrowing habit—probably to maintain the oxygenation levels of their egg masses—they remain more exposed, which facilitates the establishment of larger numbers of epibionts.
Another factor that could influence the differences in characteristics of infestation between sexes would be the host attractiveness to cyprid larvae. Many cyprid larvae are known to be able to locate highly specific substrates by following chemical cues (Foster, Reference Foster and Southward1987; Pawlik, Reference Pawlik1992; Clare, Reference Clare, Schram and Höeg1995). In this way, intensity of infestation by cyprid larvae would be a good indicator of attractiveness. However, the values of intensity of infestation by cyprid larvae showed no differences between sexes, suggesting that there is no visible preference of the larvae with respect to the sex of the host.
Ritchie & Höeg (Reference Ritchie and Höeg1981) observed that the grooming habit of gill cleaning avoided infestation by rhizocephalan cyprid larvae (Lernaeodiscus porcellanae Müller, 1862) in a porcelanidae crab (Petrolisthes cabrilloi Glasell, 1945). Besides the probable respiratory reversion mechanism (Walker, Reference Walker2001), specialized structures (scaphognatites) involved in gill cleaning could represent another important mechanism to prevent fouling in Libinia spinosa. The higher median intensity of infestation of males than of females in all size-classes, the high prevalence of infestation in ovigerous females (in spite of a generally low median intensity of infestation) and the lack of difference in the number of cyprid larvae between sexes indicate the existence of different infestation-controlling mechanisms that could increase cyprid post-settlement mortality, such as higher brachial cleaning efficiency in females.
Differences in infestation intensities by ovigerous epibionts between the two host sexes could be the consequence of a seasonal reproductive variation, density-dependent controlling mechanisms, or epibiont crossing-rates. As Octolasmis is an obligatory cross-fertilizing hermaphrodite (Jens T. Höeg, personal communication), the crossing-rate of these epibionts would be facilitated by the proximity of barnacles in high density aggregations—which is suggested by the positive correlation between non-ovigerous and ovigerous epibionts.
The presence of established conspecifics could be also an attracting factor for cyprid larvae (Pawlik, Reference Pawlik1992; Clare, Reference Clare, Schram and Höeg1995), since a positive correlation was observed between intensity of infestation by cyprid larvae and previously settled epibionts (non-ovigerous and ovigerous epibionts) in the present study. Additionally, in spite of the fact that larger hosts have more epibionts, the association with size is weaker than the association between the number of sessile epibionts (non-ovigerous and ovigerous) and cyprid larvae. As L. spinosa has a terminal moult we can assume that the ‘oldest’ crabs will have higher infestation levels. Rotllant & Takac (Reference Rotllant and Takac1999) observed a great variation in the final size of the crab Libinia emarginata after the terminal (puberty) moult, and this may explain the presence of smaller but older hosts. As such, these older individuals with greater intensities of infestation may be weakening the correlation between intensity of infestation and host size, possibly reflecting a positive attraction between epibionts.
The main differences in rates of infestation between regions refer to characteristics of host populations found in each area. In general, crabs from São Sebastião area had higher prevalence of infestation for being mostly females. The higher intensity of infestation may be associated with larger hosts on this area, since there is more available area for colonization by epibionts. As L. spinosa shows continuous reproduction and wide distribution in the sampled region throughout the year (Braga et al., Reference Braga, Fransozo, Bertini and Fumis2007), the main influence of environmental factors would be related to the availability of O. lowei cyprid larvae in the water column.
In light of the fact that L. spinosa shows high prevalence and infestation intensities by O. lowei when compared with other sympatric hosts (Santos & Bueno, Reference Santos and Bueno2001; Santos, Reference Santos, Escobar-Briones and Alvarez2002; Mantelatto et al., Reference Mantelatto, O'Brien and Biagi2003), and as this crab is very abundant in the study area (Pires, Reference Pires1992; Bertini & Fransozo, Reference Bertini and Fransozo2004; Braga et al., Reference Braga, Fransozo, Bertini and Fumis2007) it appears to have an important role in the maintenance of the adult population of O. lowei and, almost certainly, equal importance in maintaining the larval stock of this epibiont.
The presence of O. lowei in the gills of L. spinosa is probably a good indicator of host maturity. The main factors limiting the infestation observed in this study were: the intermoult length and size of the host; changes in rates of infestation between sexes related to the behaviour and efficiency of antifouling mechanisms. The action of these mechanisms is suggested by the difference in intensity of infestation by sessile epibionts. And, the lack of difference in intensity of infestation by O. lowei cyprid may be a sign of lack of preference to these epibionts between the sexes in L. spinosa.
Additional experimental studies of host and epibiont biology will be needed, however, to elucidate the details of the interaction of population parameters between epibionts and hosts, and to provide more data for new hypothesis and future studies.
ACKNOWLEDGEMENTS
The authors would like to thank FAPESP for their financial support of the BIOTA/FAPESP marine benthic invertebrate project (No. 98/07090-3), sub-project ‘Non-consolidated infra-littoral’; for the scholarship grants to both authors (No. 01/00886-1) and (No. 05/01617-5). We are also thankful to Dr Jens Höeg for valuable comments and to the two anonymous referees for helpful comments that improved this paper.
