Introduction

The Mediterranean Sea is one of the world areas most affected by biological invasions with about 955 introduced species, 153 of these representing crustaceans (Zenetos et al. 2010). Among alien crustaceans, the taxa most frequently recorded in the Mediterranean are Decapoda, followed by Calanoida and Amphipoda (Galil 2011). However, although the number of alien crustaceans has increased noticeably in the last two decades, probably reflecting both an increase in introductions and an interest in their study (Galil 2009), the number of alien crustaceans in the groups of amphipods, cirripedes, cumaceans, isopods and tanaidaceans is still underestimated (Zenetos et al. 2010). Caprellid amphipods, commonly known as skeleton shrimps, are small marine crustaceans that are common in many littoral habitats, where they form an important trophic link between primary producers and higher trophic levels (Woods 2009). The morphology of caprellids, with reduced abdominal appendages which in other amphipods are used for swimming (Takeuchi and Sawamoto 1998) as well as a lack of a planktonic larval stage, suggests that the cosmopolitan distribution of many littoral caprellids is facilitated by the fact that they are often associated with fouling communities on floating objects and vessels (Thiel et al. 2003).

The Mediterranean Sea has one of the best-documented amphipod faunas in the world (Ruffo 1982, 1989, 1993, 1998), but new species are still being described, especially in the case of caprellid amphipods (e.g. Caprella tavolarensis Sturaro and Guerra-García 2011, based on specimens collected from Posidonia oceanica), indicating that further sampling should be conducted to complete our knowledge about Mediterranean caprellids. This is particularly important in the case of fouling communities in harbours and marinas which are still scarcely sampled in some areas of the Mediterranean. Fouling communities include arborescent substrates such as bryozoans and hydroids, which may act as suitable reservoirs for introduced caprellids that have remained unrecorded as yet (Ros et al. 2013). Bellan-Santini and Ruffo (1998) list three caprellid species native to the Mediterranean but known for their propensity for passive dispersal and presence in Mediterranean harbour fouling communities: Caprella acanthifera, C. dilatata and C. equilibra. In 1994, an unusual-looking caprellid, characterized by an acute cephalic projection, was found associated with the fouling community of the wooden piles in the Lagoon of Venice (Sacchi et al. 1998). This caprellid, identified later as Caprella scaura (Templeton 1836) by Sandro Ruffo (Krapp et al. 2006), represented the first and only introduced caprellid reported in the Mediterranean Sea. During the last decade, this Indopacific species has spread very fast across the Mediterranean Sea and has expanded its non-native range to the East Atlantic coast (Sconfietti et al. 2005; Krapp et al. 2006; Galil 2008; Martínez and Adarraga 2008; Ben Souissi et al. 2010; Bakir and Katagan 2011; Guerra-García et al. 2011a; Eleftheriou et al. 2011). In September 2010, an established population of another alien caprellid, the tropical species Paracaprella pusilla Mayer 1890, was found for the first time in European waters, in the fouling community of a marina on the southwest Atlantic coast of Spain (Ros and Guerra-García 2012). This tropical/subtropical species, originally described from Rio de Janeiro, Brazil, by Mayer (1890), was found associated with the native hydroid Eudendrium racemosum.

This study reports the result of a survey on the fouling communities of marinas of the Balearic Islands to determine the presence and quantify abundances of non-indigenous caprellids (NICs) in the Western Mediterranean region. Considering our scant knowledge about the ecology of P. pusilla, some reproductive biology traits were studied for the first time for the species, and its fecundity was compared with the invasive C. scaura. The likely vector and pattern of introduction of P. pusilla in the Mediterranean Sea as well as the species’ current status were analyzed.

Materials and methods

Study area

The Balearic Islands, located in the centre of the Western Mediterranean, are one of the most important tourist destinations in the Mediterranean Sea and are among the preferred destinations for cruise ships crossing the Mediterranean (Minchin et al. 2006). They are characterised by an intense maritime traffic and are a potential hot spot of marine biological invasions (see Drake and Lodge 2004). The region comprises the four main islands of Mallorca, Menorca, Ibiza and Formentera, as well as the small island of Cabrera (Fig. 1).

Fig. 1
figure 1

Map of the Balearic Islands showing sampling stations and presence/absence of NICs (non-indigenous caprellids). See also Table 1

Full size image

Sampling

The sampling programme was conducted between November 2011 and August 2012. A total of 20 recreational marinas along the coast of the Balearic Islands were sampled to ensure a complete review of the total fouling communities which proliferate on artificial hard substrate including pilings, floating pontoons, ropes, buoys, wheels and ship hulls (Table 1). When caprellids were detected in a type of fouling substrate (hydroids, bryozoans or macroalgae), three random replicates of each substrate were taken by hand and fixed in situ in 90 % ethanol. Environmental parameters (water temperature, salinity and turbidity) were measured in situ at each sampling station. Three haphazard measurements were made for each parameter across the floating pontoon system, and mean values and standard deviations were calculated. Salinity and temperature were measured using a conductivity meter CRISON MM40 and turbidity in nephelometric turbidy units (ntu) using a turbidimeter WTW 335 IR.

Table 1 Locations and environmental characteristics of marinas surveyed in the present study
Full size table

Laboratory processing and statistical analysis

All caprellids were sorted and identified to species level. Abundance of caprellids was expressed as number of individuals/1,000 ml of substrate because of the different structure of the substrate types (see Pereira et al. 2006; Guerra-García et al. 2011b). Volume of substrates was estimated as the difference between the initial and final volume when placed into a graduated cylinder with a fixed amount of water. Subsequently, all non-native caprellids collected were photographed on a stereomicroscope Motic K-400L with a Nikon D90 digital camera. Body length of males, mature females (with the brood pouch fully developed) and premature females (with the brood pouch underdeveloped) was measured from the front of the head to the end of pereonite 7, using the PC-based digitizing software Scion Image Alpha 4.0.3.2 © (2000–2001 Scion Corporation). A total of 145 individuals of C. scaura and 106 of P. pusilla were measured. For each non-native caprellid species found, eggs from 15 ovigerous females with the brood pouch completely closed were counted by removing them from the brood pouch with a dissecting needle. To test possible relationships between female size and number of eggs, Pearson’s correlation coefficient was calculated for each species, and differences between the slopes of regression lines of both species were tested using parallelism and equality of lines tests.

Reproductive traits

Five reproductive traits were selected to compare the fecundity of alien species found in the present study (modified after Grabowski et al. 2007): (a) mean size of ovigerous females, (b) brood size (mean number of eggs per brood pouch), (c) maximum number of eggs, (d) partial fecundity index (mean brood size/mean size of ovigerous females) and (e) relative age at reaching maturity (minimal size/mean size of ovigerous females). The comparisons were carried out with alien caprellids from the same region (Mallorca) and collected in the same season (November 2011) to avoid confounding factors.

Results

Two non-native caprellids were found in the Balearic Islands, Paracaprella pusilla and Caprella scaura. The morphological characteristics used to define P. pusilla are (1) the large anterolateral projection of pereonite 2, (2) small dorsal tubercle on pereonite 2, (3) proximal knob on the basis of gnathopod 2 and (4) lateral pleura in pereonites 3 and 4, especially developed in pereonite 3 (see plate 2, Figs 36 and 37 in Mayer (1903)). Individuals collected in Mallorca and Ibiza display these features (Fig. 2). Drawings of P. pusilla from different world areas (Guerra-García 2006 from Colombia, Guerra-García et al. 2010 from India and Díaz et al. 2005 from Venezuela) and our own examination of specimens from the Gulf of Mexico, Brazil, India, Southern Spain and the Balearic Islands showed little intraspecific variation in morphology. The morphological characteristics used to identify C. scaura from the Mediterranean are (1) cephalon with an acute, bent forward, dorsal projection, (2) pereonites 1 and 2 elongate in males, (3) basis of gnathopod 2 long but shorter than pereonite 2 and (4) absence of ventral projection between the insertion of gnathopods 2 (Templeton 1836; Mayer 1890; Krapp et al. 2006). Individuals collected in Mallorca and Menorca display these features and are similar to other populations from the Iberian Peninsula, the Canary Islands, Italy and Greece, which were examined by the authors. Both alien species are easily differentiated from native species common in fouling communities (Ros et al. 2013).

Fig. 2
figure 2

a Lateral view of an adult male and an adult female of Paracaprella pusilla collected from Spain; b detail of the lateral pleura in pereonite 3 (see arrow) of an adult male; c detail of the anterior part of an adult male showing the small dorsal tubercle, the large anterolateral projection of pereonite 2 and the proximal knob on the basis of gnathopod 2 (see arrows); d detail of the gnathopod 2 of an adult male

Full size image

Alien caprellids were present in the three islands of Mallorca, Menorca and Ibiza and were absent from the islands of Formentera and Cabrera (Fig. 1). Paracaprella pusilla was found in Palma marina (Mallorca; 39°34′N, 2°38′E) and Ibiza marina (Ibiza; 38°54′N, 1°26′E). The species was found in a water temperature range from 20.5 °C (Ibiza) to 21.5 (Mallorca), a salinity range from 33.4 (Mallorca) to 35.3 (Ibiza) and a turbidity range from 2.2 ntu (Ibiza) to 3.8 ntu (Mallorca) (Table 1). In both marinas, P. pusilla was found associated with the hydroid Eudendrium racemosum where it exhibited similar densities (4,611 ± 2,204 ind/1,000 ml in Mallorca and 4,100 ± 2,055 ind/1,000 ml in Ibiza, Mean ± Standard Error). In Palma marina (Mallorca), the species was also found with the hydroid Pennaria disticha (Table 2). The maximum total length recorded for males was 8.2 mm, whereas for females, the maximum was 5.6 mm (Fig. 3). Caprella scaura was found in three marinas located in the northeast coast of Mallorca: Cala Ratjada (39°43′N, 3°28′E), Cala Bona (39°37′N, 3°23′E) and Porto Colom (39°25′N, 3°15′E) and in one marina in Menorca (Mahón; 39°52′N, 4°18′E). The species was found in a water temperature range from 16.3 °C (Porto Colom, Mallorca) to 26.3 °C (Menorca), a salinity range from 36.0 (Cala Bona, Mallorca) to 37.6 (Porto Colom, Mallorca) and a turbidity range from 1.7 ntu (Menorca) to 36.3 ntu (Porto Colom, Mallorca) (Table 1). Caprella scaura was associated with eight different substrates, including hydroids, bryozoans and macroalgae of the marinas’ fouling community (Table 2). The highest abundance was found in Cala Ratjada, associated with the bryozoan Bugula neritina (18,333 ± 8,647 ind/1,000 ml). The maximum total length recorded for males was 13.2 mm, whereas for females, the maximum was 7.9 mm (Fig. 3). The large individuals of both sexes were found associated with bryozoans.

Table 2 Density of non-indigenous caprellids (C. scaura and P. pusilla) found on different fouling species and in different locations of the Balearic Islands
Full size table
Fig. 3
figure 3

Box-and-Whisker plot for each sex/age group measured for the different populations. Median values are included; the rectangles contain values between the first and the third quartiles; the bars connect the extreme values

Full size image

Regarding the reproductive traits, we found that the mean number of eggs per female was 29.07 for P. pusilla and 26.67 for C. scaura (Table 3). The partial fecundity index was 7.20 for P. pusilla and 5.32 for C. scaura, and the maturity index was 0.66 and 0.73 for P. pusilla and C. scaura, respectively. A significant correlation was found between female size and number of eggs for both species (P. pusilla: r = 0.62, p < 0.05; C. scaura: r = 0.96, p < 0.01) (Fig. 4). Although the parallelism test did not show differences between the slopes of regression lines of the two species (F 1,26 = 0.006, p = 0.94), the equality of lines test showed significant differences between C. scaura and P. pusilla (F 2,26 = 10.89, p = 0.0004). The graph shows that for a given body size, females of P. pusilla had a higher number of eggs than females of C. scaura.

Table 3 Reproductive traits for non-indigenous caprellid populations collected at Mallorca in November 2011
Full size table
Fig. 4
figure 4

Correlations between female size and number of eggs per brood in Paracaprella pusilla and Caprella scaura collected at Mallorca

Full size image

Discussion

Non-native caprellids from the Balearic Islands and possible vectors of their introduction

We recorded, for the first time, the presence of the invasive C. scaura at the Balearic Islands as well as the presence of the introduced P. pusilla in the Mediterranean Sea, which represents the first record of the genus Paracaprella in the Mediterranean. These species were present at the islands of Mallorca, Menorca and Ibiza but were absent from the small islands of Formentera and Cabrera. Caprella scaura was found associated with a wide variety of fouling substrates, including macroalgae, hydroids and bryozoans, thus showing a high plasticity to colonize different habitat structures, while P. pusilla was only found associated with fouling hydroids. This pattern of habitat use was also found for both species in the fouling community of a marina in southern Spain (Ros et al. 2013) where P. pusilla was positively correlated with the native hydroid E. racemosum (Ros and Guerra-García 2012), reflecting a clear preference for hydroids in the non-native area. Although Caprella species have been found to survive transport in ballast tanks (Carlton 1985), for a fouling species frequently recorded from ports, transport via hull fouling is assumed to be the most probable vector (Galil 2011). As both P. pusilla and C. scaura were found associated with the fouling communities adherent to artificial hard substrates including ship hulls, ship fouling is assumed to be the most probable vector for the introduction of the species to the Balearic Islands. This may be related to the absence of both species from the islands of Formentera and Cabrera which have only few ports (2 and 1, respectively; FEAPDT 2011) and are therefore exposed to much lower boating pressure than the islands of Mallorca (39 ports), Menorca (9 ports) and Ibiza (8 ports).

Reproductive traits

Reproduction appears to be a major factor in the success of invasive amphipods (Weis 2010). Grabowski et al. (2007) studied six reproductive and two additional traits (salinity tolerance and tolerance to human impacts) to compare six invasive versus seven native gammarid species occurring in Central European waters. They found that invasive gammarids were characterized by a combination of large brood size, high partial fecundity, early maturation and by the appearance of higher number of generations per year. In the present paper, we studied four of the six reproductive traits and an additional one (maximum number of eggs) in the newly introduced P. pusilla and C. scaura, an invasive species which has been spreading very fast across the Mediterranean and the East Atlantic coast (Guerra-García et al. 2011a). We found that C. scaura and P. pusilla females produce a larger mean number of eggs when compared with native species from the Mediterranean Sea with similar female size such as C. grandimana, with an average brood size of 7.6 eggs (Baeza-Rojano et al. 2011). When comparing P. pusilla with C. scaura, we found that P. pusilla produces more eggs per brood than C. scaura and has a higher partial fecundity index. This implies that for a given size of the female, P. pusilla has a higher number of eggs than C. scaura. However, as females of C. scaura can attain larger body sizes, the maximum number of eggs per female was higher in this species. Moreover, the maturity index and thus the relative age at reaching maturity are less in P. pusilla than in C. scaura. These traits may facilitate the secondary spread of the P. pusilla to new areas of the Mediterranean as has already happened with C. scaura.

Current status of P. pusilla

Similar to other alien caprellids in Europe such as Caprella mutica in Scotland (Willis et al. 2004), the non-indigenous status of Paracaprella pusilla in European waters can be assessed using the criteria of Chapman and Carlton (1994): (1) previously unknown in local region; (2) post-introduction range expansion; (3) associated with a human dispersal mechanism; (4) associated with or dependent on other introduced species; (5) associated with artificial environments; (6) restricted or discontinuous distribution in the region; (7) disjunct global distribution; (8) insufficient life history adaptations for natural global dispersal; and (9) exotic evolutionary origin. Paracaprella pusilla scores positively on criteria 1, 2, 3, 5, 6, 7, 8 and 9, suggesting it to be an alien species to the area. As this is the first record of the genus Paracaprella in the Mediterranean Sea, it increases the known diversity of the Caprellidea in this region.

Global distribution of P. pusilla

According to Mayer (1903), the species’ natural area of distribution is the Atlantic coast of Central and South America. Most records of P. pusilla are from the Gulf of Mexico and the Caribbean coast (Ros and Guerra-García 2012), and the species is one of the most abundant caprellids along the Caribbean coast of Venezuela and Colombia (Díaz et al. 2005; Guerra-García 2006). Therefore, the species appears to have a strong Caribbean affinity (Carlton and Eldredge 2009). Nevertheless, the origin of Paracaprella pusilla is unknown (Mead et al. 2011). Records on geographically disjunct occurrences of P. pusilla date back to the early 1900s, a short time after the species had been described by Mayer in 1890 (see Ros and Guerra-García 2012). This, along with the facts that most of the records refer to fouling communities of harbours, and that the species may be able of travelling long distances attached to vessel hulls, has prevented a clear determination of the origin of the species. Actually, the species’ global area of distribution includes the Atlantic coasts of Central and South America, tropical West Africa, East Africa, Hawaii, India, Australia and the southwest coast of Spain.

Introduction pattern of P. pusilla to the Mediterranean Sea

Due to a lack of previous studies on caprellids associated with fouling communities in the study area, the exact time of introduction to this site remains unknown for both alien species. However, extensive biological surveys in the Mediterranean over the twentieth century allow for a reasonable measure of confidence in separating alien and native biota (Galil 2009). Never recorded in the Mediterranean waters before, P. pusilla is not mentioned neither in the handbook of the Mediterranean amphipods fauna (Ruffo 1993) nor in the study by Guerra-García et al. (2011a, b) on the intertidal and shallow water caprellids of the Iberian Peninsula. Moreover, Caprella scaura and P. pusilla were not recorded neither in a study on amphipods of Ibiza (Ballesteros et al. 1987) nor in a more recent study on amphipods of Mallorca (Box 2008). Therefore, the introduction of these caprellids to the European waters of southern Spain and the Western Mediterranean Sea may have occurred during the last decade. The date of the introduction of the inoculum is significant for the study of the patterns and processes of invasion but is extremely difficult to ascertain for unintentional or undocumented intentional introductions (Galil 2011).

There are two main alternatives to explain the presence of P. pusilla in the Mediterranean Sea (Fig. 5): The species entered (a) via the Suez Canal (Port Said) on vessels from the Indo-Pacific, or (b) through the Strait of Gibraltar, on vessels arriving from the Atlantic coast of America or from the established population in southwest Spain. The presence of P. pusilla in the Suez Canal was only reported by Schellenberg (1928), who recorded the species in three stations: Kantara (46 km from Port Said), Kabret (between Little Bitter Lake and Great Bitter Lake) and Port Taufiq. This seems to support hypothesis (a). However, so far, the species has not been recorded neither in the Red Sea nor along the Mediterranean Sea, and recent studies on the fouling community in the Suez Canal by Emara and Belal (2004), including Little Bitter Lake, Kabret, Great Bitter Lake and Port Taufiq, reported only the presence of the caprellid species Caprella equilibra, which was also the only caprellid species found in the fouling communities of the Suez Canal by El-komi (1998). Probably, the population found by Schellenberg in 1928 did not succeed in adapting fast enough to the new environment and failed in spreading to adjacent areas. Moreover, the absence of P. pusilla in the Red Sea and its relatively recent record in the Indian Ocean (Sivaprakasam 1977) suggest that the populations recently found in European coastal waters originate from the Atlantic coasts of Central and South America, where the species is highly abundant. In this case, an introduction through the Strait of Gibraltar (hypothesis b) would be more probable than the alternative (hypothesis a). Interestingly, many small craft of Mallorca overwinter in marinas in the south of Spain (Minchin et al. 2006) and thus could represent a suitable vector for the secondary spread of the species from the established population of southern Spain to the Balearic Islands. Marinas seem to provide a network of suitable habitats for the secondary spread of a species via recreational yachting activity (Ashton et al. 2006).

Fig. 5
figure 5

Distribution map of Paracaprella pusilla in the Mediterranean Sea with years of first record for the different areas. Arrows indicate two possible ways of introduction to the Mediterranean Sea (see text)

Full size image

Unlike what happened when Schellenberg found the species in the Suez Canal in 1928, the last decades of the twentieth century saw pronounced thermal fluctuations and a significant increase in the average seawater-surface temperature in the Mediterranean (Nykjaer 2009). This may favour survival, growth and reproduction of tropical aliens, giving them a distinct advantage over native temperate Mediterranean taxa (Galil 2011). Along with the increasing role of the Mediterranean as a hub of international commercial shipping (Dobler 2002), this might explain the fact that P. pusilla has successfully reached the Western Mediterranean Sea only most recently.

The precautionary principle suggests to considering each alien species ‘guilty until proven innocent’ and calls for analyzing possible impacts on native communities (Occhipinti-Ambrogi et al. 2011). Taking into account that the occurrence of P. pusilla in the Mediterranean Sea is probably a consequence of secondary spread from the established population in the Strait of Gibraltar, and that the fraction of alien species that spread following establishment is considered one of the measures of invasion success (Galil 2011), the presence of P. pusilla at the Balearic Islands suggests a future invasion along marinas of the Mediterranean Sea.