Effects of the invasive macroalga Lophocladia lallemandii on the diet and trophism of Pinna nobilis ( Mollusca : Bivalvia ) and its guests Pontonia pinnophylax and Nepinnotheres pinnotheres ( Crustacea : Decapoda )

Pinna nobilis and its guests Pontonia pinnophylax and Nepinnotheres pinnotheres were sampled from Posi­ donia oceanica meadows invaded and non-invaded by the red alga Lophocladia lallemandii. understanding the interactions among native and invasive species was the primary aim. Therefore, the effects of the invasive L. lallemandii on the percentage contribution of the food sources of P. nobilis and its guests and their trophic associations were investigated by applying mixing models to the δ13C and δ15n stable isotopes. isotopic analyses revealed that the guests of P. nobilis occupied a higher trophic level than their host due to their capacity for food selection. The invasion of L. lallemandii altered the percentage contribution of the food sources to the consumers’ diet. Whereas the percentage contribution of the food sources to the δ13C signatures did not really change, L. lallemandii produced variations in the percentage contribution to the δ15n signatures. This red macroalga represented one of the main food sources for P. nobilis, with a real contribution of 19.4% in the invaded meadows. Moreover, L. lallemandii slightly decreased the trophic level of the consumers. This study provides the first isotopic and trophic data for the pea crab N. pinnotheres.

inTRoduCTion interactions between native and introduced species represent a threat to biodiversity and ecosystem functioning (galil, 2007). in fact, competitive exclusion by invasive species has been reported as a cause for the progressive regression of seagrasses (Williams, 2007).These important habitats are very sensitive to environmental degradation and physical disturbance (Hemminga and duarte, 2000).The most widespread seagrass species in the Mediterranean sea is Posidonia ocea nica (l.) delile.P. oceanica communities include primary producers such as red algae, epibionts and vagile biota (Templado et al., 2004).Faunal communities associated with seagrass encompass many taxonomic groups, including bivalves such as Pinna nobilis linnaeus (1758) (garcía-March et al., 2002).The bivalve P. nobilis is a long-lived species (butler et al., 1993;galinou-Mitsoudi et al., 2006) that is endemic in the Mediterranean sea, being one of the largest bivalves in the world (garcía-March, 2003).However, it is an endangered species (eeC 1992;Centoducati et al., 2007) due to the increasing human impact on the seabed (Richardson et al., 1999).The fan mussel P. nobilis is a benthic suspension feeder that can colonise different substrates such as bare sand bottoms (Katsanevakis, 2007) and soft-bottom areas overgrown by seagrass meadows at depths ranging from 0.5 to 60 m (Zavodnik et al., 1991). in these areas, the fan mussel lives partially buried in the sand, usually anchored among P. oceanica shoots and hidden by the leaves of the canopy (garcía-March et al., 2002).
invasive species represent a risk to natural ecosystems by damaging biodiversity and altering the structure and functioning of ecosystems (boudouresque and verlaque, 2002;Macdougall and Turkington, 2005).The red macroalga Lophocladia lallemandii (Montagne) F. schmitz is an alien species that was introduced into the Mediterranean through the suez Canal and is widespread throughout the tropics and subtropics (boudouresque and verlaque, 2002).L. lallemandii grows on a wide range of substrates (Patzner, 1998;ballesteros, 2006;ballesteros et al., 2007), including P. oce anica meadows (ballesteros et al., 2007), where it induces shoot mortality and affects the invertebrate community (Patzner, 1998;ballesteros, 2006).Therefore, invasion by L. lallemandii affects the characteristics of microhabitats and faunal communities (Piazzi et al., 2002).
analysis of isotopic composition has become an effective method for studying trophic food webs (Pinnegar and Polunin, 2000;Fisher et al., 2001), since organisms assimilate stable carbon and nitrogen isotopes from their food sources (Pinnegar and Polunin, 1999).The values of stable δ 13 C isotope have mostly been used to indicate primary food sources, whereas values of stable δ 15 n isotope have allowed to determine the trophic level (Post, 2002).isosource software (Phillips and gregg, 2003) has made it possible to quantify the source contribution to a mixture by the application of mixing models (Phillips et al., 2005).Previous studies using stable isotopes (δ 18 o-δ 16 o; Mg:Ca-sr:Ca ratios; skeletal δ 18 o-skeletal δ 13 C) in the shell of the fan mussel P. nobilis shell have mainly been performed to estimate growth and age (Richardson et al., 1999) and to reconstruct sea surface temperatures and ontogenetic records of metabolic Co 2 (Kennedy et al., 2001b).However, only in one study that analysed δ 13 C and δ 15 n signatures was it demonstrated that the fan mussel and its guest shrimp P. pinnophylax had assimilated similar food sources and, consequently, were at a similar trophic level (Kennedy et al., 2001b). in contrast, studies on the diet and trophic relationships of P. nobilis and its guest pea crab N. pinnotheres have not been undertaken.
Therefore, this study aimed mainly to determine the effect of invasive L. lallemandii on the contribution of food sources to the diet of the fan mussel P. nobilis and its two guests P. pinnophylax and N. pinnotheres and their trophic associations.

Sample collection
at each site, ten linear transects of 30x3 m were sampled by scuba diving.The sites were separated by hundreds of metres.all transects were laid over seagrass beds at 7-10 m depth (except for the esponja sampling station, which was at a mean depth of 22.5 m); sand patches were also present.a total of 2700 m 2 were sampled for each treatment (P.oceanica invaded and non-invaded).according to the european Council directive 92/43/eeC on the conservation of natural habitats and wild fauna and flora, the fan mussel P. nobilis is listed as an endangered species and is under strict protection (annex iv), and all forms of deliberate capture or killing of this bivalve are prohibited (eeC 1992;Centoducati et al., 2007).Therefore, in order to minimise the impact on local populations, only 24 individuals of P. nobilis were randomly collected (11 individuals in a P. oceanica meadow invaded by L. lallemandii and 13 individuals in a non-invaded meadow) in summer 2007 by experienced scuba divers under license from the appropriate institutions (government of the balearic islands).The size range of the samples was kept as constant as possible.
The fan mussels were quickly transferred to the laboratory where they were carefully dissected ac- cording to the procedure described by yonge (1953).
For each P. nobilis, the maximum shell width (W m ), maximum shell length (l m ) and maximum length of the posterior adductor muscle scars (l ad ) were measured (Rabaoui et al., 2007).age was determined by counting the number of abductor muscle scar rings (R) on the shell.However, because the muscle scar ring of the first year is absent or inconspicuous (Richardson et al., 1999), the age was estimated as the number of rings plus one. in addition, the presence or absence of decapod guests was recorded for each fan mussel.The muscle of each P. nobilis and the muscle of the associated guests (P.pinnophylax and N. pinnotheres) were extracted.after extraction, tissues were immediately frozen and stored until further processing for stable isotopic analysis (δ 13 C and δ 15 n).Potential common food sources were also collected: particulate organic matter (PoM), sediment organic matter (soM), P. oceanica leaves as a representation of the potential contribution of its remains, epiphytes of P. oceanica leaves (ePol) and the invasive alga L. lallemandii.Water samples (10 l) for PoM determination were collected at the maximum depth (∼10 m) and filtered through pre-combusted fibreglass filters (Whatman gF/C) at 450°C for 4 h.soM was collected by scuba divers from a quadrat surface of 400 cm 2 using steel quadrats.ePol samples were manually isolated from P. oceanica leaves at the laboratory using stainless-steel tools, and were pooled and treated as a single sample for each meadow (Cardona et al., 2007).soM and ePol were acidified with HCl (2 n) drop-by-drop (the cessation of bubbling was used as the criterion to determine the amount of acid to add) and left for 3 h (Carabel et al., 2006).

Isotopic analysis and processing
Muscle and food source samples were dried at 60ºC to constant weight and then ground to a fine powder using a mortar and pestle.Homogeneous dried powder (2 ± 0.1 mg) of each sample was placed in cadmium tin cups and then combusted to determine 13 C and 15 n stable isotope composition by continuous flow isotope ratio mass spectrometry (CF-iRMs) using a THeRMo delTa X-Plus mass spectrometer.samples of an internal reference material were analysed after every eight samples to calibrate the system and to compensate for drift over time.The global standard for δ 13 C is Co 2 and for δ 15 n is atmospheric nitrogen.The reference material used for the stable isotopes analyses was the bovine liver standard (bls: 1577b; u.s.department of Commerce, national institute of standards and Technology, gaithersburg, Md 20899).The analytical precision was based on the standard deviation of bls replicates: 0.02‰ and 0.10‰ for δ 13 C and δ 15 n, respectively.stable isotope abundances were measured by comparing the ratio of the most abundant isotopes ( 13 C: 12 C and 15 n: 14 n) in the samples with the international isotopic standards.Carbon and nitrogen stable isotope ratios were expressed in δ notation in terms of parts per thousand (‰) deviations from the standards, according to the following equation: where X is 13 C or 15 n and R is the corresponding 13 C/ 12 C and 15 n/ 14 n ratio.
in order to determine the trophic level of the organisms, the following formula was applied: Trophic level = (δ 15 n consumer -δ 15 n First trophic level )/ 3.4+1 where 3.4% is the assumed δ 15 n trophic enrichment factor (le loc 'h et al., 2008).

Food source contribution (%)
in order to determine the percentage contribution of the food sources to the fan mussel P. nobilis and its guests P. pinnophylax and N. pinnotheres, feasible contributions for every source were estimated by isotope mixing models using isosource software version 1.3.1 (Phillips and gregg, 2003).The red macroalga L. lallemandii was considered as a feasible food source in order to analyse its relative contribution and its effects on the diet of the consumers P. nobilis and its guests P. pinnophylax and N. pinnotheres, regardless of the situation (invaded or non-invaded).by calculating the difference between the hypothetical contribution of L. lallemandii in a non-invaded meadow and the L. lallemandii contribution in an invaded meadow, the real contribution of L. lallemandii was obtained.
The model was used to estimate the potential contributions of the primary producer groups to the fan mussels and their guests.at increments of 1% and a tolerance of 0.1, the mean of the isotopic values, 1 st to 99 th percentiles, and the range of probable contributions to the organisms were determined for every primary producer (decottignies et al., 2007;ince et al., 2007;Pitt et al., 2008). in the absence of consumer-specific isotope discrimination factors for these two consumers, an assumed discrimination of 0.5‰ was applied for carbon (Pitt et al., 2008).nitrogen could not be incorporated into the model since the discrimination levels were unknown, and applying an assumed 3-4‰ value did not yield results because the values of the consumers lay outside the polygon created by the potential sources (ince et al., 2007;Pitt et al., 2008).

Statistical analysis
statistical analysis was carried out using sPss  (v.16.0 for Windows  ).one-way anova was performed in order to analyse the δ 13 C and δ 15 n isotopic signatures in the P. nobilis individuals inhabiting P. oceanica meadows invaded by the red alga L. lallemandii and the fan mussels that occurred in non-invaded seagrasses.one-way anova was also used to determine the δ 13 C and δ 15 n differences be-tween the P. pinnophylax individuals that colonised P. nobilis inhabiting invaded meadows and those in non-invaded meadows.

Biometry of Pinna nobilis and presence/absence of guests
The mean values of maximum shell width (W m ), maximum shell length (l m ), maximum length of the posterior adductor muscle scars (l ad ) and number of muscle scar rings (R) of the 24 P. nobilis (11 in invaded seagrass and 13 in non-invaded seagrass) were 16.12 ± 0.61, 37.71 ± 2.44, 18.28 ± 1.42 and 5.45 ± 0.53 (mean ± se), respectively (Table 1).
The fan mussels P. nobilis hosted the shrimp P. pinnophylax in 54.2% and the pea crab N. pinnoth eres in 8.3% of the samples.P. pinnophylax couples (female and male) were present in 33.33% and a solitary shrimp in 20.8% of the samples.both guests (shrimp and pea crab) were present in 4.2% of the

Stable isotopes
The δ 13 C and δ 15 n mean values (±se) are shown in Figure 2. The consumers P. nobilis and its guests P. pinnophylax and N. pinnotheres showed little variation in the δ 13 C values.For δ 5 n values, these consumers showed more variation, being ranked as follows: N. pinnotheres > P. pinnophylax > P. nobi lis.The isotopic δ 13 C and δ 15 n signatures for invaded (fan mussels from seagrass invaded by L. lalleman dii) and non-invaded (fan mussels from non-invaded P. oceanica meadow) P. nobilis were similar.The δ 13 C and δ 15 n values for invaded and non-invaded P. pinnophylax were also similar.nevertheless, the δ 13 C and δ 15 n isotopic signatures of invaded P. no bilis and P. pinnophylax showed values below the isotopic signatures of non-invaded P. nobilis and P. pinnophylax (Fig. 2 and Table 2).statistical analysis demonstrated differences between invaded and non-invaded P. nobilis δ 15 n signatures (anova, P<0.05). in contrast, the δ 13 C isotopic signatures of invaded and non-invaded P. nobilis were not significantly different (anova, P>0.05).no differences between δ 13 C and δ 15 n isotopic signatures were found in the shrimp P. pinnophylax (invaded and non-invaded) (anova, P>0.05).
The food sources ePol and PoM ranged in the same group constituted by the consumers P. nobi lis, P. pinnophylax and N. pinnotheres.ePol was positioned near the fan mussel P. nobilis and PoM  was located near its guests (Fig. 2 and Table 2).The invasive red alga L. lallemandii was only sampled in sa dragonera where the seagrass was invaded by this red alga, and contained the lowest δ 13 C isotopic value. in contrast, P. oceanica was the most enriched in δ 13 C. in relation to the δ 15 n isotopic values, similar results were found in these food sources.soM was located near P. oceanica (Fig. 2 and Table 2).Two trophic levels were established (Table 2).The first included the primary producers L. lallemandii and P. oceanica, soM and primary consumers such as ePol and invaded and non-invaded P. nobilis; the second incorporated the fan mussel guests (invaded and non-invaded P. pinnophylax and N. pinnotheres) and PoM.

Food source contributions
isosource results determined the percentage contribution of the potential food sources to the fan mussel P. nobilis and its guests P. pinnophylax and N. pinnotheres from P. oceanica meadows, either invaded or non-invaded by the red macroalga L. lal lemandii (Fig. 3).
For P. nobilis, the percentage contribution of each food source to the δ 13 C signature was almost identical in the two situations (invaded and non-invaded).The greatest contribution based on δ 13 C came from the ePol (24.4%), followed by PoM, soM and P. oceanica particulate remains.The red macroalga L. lallemandii made a minor contribution to the P. nobilis (invaded and non-invaded) δ 13 C value.However, L. lallemandii made a major contribution to the P. nobilis δ 15 n signature (47.2% invaded and 27.8% non-invaded).Consequently, the real contribution of L. lallemandii to the P. nobilis δ 15 n signature was 19.4% (Fig. 3).invaded and non-invaded P. nobilis showed differences in the percentage contribution of the food sources based on the δ 15 n signature.The highest contributions to the δ 15 n signature in invaded mussels were made by L. lallemandii and P. oceanica particulate remains, showing an overall contribution of 83.7%.For the non-invaded P. no bilis δ 15 n signature, L. lallemandii and P. oceanica particulate remains represented 55.5% of the contribution.soM also made a significant contribution (26.4%) to non-invaded fan mussels (Fig. 3).
on the other hand, in the shrimp P. pinnophylax δ 13 C signature, the food sources soM, P. oceanica, ePol and PoM made similar contributions within the range of 19.1-24.3%.L. lallemandii made a minor contribution to P. pinnophylax (invaded and non-invaded) in the δ 13 C signature. in terms of the δ 15 n signature, the epiphytes of P. oceanica leaves made a major contribution (26.0%) to invaded P. pinnophylax.PoM and ePol made the highest contribution to non-invaded P. pinnophylax at 31.8% and 24.1%, respectively.L. lallemandii showed little real contribution (4.1%) to the P. pinnophylax δ 15 n signature (14.0%non-invaded shrimp, 18.1% invaded shrimp) (Fig. 3).
in the pea crab N. pinnotheres, strong differ- ences in the contribution percentages were observed between δ 13 C and δ 15 n signatures.all food sources contributed to the δ 13 C signature, with a range of 16.6-25.1%,the highest being from the ePol. in contrast, PoM (80.5%) made the highest contribution to the δ 15 n signature.

disCussion
The invasion of L. lallemandii altered the δ 15 n percentage contribution of the food sources to the consumers' diet and slightly decreased the trophic level.The present study revealed that the guests of P. nobilis (P.pinnophylax and N. pinnotheres) occupied a higher trophic level than their host.The food selection capacity of the guests may allow them to increase their trophic level.isosource results showed that the epiphytes of P. oceanica leaves made the highest contribution to the δ 13 C signatures of P. no bilis and its guest N. pinnotheres.For P. pinnophylax the food sources soM, P. oceanica, ePol and PoM made similar contributions; L. lallemandii made a minor contribution.in contrast, according to the δ 15 n isotopic values, PoM and ePol represented the main food sources for N. pinnotheres and P. pin nophylax, respectively.in the case of the P. nobilis δ 15 n signature, L. lallemandii and P. oceanica were the main food source, with L. lallemandii presenting a real contribution of 19.4% in the invaded meadow.
several studies have reported seasonal, spatial, food web, and organism size variations and fractionation among tissues as a consequence of differences in stable isotope signatures (decottignies et al., 2007).The experimental design of the present work avoided the interference of factors such as geographical differences or physicochemical water parameters, since the samples were collected from the same extended area and during the same period.Moreover, all P. nobilis had comparable dimensions.Therefore, the study design avoided biases in the stable isotopic values as far as possible.
The isotopic δ 13 C and δ 15 n isotopic values obtained in the present study were within the range of previous studies (Kennedy et al., 2001a;Cardona et al., 2007;deudero et al., 2009).The variations in some isotopic values such as those in ePol occurred because seagrass epiphyte δ 15 n values can vary by 3‰ depending on the proportion of animals and plants that form this epiphyte community (Kennedy et al., 2001a).according to the δ 15 n isotopic results obtained in the current work, which determine the trophic level (Post, 2002), the guests of P. nobilis (P.pinnophylax and N. pinnotheres) occupied a higher trophic level than their host.P. pinnophylax and N. pinnotheres showed more enriched δ 15 n isotopic values than their host.The difference in δ 15 n between P. nobilis and P. pinnophylax was 1.97‰, and between P. nobilis and N. pinnotheres it was 3.49‰.However, the fan mussel guests had similar δ 13 C isotopic values.a slight increase in the δ 15 n isotopic value of P. pinnophylax with respect to its host P. nobilis has previously been reported (Kennedy et al., 2001b).nonetheless, this study provides the first isotopic and trophic data for the pea crab N. pinnotheres, which had the highest δ 15 n enrichment of the two guests.The isotopic enrichment of the fan mussel guests in the present work was consistent with the estimated trophic level of the consumers (le loc 'h et al., 2008).The capacity for food selection may explain the differences in the contribution of the different food sources to their diets, since the morphology of both studied guests allows them to select food sources (Rabaoui et al., 2008).a previous study also demonstrated that the capacity for food selection generated variations in isotopic composition (decottignies et al., 2007).
since P. nobilis feeds on suspended materials (Kennedy et al., 2001b), L. lallemandii could become a new potential food source that contributes to the fan mussel δ 15 n isotopic signatures; however, P. nobilis guests can select food to avoid feeding competition, as reported in other species with the capacity for food selection (decottignies et al., 2007). in fact, a previous investigation suggested the need to include the qualitative selection capacities of consumers in future interpretations of trophic relationships in marine coastal ecosystems (decottignies et al., 2007).isosource results showed that ePol made the highest contribution to the δ 13 C signatures of all studied consumers, while for δ 15 n isotopic values, PoM and ePol represented the main food sources for N. pinnotheres and P. pinnophylax, respectively.There was greater variability in the contribution of food sources to P. nobilis δ 15 n isotopic values, with L. lallemandii and P. oceanica remains being the main food sources in the P. oceanica meadow invaded by L. lallemandii.in agreement with these results, a previous study reported food source contributions in a non-invaded P. oceanica meadow in which ePol contributed to the P. pinnophylax diet and P. oceanica remains were a potential food source for P. nobilis (Kennedy et al., 2001b).
The invasion of L. lallemandii altered the percentage contribution of the food sources to the consumers' diet. in fact, it has been reported that the incorporation of new sources resulted in changes in the contribution of food sources to an organism's diet (Phillips and gregg, 2003).Whereas the percentage contribution of the food sources to the δ 13 C signatures did not really change, L. lallemandii produced variations in the percentage contribution to the δ 15 n signatures.differences in the percentage contribution between δ 13 C and δ 15 n signatures by food sources such as PoM and soM have been reported previously (sarà et al., 2004).The contributions to δ 15 n signatures of invaded versus non-invaded P. pinnophylax showed little variation, almost avoiding the effect of the red macroalga L. lallemandii (4.1% of real contribution), perhaps due to its capacity for food selection (Rabaoui et al., 2008). in invaded versus non-invaded P. nobilis, the food sources showed variations in their percentage contribution to the δ 15 n signatures, with L. lallemandii making a real contribution of 19.4% in P. nobilis located in the invaded seagrass meadows.Therefore, invasion by L. lallemandii led to a slight decrease in the trophic level of the consumers P. nobilis and the shrimp P. pinnophylax.
Changes in the trophism of P. nobilis in a P. oceanica meadow invaded by L. lallemandii could be related to the physiological effects of the lophocladines, bioactive alkaloids from the genus Lo phocladia (gross et al., 2006).The invasive alga L. lallemandii has an injurious effect on the sea grass P. oceanica meadow (ballesteros et al., 2007) and contains lophocladines (sureda et al., 2008) that might reverberate in the physiology of P. nobilis, as demonstrated for other organisms (boudouresque and verlaque, 2002; ballesteros, 2006; sureda et al., 2006).Future studies are thus required to determine the physiological effect that the invasive red alga L. lallemandii might induce in endemic species.

aCKnoWledgMenTs
The authors acknowledge the collaboration in isotopic analysis offered by the scientific and Technical services (uib), especially the support offered by M. Ribas (iuniCs).We are also grateful to the logistical support from e. Álvarez, s. sardu, a. Martín, y. bertrand, j. jiménez and a. box.additionally, we thank d. March for his cartographic support.This study was partly financed by, and the first author received a fellowship from, the acció especial "desarrollo de técnicas para evaluar juveniles del bivalvo endémico Pinna nobilis mediante dispositivos de asentamiento" of the dg Recerca, desenvolupament Tecnològic i innovació, Conselleria d'economia Hisenda i innovació, Caib. it is important to mention that the experiments complied with the current laws of sPain and individuals of P. nobilis were collected under the license of autonomic institutions (government of the balearic islands).

Fig. 1 .
Fig. 1.geographic location of the study areas in the western Mediterranean.Continuous lines show isobaths every 10 metres up to 100 metres.a) sampling areas that showed non-invaded seagrass meadow.¢ indicates the location of the study sites espardell, esponja and Talamanca; b) sampling area that showed P. oceanica meadows invaded by L. lallemandii.p indicates the three sites selected for the study in sa dragonera natural Park.

Fig. 3 .
Fig. 3. -Mean values of potential food sources contribution (%) to the δ 13 C and δ 15 n diet of the fan mussel Pinna nobilis, the shrimp Pontonia pinnophylax and the pea crab Nepinnotheres pinnotheres in an invaded and non-invaded situation.i: invaded; ni: non-invaded; Po: Posidonia oceanica; PoM: Particulate organic matter; soM: sediment organic matter; ePol: epiphytes of Posidonia oceanica leaves; lo: Lophocladia lallemandii.The real contributions (RC) of L. lallemandii are delimited by the dashed lines.

Table 1 .
-Morphometric features of Pinna nobilis specimens and the presence of its associated guests, the shrimp Pontonia pinnophylax and the pea crab Nepinnotheres pinnotheres.Maximum shell width (W m ), maximum shell length (l m ), maximum length of the posterior adductor muscle scars (l ad ), and number of muscle scar ring (R) are represented. indicates the presence of guests of only one sex.p indicates the presence of the couple guests (male and female simultaneously).

Table 2 .
-number of samples (n), stable isotopes values (δ 13 C and δ 15 n) and trophic level (Tl) of the studied species at invaded and noninvaded locations and the food sources.Results are expressed as mean ± standard error (se).PoM, particulate organic matter; soM, sediment organic matter; ---, data insufficient.