Distribution of four isopods (Peracarida: Crustacea) in shallow waters of the Bizerte Lagoon (Tunisia, SW Mediterranean Sea)

Abstract The distribution of four isopods – Idotea balthica basteri (Pallas, 1772), Idotea chelipes mediterranea Charfi-Cheikhrouha, 1996, Sphaeroma serratum (Fabricius, 1787), and Paracerceis sculpta (Holmes, 1904) – at Menzel Jemil (Bizerte Lagoon, North Tunisia) was analysed and then related to the variation of temperature, depth and plant biomass. The abundance of isopods showed monthly fluctuations with maxima in August and minima in December 2010. The most dominant species was I. b. basteri followed by S. serratum, with a spatial gradient in the distribution pattern of isopod species in the study area. The high density of juveniles in algae and seagrass highlights the important role played by this habitat as a nursery for species. Both dominant species were tolerant to temperature variations, with some preference values (19.73–27.73°C in I. b. basteri; up to 28°C in S. serratum) and salinity (36.7–38.6 psu in I. b. basteri; 39–39.2 psu in S. serratum). Multiple regression analysis between environmental variables and species abundance showed a positive relationship with biomass of Gracilaria bursa-pastoris and Ulva lactuca for the two species of Idotea and Sphaeroma, respectively, while abundance of P. sculpta was positively related to temperature. Biotic and environmental linking (BIOENV) analysis confirmed the greatest influence of temperature and vegetation biomass on the temporal distribution of these isopod species. A spatial distribution pattern according to successive depths was revealed.


Introduction
Isopods are one of the most diverse orders of crustaceans, living in a wide variety of environments from marine and fresh waters to terrestrial habitats. They are most commonly found in shallow marine waters (El-Shahawy & Desouky 2010). Marine isopods are normal constituents of the fauna inhabiting macroalgae (Healy & O'Neill 1984;Arrontes & Anadón 1990). They frequently receive less attention than amphipods in ecological and taxonomical studies, and knowledge of them is scarcer (Guerra-García et al. 2009).
Marine isopods have a wide variety of feeding habits, i.e. herbivory, scavengery, carnivory, omnivory and even foraminiferivory (Robertson & Mann 1980;Healy & O'Neill 1984;Gudmundsson et al. 2000;Orav-Kotta & Kotta 2004). Changes in food supply may influence the distribution of species and diversity patterns in the benthic environment. Food selection and habitat complexity may play an important role in the distribution and abundance of marine organisms such as peracarids (Duffy & Hay 1991;Edgar & Robertson 1992). Distribution and habitat preferences of isopods are poorly known (Holdich & Lincoln 1974) despite their great importance, both qualitative and quantitative, as main secondary and tertiary producers in marine macrobenthic assemblages. Isopods of southern Mediterranean lagoons are still relatively unexplored compared with those of northern lagoons (Giordani Soika 1950;Sconfietti 1988;Fava et al. 1992;Mancinelli 2010;Vincenzi et al. 2013;Longo & Mancinelli 2014).
Tunisian lagoons are extremely interesting due to their geographic distribution from north to south and their location in different climatic zones related to rainfall patterns. They cover a very large area, 110,000 hectares, and offer a wide range of biotopes since they are of different sizes and depths.
To improve the knowledge of isopods in Tunisian wetlands, we identify the isopod species assemblage with algae and seagrass beds at Menzel Jemil in Bizerte Lagoon, study their spatial-temporal variation, analyse their distribution pattern and define the relationship between their abundance and the environmental factors.

Study area
The study site, located in the northeast of Bizerte Lagoon (37°14ʹ10.25"N, 9°54ʹ52.13"E), on a sandy mud bottom, was selected because of its abundance of isopods and its accessibility. According to Zaabar et al. (2015b), the vegetal cover at Menzel Jemil mainly consists of Ulva lactuca (Linnaeus, 1753); during spring substantial numbers of Gracilaria bursa pastoris (S. G. Gmelin) P. C. Silva 1952 and Gracilariopsis longissima (S. G. Gmelin) M. Steentoft, L. M. Irvine and W. F. Farnham 1995 are present. Cladophora sp. is found occasionally, and the seagrass Cymodocea nodosa (Ucria) Ascherson, 1870 is present throughout the year, exhibiting a growth increase in summer. The plant biomass at Menzel Jemil is related to temperature reaching a minimum in January and a maximum in June (Zaabar et al. 2014).

Sampling and laboratory procedures
Sampling was conducted monthly from October 2009 to September 2010, using a metal square of 0.25 m 2 with three replicates at each depth of 20, 40, 60 and 80 cm. The surface was scraped, and the vegetation and associated macrofauna species, including isopods, were removed after washing the algae and seagrass in a plastic tray and sieving with 1-mm mesh, and were then transported to the laboratory.
In the laboratory, isopod specimens found at each depth were preserved in 70% alcohol, identified and counted. Individuals of each species were separated and sorted according to their differentiation state. Then, the plant biomass was estimated by weighing all plant matter collected from each species of algae and seagrass after being dried at 70°C for 48 h.

Data analysis
Species richness (S), total abundance (N), density (ind.m −2 ), Shannon index (H'; Shannon & Weaver 1963), evenness index (J'; Pielou 1969), frequency index (F%; Soyer 1970) and dominance (Di%; Bellan-Santini 1969) were calculated monthly. Possible differences in these parameters betwen months were tested by one-way analysis of variance (ANOVA), after verifying normality (Kolmogorov-Smimov) and homogeneity of variances (Levene). To investigate the spatial-temporal distribution of isopods, the same procedure was carried out for testing statistical differences in the values of diversity indices and abundance of the species according to depths and seasons, respectively.
To determine preferences of temperature and salinity, values measured monthly were plotted vs. the abundance of specimens obtained from the algae and seagrass.
To better establish temporal distribution patterns, multidimensional scaling (MDS) based on the Bray-Curtis similarity was applied to abundance. The quantitative data from monthly samples was analysed, grouped according to seasons and compared statistically using analysis of similarity (ANOSIM). The SIMPER (SIMilarity PERcentage) procedure was used to identify those species that contributed most to the dissimilarity between these groups of samples.
Relationships between environmental variables (temperature, dissolved oxygen, turbidity, plant biomass) and isopod assemblages were examined using multiple stepwise regression analyes. In these analyses, only environmental variables (independent variables) were accepted in each regression model (p < 0.05).
The faunal and environmental ordinations of samples were also studied using biotic and environmental linking (BIOENV) analysis. Previously, environmental variables showing high correlation (after Pearson test) were redundant and not considered. The BIOENV analysis resulted in different combinations of environmental variables that are highly correlated with the ordination of faunal samples. These multivariate analyses were carried out using the PRIMER v. 5 software package.

Environmental variables
Regarding abiotic factors (Figure 1), the highest water temperature (30.86°C) and salinity (39.2 psu) were recorded in August; dissolved oxygen was high in May (7.3 mg.L −1 ), pH in February (8.65) and turbidity in November (15.44 NTU). Seasonal variations in the environmental parameters of Menzel Jemil (sampling site) outlined by ANOVA test were significantly different (p < 0.001). Pearson test showed that temperature and salinity were positively correlated (r = 0.763; p = 0.004); this was also the case for turbidity and pH (r = 0. 798; p = 0.004). Data showed a positive correlation between plant biomass and temperature (r = 0.868; p = 0.000).
With respect to Soyer's frequency (F) index, the four collected species were continuous (F ≥ 50%). Idotea balthica basteri and S. serratum were collected monthly (F = 100%, Table I) and many individuals of both species were juveniles. The non-indigenous species P. sculpta was also present in the algae and seagrass throughout the study year (F = 91.66%), except in June 2010. The presence of ovigerous females and juveniles in the vegetation revealed the establishment of the species in this area. Idotea chelipes mediterranea (F = 75%) was sparsely present, and totally absent from November 2009 to March 2010.
The Shannon-Wiener diversity index (H') ranged between 0.74 bits in May and 1.25 bits in July 2010, with significant seasonal variation (one-factor ANOVA; F = 19.31, p < 0.001). The evenness (J') index showed the same fluctuation, with a peak in December (0.76) and a low value in February (0.37). Significant seasonal differences were evident (onefactor ANOVA; F = 33.27; p < 0.001).
A spatial distribution pattern according to successive depths was revealed for the four species (Figure 3(b)), with evident differences: maximum abundance (79.69%) at 80 cm for I. b. basteri and 55.72% at 20 cm for S. serratum. No clear difference between I. c. mediterranea and P. sculpta was recorded, probably due to their reduced number. There is a spatial gradient in the distribution of isopod species in the study area. S. serratum was more related to the shoreline; its abundance decreased along the gradient depth, whereas the abundance of I. b. basteri was increasing (second author, pers. obs.). Based on the ANOVA test, S. serratum was more abundant at 20 cm depth (F = 1.09, p < 0.05), and I. b. basteri at 80 cm depth, than elsewhere (F = 0.62, p < 0.05).

Dynamics of two dominant species
We focus on the two dominant species which determined the density of the total isopod assemblage. The monthly dynamics of the I. b. basteri population is displayed in Figure 4 same pattern was recorded for juveniles, contributing more to the profile of population density in the algae and seagrass, which highlights the use of vegetation as a nursery ground. Sphaeroma serratum was also present in the algae and seagrass throughout the year (Figure 4(b)). Density values increased from July to October, with a maximum in August (160 ind.m −2 ). Similarly, the density of juveniles peaked in August, with a minimum in January (2.5 ind.m −2 ). In March, the bulk of the population was made up of juveniles, probably due to mortality of adults after their reproductive period.

Tolerance to temperature and salinity variation of two dominant species
The abundance of the dominant species, I. b. basteri and S. serratum, exposed to specific temperature and salinity conditions was computed. Figure 5(a) indicates a wide tolerance to temperature 9-30.86°C for both species. The ANOVA test revealed that the greatest abundance of I. b. basteri occurred at a temperature range of 19.73-27.73°C (F = 40.12, p < 0.05), and the number of specimens decreased when the temperature reached 28.46°C. It was found that temperatures up to 28°C were preferred by the population of S. serratum (F = 9.76, p < 0.05).

Affinity between samples
Multivariate analysis using fourth root transformed abundance data ( Figure 6) revealed three distinct groups of samples corresponding to different periods of the year in MDS representation. Group I was made up of months from December to February (winter), in which abundance of isopod species and temperature were low. Group II included samples from July and August (summer) where Sphaeromatidea was abundant and temperature high. Group III, composed of the remaining months (autumn and spring), was located at the opposite side of the MDS ordination plot to the winter and summer months. The latter group results in an intermediate temperature with a  little variation of I. b. basteri abundance. According to seasons, significant differences were found between the isopod assemblage (ANOSIM, R = 0.32; p = 0. 016). This was statistically confirmed when pairwise tests were applied. The highest dissimilarities of the SIMPER analysis were obtained between winter and summer (dissimilarity: 54.7%), autumn and winter (48.37%) and winter and spring months (44.48%). SIMPER analysis indicated that the species contributing the most to the difference between seasons were the two dominant species, I. b. basteri and S. serratum.

Linking isopod fauna and environmental variables
Using multiple stepwise regression analyses, the abundance values (N) of isopod species displayed a significant relationship with environmental variables (Table II).
The abundance of I. b. basteri and I. c. mediterranea was related positively to biomass of Gracilaria bursapastoris and negatively to biomass of Gracilariopsis longissima. In contrast, abundance of Sphaeroma serratum and Paracerceis sculpta was related positively to biomass of Ulva lactuca, which was also positively related to temperature.
Using BIOENV as a different multivariate approach, the composition and structure of isopods displayed the highest correlation when combining two variables such as temperature-biomass of Ulva lactuca and temperature-biomass of Gracilaria bursapastoris, and three variables such as temperature-Ulva lactuca biomass-Gracilaria bursa-pastoris biomass (Table III).

Discussion
This paper is the first investigation of the influence of environmental variables on the distribution of marine isopods associated with algae and seagrass at Menzel Jemil (Bizerte Lagoon). Environmental conditions during this study were relatively comparable with those measured during the last few years (Harzallah 2003;Béjaoui et al. 2008;Ben Garali et al. 2010). In this Mediterranean lagoonal complex, environmental variables strongly fluctuated with seasonal variations related to precipitation, temperature and currents (Afli et al. 2009b). Therefore, only some invertebrate species were tolerant to the extreme salinity and temperature changes between winter and summer, and able to survive in Bizerte Lagoon (Afli et al. 2009a).
In the present study, four isopod species inhabiting algae and seagrass were examined and strong differences in their abundance and distribution were reported. Isopod richness associated with algae and seagrass at Menzel Jemil station was lower than that recorded in Bizerte Lagoon (10 species) by Dridi and Prunus (1980) and Zaouali (1980); this could be due to the increase in salinity after the construction of dams, reducing freshwater input from Ichkeul Lake during the last few decades.
Idotea balthica basteri was the most dominant species, particularly in spring and early summer inhabiting Gracilariopsis longissima, Gracilaria bursa-pastoris, Ulva lactuca and decaying seagrass. Despite many ecological changes in the Bizerte Lagoon in the last few decades (Afli et al. 2009a), I. b. basteri inhabiting algae and seagrass was still the most dominant species, while Idotea chelipes mediterranea has significantly decreased compared to the years 1980-1990 (second author pers. obs.). The abundance of I. balthica could be explained by its competition with I. chelipes for food resources and space (Korheina 1981). Kotta et al. (2000) proposed that the decline of I. chelipes probably reflects the decreasing cover for benthic vegetation in the Gulf of Riga since other phytophilous species have declined. Ingólfsson and Agnarsson (2003), studying dispersal and movements of isopods and amphipods during high tide, and Franke et al. (2007) in their work about habitat selection and habitat segregation in I. balthica and I. emarginata, showed that I. balthica was the dominant species among surface-floating algae.
The low abundance of P. sculpta with respect to Sphaeroma serratum could be related to its recent colonisation of this area. Paracerceis sculpta is the most acclimated isopod in the Tunisian brackish waters (Ounifi Ben Amor et al. 2010). According to Zibrowius (1992), it was also the most widely ship-transported alien species in the world, and was recorded in the Mediterranean during the earlier part of the last century. The presence and establishment of this non-indigenous species, native to the northeast Pacific, in Bizerte Lagoon might be explained by biofouling on ships due to the proximity and connection of this lagoon to Bizerte Port. Furthermore, the position of Tunisia between the two Mediterranean basins will be a key area for understanding the influx of alien species in the whole Mediterranean Sea (Guidetti et al. 2010;Azzurro et al. 2014).
Temporal distribution of isopod species in Bizerte Lagoon might be due to a response to the same environmental factors (e.g. temperature, salinity, nutrients) causing the change and distribution in algae and seagrass cover on which animals live  (Arrontes & Anadón 1990). There are several potential environmental factors that might influence a species' distribution and abundance, including water temperature, salinity, light level, shelter from predators, and food resources (Dufour et al. 2007). Based on the recorded temporal distribution of Idotea and temperature and salinity fluctuations, we found that I. b. basteri, with a marine tendency, occurs at higher temperature and salinity than lagoonal species I. c. mediterranea, and contrary to Sphaeromatidea distribution. Similar patterns were observed by Leidenberger et al. (2012) for Idotea distribution in the Baltic Sea. I. chelipes prefers warmer temperatures (winter temperatures in the Bizerte Lagoon) coinciding with its preferred habitat in lagoons and estuaries (Zettler 2000(Zettler , 2001. On the other hand, I. chelipes was described as a brackishwater species, tolerating 3-30 psu (Łapucki & Normant 2008). Moreover, the decrease of I. c. mediterrranea over the last few decades in the Bizerte Lagoon could be explained by the increase of salinity due to the construction of dams in upstream Ichkeul Lake, causing a marinisation of the lagoon which becomes colonised by I. b. basteri. In addition, the Bizerte Lagoon ecosystem is currently destabilised by waste water coming from the  bordering cities (Dellali et al. 2001) and shoreline industries, whereas I. c. mediterranea was a more sensitive species to pollution. Consequently, the rearing of this species in the laboratory is more difficult (experiment by second author).
On the other hand, I. b. basteri, S. serratum and P. sculpta, as well as other typically marine species, entered and populated the Bizerte Lagoon due to favourable environmental conditions: relatively low temperature, high salinity, good oxygenation, low turbidity, high insolation, and abundance of food (Alves Martins et al. 2015).
Among isopods, I. balthica and S. serratum were the most common species in the littoral zone of the Mar Piccolo basin (Ionian Sea, southern Italy); numerically dominant, they constituted an important food for many predators such as crustaceans, fish, and birds in this ecosystem (De Nicola Giudici & Guarino 1989).
In the study area, I. b. basteri was the most dominant species followed by S. serratum, exhibiting a spatial distribution pattern. According to Casagranda et al. (2006), a similar distribution pattern was found in Lake Ichkeul. Indeed, I. chelipes was found in relatively deep stations with living vegetation and hard substrate (sand, gravel), whereas Sphaeroma hookeri mainly occurred at relatively shallow stations with plenty of detritus and silty substrate.
S. serratum usually lives in coastal marine or brackish waters, which are often subject to variations in salinity. It lives under stones close to the shore or among seaweed, in 5-80-cm-deep water (Prato et al. 2012).
Using multivariate analyses, Gracilaria bursa-pasteri biomass was the variable which better explained the temporal distribution of the Idotea genus, particularly I. b. basteri, the most investigated species in the Bizerte Lagoon, whereas the distribution of S. serratum and P. sculpta was related to Ulva lactuca biomass. Casagranda et al. (2006) reported that the functional relationship between shredders (e.g. I. chelipes, S. hookeri) and macrophytes may be nutritional. This meadow also provides the shredders with a nursery and some refuge from wave action, predation, salinity and temperature variations. According to Raberg and Kautsky (2008) I. balthica was the second most common crustacean meso-grazer besides Gammarus spp. on the Swedish coast, and its mean abundance appeared to be highest in the Cladophora glomerata zone.
The most dominant species in this study, I. balthica and S. serratum, were tolerant to some variations in temperature and salinity. Similar results have been reported for many other isopod species in a tropical coastal lagoon (Garcia-Guerrero & Hendrickx 2010). From their point of view, water and vegetation parameters may influence the seasonal distribution of isopods.
In summary, algae and seagrass in Bizerte Lagoon represent a habitat where isopods display spatial and temporal distributions. Moreover, this Mediterranean ecosystem, located between Ichkeul Lake and the Mediterranean Sea, under the influence on the one hand of fresh water which decreased after the construction of dams in upstream Ichkeul Lake, and marinisation of the lagoon due to climate change on the other hand, may favour the entrance of non-indigenous species. Finally, in order to increase knowledge about the dispersion of non-indigenous species in the Mediterranean lagoons, studies on the macrofauna of different coastal lagoons of Tunisia should be conducted regularly.  Table III. Results of biotic and environmental linking (BIOENV) analysis displaying the three highest Spearman rank correlations using square root quantitative data of isopods collected from October 2009 to September 2010. T, temperature (°C); U.lac, Ulva lactuca biomass (g.m −2 ); G.b.p, Gracilaria bursa-pastoris biomass (g.m −2 ).