Tunisia has many Artemia populations, but little information is available concerning their taxonomy, biometry and morphology. This work is an updated systematic inventory of Artemia populations in Tunisia, based on the comparison of different morphometric parameters measured in cultured adult individuals. Sixteen Tunisian Artemia populations were examined. The study included reference samples of two Artemia franciscana (San Francisco Bay, California, USA; and Great Salt Lake, Utah, USA) populations. The variability among the diverse populations was studied by statistical treatment of data through the analysis of variance. This analysis did not unveil any particular similarity among the Artemia populations studied, apart from the width of the head of male specimens (F=1.088, P=0.360). Biometrical analysis of these data was performed via multivariate discriminant analysis, and using the origin of each population as a separation factor. Results revealed that all the Tunisian Artemia populations studied can be classified as Artemia salina, with the exception of the Artemia population from Sabkhet Halk El Menzel, which belongs to Artemia franciscana.
The brine shrimp Artemia Leach, 1819 (Branchiopoda: Anostraca) comprises a number of bisexual species and parthenogenetic strains that are morphologically similar. Brine shrimps are found in inland and coastal saline and hypersaline waters throughout the world, with the exception of Antarctica (Triantaphy llidis et al. 1998; Van Stappen 2002). The present Artemia distribution is not only due to natural means (high dispersal potential of diapausing cysts) but also to human activity. Artemia species and strains have a remarkable genetic variability that can be expressed in various phenotypic characteristics such as morphometry, life history, breeding mechanisms and growth rate, according to their genetic differentiation and population structure (Gajardo et al. 2004; Muñoz et al. 2008).
The taxonomic study of Artemia populations in the Western Old World, including Italy, the South of France and the Iberian Peninsula (Spain and Portugal), together with North Africa, is interesting owing to the presence of the Mediterranean bisexual Artemia salina (L., 1758) and at least two different parthenogenetic forms, diploid and tetraploid (Gilchrist 1960; Vieira & Amat 1985; Vanhaecke et al. 1987; Amat et al. 1995). This Mediterranean region also holds the recent invader species, Artemia franciscana Kellogg, 1906 (Narciso 1989; Amat et al. 1995; Amat et al. 2005; Mura et al. 2004; Amat et al. 2007; Scalone & Rabet 2013). In Tunisia, Artemia has been reported in 23 different locations, characterized as temporal or ephemeral catchments, and is distributed from semi-arid to Saharan hydrogeographical zones (Ben Naceur et al. 2009, 2010, 2012a). However, there is a paucity of data regarding their taxonomy, biogeography and genetic characteristics. Kaiser et al. (2006) published the most recent checklist of the zoogeography of Artemia and included eight Tunisian Artemia populations, three of which were cited as being of unknown specific identity.
The most relevant methods used for Artemia species characterization are the comparison of morphometric traits for specimens cultured under standard conditions (Hontoria & Amat 1992a, b; Gajardo et al. 1998; Amat et al. 2005; Mura et al. 2005), their genetic characterization (Zhang & King 1992; Abatzopoulos et al. 2002; Camargo et al. 2002; Kappas et al. 2004; Van Stappen et al. 2007; Muñoz et al. 2008; Tizol-Correa et al. 2009), electron microscopic examination of some morphological traits (Criel & Macrae 2002; Mura & Nagorskaya 2005; Mura et al. 2005; Ben Naceur et al. 2011a), and cross-fertility tests (Gajardo et al. 2001; Abatzopoulos et al. 2002).
Although modern methods rely on genetic characterization through molecular markers for the differentiation of anostracan populations, it is well known that among Artemia, morphometrical traits have been the basis for describing species and strains (e.g. Hontoria & Amat 1992a, b; Triantaphyllidis et al. 1997; Camargo et al. 2003). Work on morphometrics of Artemia (Schmankewitsch 1873 in Litvinenko et al. 2007; Gaevskaya 1916; Gilchrist 1960; Ben Naceur et al. 2011 b, 2012b) has shown that Artemia individuals exhibit morphological changes in accordance with environmental conditions. Moreover, Hontoria and Amat (1992a) reported that individuals from different Artemia populations, although similar in body shape, show morphological traits that enable morphometric differentiation when they are cultured under standard laboratory conditions.
In the present work, variation in different morphometric variables measured on adult Artemia specimens from Tunisian populations, and cultured under laboratory conditions, has been studied. Multivariate discriminant analysis has been used in order to support current knowledge on the taxonomic status of Artemia in these populations.
MATERIAL AND METHODS
Specimens from 16 Tunisian Artemia populations (Table 1) were cultured in the laboratory, characterized morphometrically, and compared with cultured specimens from two Artemia franciscana populations (San Francisco Bay, California, USA; and Great Salt Lake, Utah, USA).
Experimental Artemia populations were reared under standardized culture conditions in order to minimize the strong environmental influences on body morphometry. Nauplii hatched from original cysts were reared up in 2-L plastic containers filled with 90 ± 10 g/L filtered and autoclaved brine (seawater plus crude sea salt) (Amat et al. 2005). Temperature was maintained at 24 °C during a 16 h light/8 h dark cycle. Shrimps were fed the unicellular alga Chlorella sp. at a density of 100-200× 103 cells/mL. The medium was renewed twice a week with a new microalgal culture.
As soon as specimens reached reproductive maturity (i.e. well-developed antennae on males and a well-developed brood pouch on females), a random sample of 20 individuals was removed from the culture (for each sex and population), anaesthetized with a few droplets of water saturated with chloroform and measured under a microscope using a calibrated micrometer eye piece. The following were measured for each male and female: total length (tl), abdominal length (al), width of third abdominal segment (wts), length of cercopods (lc), number of setae on left cercopod (nlc), number of setae on right cercopod (nrc), width of head (wh), diameter of compound eyes (dy), maximal distance between compound eyes (dby), length of first antenna (la), width of brood pouch (wb) (for females), width of second abdominal segment (wss) and width of frontal knob (fk) (for males); abdomen length : total length ratio (ra, %) was also calculated for both sexes (Amat et al. 2005). The morphometric variability among the diverse populations was established by statistical treatment of data through one-way ANOVA (Tukey, P<0.05). The homogeneity of variances was tested using the Leven's test (Norušis 1993). Biometrical analysis of these data was performed via multivariate discriminant analysis using the statistical package SPSS for Windows, version 10.0 (Norušis 2000).
Sources of Tunisian Artemia populations used for morphometrical characterization and literature references for autochthonous populations with verified species status.
The statistically significant inter-population differences in morphometrics have been observed for the male and female specimens (Tables 2, 3). Results of variance demonstrate different degrees of variation and do not show any particular similarity among the Artemia populations studied, apart from the width of the head of male specimens (F=1.088,P = 0.360).
When the origin of populations was used as a separation factor for the multivariate discriminant analysis applied to morphometric data, this procedure showed 11 discriminant functions for males and 10 discriminant functions for females (Fig. 1, Table 4). The first five of these functions gave cumulative variance percentages of 85.7% and 89.8% for males and females, respectively. Morphometric characteristics showed a clear differentiation between Tunisian populations (except that from Sabkhet Halk El Menzel) and the two reference populations.
The centroids (Fig. 1) demonstrate that populations belonging to the same species tend to be grouped. In the case of males (Table 5), the analysis shows that morphometric characters correlated with the first discriminant function are total length, length of cercopod, width of second abdominal segment and abdominal length. Those related to the second function are total length, abdominal length and total length and abdomen length/total body length ratio. For females, total length and abdominal length are morphometric traits correlated with the first discriminant function, while abdominal length, total length and abdomen length / total body length ratio, are related with the second function.
Tunisian Artemia populations, except for Sabkhet Halk El Menzel, were evidently grouped and separated from the two A. franciscana reference data sets (Fig. 1). Tunisian populations previously identified as A. salina (Table 1) were well integrated among unidentified Tunisian populations studied here, supporting their taxonomic assignment to the bisexual A. salina species. The Artemia population from Sabkhet Halk El Menzel was grouped with the two American populations and can be considered as belonging to the invasive A. franciscana species.
Morphometrical differentiation has been one of the most useful methods for Artemia taxonomy and biosystematics (Asem & Rastegar-Pouyani 2008). The results obtained here with multivariate discriminant analysis for 16 Tunisian Artemia populations and two American A. franciscana populations, using the origin of each population as a separation factor, provided two different phenotypic groups (the first group formed by populations identified as A. salina and unknown Artemia and the second group composed of the two American A. franciscana populations, together with the Tunisian population found in Sabkhet Halk El Menzel). Ben Naceur et al. (2011a, b) studied morphological characteristics (ovisac morphology in females and the frontal knob morphology and ornamentation (number of spines and mechanoreceptors), as well as the basal part of the penis, in males) of some Tunisian Artemia populations. Their results show that the male Artemia specimens from the Sabkhet Halk El Menzel population have a sub-spherical frontal knob shape, and spine-like projections present in the non-retractile basal part of the penis, leading to the conclusion that this population belongs to a species other than A. salina. In fact, based on the presence or absence of a spine on the basal part of penis, A. salina can be distinguished from all the other bisexual Artemia populations, since all non-Mediterranean bisexual Artemia (from both the New and the Old World) have spines basally on the penis (Triantaphyllidis et al. 1997; Mura & Nagorskaya 2005).
Several publications recently reported the presence of the invasive A. franciscana in different countries around Tunisia. This event was initially described in Portugal (Hontoria et al. 1987, in Amat et al. 2007), France (Thiery & Robert 1992; Scalone & Rabet 2013), Italy (Mura et al. 2004) and in Spain and Morocco (Amat et al. 2005; Green et al. 2005; Amat et al. 2007). The occurrence of A. franciscana in Portuguese hypersaline water bodies might date from the 1980s, and although the time and the place of the original introduction remain unknown, the introduction could have been intentional (Amat et al. 2007), especially as the beneficial effect of the branchiopod Artemia in salt production and their use in aquaculture as a food source were well known (Dhont & Sorgeloos 2002). This species was probably then dispersed via waterfowl northwards along the East Atlantic fly way (Green et al. 2005) and eastwards, reaching the Spanish, Moroccan and Italian sites where it was found (Amat et al. 2007). This hypothesis about the natural dispersion of A. franciscana via water birds should not be dismissed in the light of intentional inoculations carried out in industrial salt pans exploited around the Mediterranean basin (Amat et al. 2007). Concerning the Tunisian Artemia populations, the results obtained herein suggest that Artemia harvested from Sabkhet Halk El Menzel belongs to the invasive A. franciscana. Ben Naceur et al. (2009) prepared a previous check list on the distribution of Artemia in Tunisia and signalled the presence of this branchiopod in 21 sites, with the exception of the Sabkhet Halk El Menzel population. Furthermore, Sabkhet Halk El Menzel has the main marine aquaculture (fish farm) in Tunisia, in its southeast part. This private farm, concerned with the intensive production of sea bass and sea bream, includes a hatchery where commercial Artemia cysts are used to obtain brine shrimp nauplii for feeding fish larval stages.
Mean values ‘mm’ (standard deviation in parentheses) of morphometric characters measured in males of different Tunisian Artemia populations. Total length (tl), abdominal length (al), width of second abdominal segment (wss), width of third abdominal segment (wts), length of cercopods (lc), number of setae inserted on left cercopod (nlc), number of setae inserted on right cercopod (nrc), width of head (wh), maximal distance between compound eyes (dby), diameter for compound eyes(dy), length of first antenna (la), width of frontal knob (fk) and abdomen length : total length ratio (ra, %). Abbreviations for populations are explained in Table 1. Same letters show non-significant differences between mean in each row of main column (P=0.05).
Mean values ‘mm’ (standard deviation in parentheses) of morphometric characters measured in females of different Tunisian Artemia populations. Total length (tl), abdominal length (al), width of brood pouch (wb), width of third abdominal segment (wts), length of cercopods (lc), number of setae inserted on left cercopod (nlc), number of setae inserted on right cercopod (nrc), width of head (wh), maximal distance between compound eyes (dby), diameter for compound eyes (dy), length offirst antenna (la), and abdomen length : total length ratio (ra, %). Abbreviations for populations are explained in Table 1. Same letters show non-significant differences between mean in each row of main column (P=0.05).
Results for discriminant analysis based on morphometrical traits for Artemia populations from Tunisia, with the two commercial A. franciscana populations used as reference.
In conclusion, all Tunisian Artemia populations, except that from Sabkhet Halk El Menzel, were identified as A. salina. The new Artemia population from Sabkhet Halk El Menzel can be considered to be the invasive A. franciscana. However, other methods must be used (such as monitoring of reproduction and genetic characterization) to confirm the taxonomic status of the Artemia population from Sabkhet Halk el Menzel. The time and origin of the introduction there of this exotic species are unknown but aquaculture activities should be suspected as having been responsible.