First insight of genetic diversity, phylogeographic relationships, and population structure of marine sponge Chondrosia reniformis from the eastern and western Mediterranean coasts of Tunisia

Abstract Despite the strategic localization of Tunisia in the Mediterranean Sea, no phylogeographic study on sponges has been investigated along its shores. The demosponge Chondrosia reniformis, descript only morphologically along Tunisian coasts, was chosen to estimate the influence of natural oceanographic and biogeographic barriers on its genetic differentiation and its Phylogeography. The cytochrome oxidase subunit I (COI) gene was amplified and analyzed for 70 Mediterranean Chondrosia reniformis, collected from eight localities in Tunisia. Polymorphism results revealed high values of haplotype diversity (H d) and very low nucleotide diversity (π). Thus, these results suggest that our sponge populations of C. reniformis may have undergone a bottleneck followed by rapid demographic expansion. This suggestion is strongly confirmed by the results of neutrality tests and “mismatch distribution.” The important number of haplotypes between localities and the high genetic differentiation (F st ranged from 0.590 to 0.788) of the current C. reniformis populations could be maintained by the limited gene flow Nm (0.10–0.18). Both haplotype Network and the biogeographic analysis showed a structured distribution according to the geographic origin. C. reniformis populations are subdivided into two major clades: Western and Eastern Mediterranean. This pattern seems to be associated with the well‐known discontinuous biogeographic area: the Siculo‐Tunisian Strait, which separates two water bodies circulating with different hydrological, physical, and chemical characteristics. The short dispersal of pelagic larvae of C. reniformis and the marine bio‐geographic barrier created high differentiation among populations. Additionally, it is noteworthy to mention that the “Mahres/Kerkennah” group diverged from Eastern groups in a single sub‐clade. This result was expected, the region Mahres/Kerkennah, presented a particular marine environment.


| INTRODUC TI ON
Phylum Porifera, commonly known as sponges, are the evolutionary oldest multicellular animal. These invertebrates are found in all oceans and at all depths (Hooper & Van Soest, 2002;Van Soest et al., 2012). According to the World Porifera Database, the number of described taxa are more than 9000 (Van Soest et al., 2021).
The latest classification divides sponges into four classes; Calcarea, Hexactinellida, Homoscleromorpha, and Demospongiae, to which more than 90% of sponge species belong. Unlike the other animals, sponges are the simplest group that lacks true tissues and organs.
They are only formed by specialized cell types (e.g., choanocytes and pinacocytes), which are embedded in a complex matrix called mesohyl (Junqua et al., 1974;Simpson, 2012). Despite their simple morphology, sponges' genome is complex (Harcet et al., 2010). Due to their higher filtering capacity, sponges play interesting roles in biogeochemical cycling and in the benthic-pelagic coupling of nutrients within the ecosystem (De Goeij et al., 2013;Lesser, 2006;McMurray et al., 2017). Contrarily to most benthic organisms, these invertebrates have the capacity to pump large volumes of seawater through the water column. Crossing through the body of the sponge, seawater is chemically transformed due to feeding, excretion, and the activities of microbial symbionts, with significant effects on the carbon and nutrient cycling (Pawlik & McMurray, 2020). To defend against predators, and pathogens, sponges have developed numerous secondary metabolites, which present a high biotechnological potential in different domains (Genta-Jouve & Thomas, 2012).
The sponge can reproduce sexually (gametes are produced from two types of somatic stem cells) or asexually. Sexual reproduction in sponges is varied, it can be gonochoric, sequential, or simultaneous hermaphrodites, it also can be through viviparity or oviparity (Maldonado, 2006). Their larvae are characterized by low dispersal potential (Vacelet, 1999). This disperse over short distances potential has important consequences for the connectivity and genetic structuring of sponge populations (Avise, 1994;Scheltema, 1971).
Thus, larval dispersal potential is a key factor that can be used to understand the spatial patterns of genetic diversity, which is the main goal of phylogeographic studies (Cowen & Sponaugle, 2009;Palumbi, 2003). Phylogeography, which combines genetic and geographic data, allows comprehension of the distribution of genetic differentiation in terrestrial and aquatic ecosystems. Thus, this approach confers to understanding the spatial patterns of genetic diversity and both the historical and contemporary factors acting on taxa (Avise, 2000;Rissler, 2016). Moreover, phylogeographic studies are substantial for the development of effective conservation strategies in the increasingly threatened marine realm (Moritz, 2002;Moritz & Faith, 1998 Oceans and Mediterranean Sea (Di Camillo et al., 2011;Idan et al., 2020;Lazoski et al., 2001). Chondrosia reniformis lives on shady rocky coasts at a depth of up to 50 m and it can be found in shallow, mesophotic, and oligotrophic habitats (Di Camillo et al., 2011;Idan et al., 2020).
This demosponge is a gonochoric broadcaster sponge that also can reproduce asexually via drop-like propagules (Di Camillo et al., 2011;Riesgo & Maldonado, 2008). Both the dispersal capability of the lecithotrophic larvae and the gamete's dispersal are probably low. Its reproductive cycle is believed to be influenced by temperature (Idan et al., 2020). Among areas, oogenesis appears to be varied from seasonal to continuous, it is obtained before the temperature peak around May to August (Di Camillo et al., 2011;Riesgo & Maldonado, 2008). Spermatogenesis in C. reniformis seems to be rapid and probably synchronized with the last developmental stage of the oocytes (Di Camillo et al., 2011).
However, few genetic studies have been performed on this species.
Indeed, Lazoski et al. (2001) have investigated the levels of genetic variation within and between geographically distant populations of this species from the Atlantic (North and South America) and Western Mediterranean sea coasts.
In the last two decades, DNA sequences have been extensively used to understand the evolutionary history and spatio-temporal genetic divergence of species, and it is mitochondrial DNA that is commonly used. Indeed, since its maternal inheritance without recombination, high mutational rate, shorter coalescence times, and

J E L C L A S S I F I C A T I O N
Biodiversity ecology high copy numbers in the organism (Avise, 2000(Avise, , 2009Palumbi et al., 2001), this genome is commonly used as a genetic marker to identify the taxa as well as to investigate phylogeographic relationships in most marine organisms (Avise, 2000). However, no nuclear or mitochondrial DNA molecular studies have been undertaken on C. reniformis to analyze its population structuring and its phylogeography. The only studies carried out have focused on the cytochrome oxidase subunit I DNA marker (COI) in order to position the genus Chondrosia in the phylogenetic tree of demosponges (Riesgo et al., 2014;Rot et al., 2006;Rua et al., 2011;Vacelet et al., 2000;Villamor et al., 2014;Xavier et al., 2010).
The aim of this study, using COI mitochondrial DNA marker is to estimate levels of diversity and differentiation of Tunisian coastal populations of the two east and west Mediterranean basins, to analyze the effects of natural oceanographic and biogeographic barriers between these two basins and finally to establish for the first time the phylogeography of Chondrosia reniformis along its Tunisian coastal distribution ( Figure 1).

| Sample collection and genomic DNA extraction
A total of 70 specimens of Chondrosia reniformis (Nardo, 1847) were collected, between January and September 2020, from eight sampling locations along the Tunisian coasts ( Figure 2, Table 1). These DNA of sponge specimens was extracted using EZ-10 Spin Column Kits (Bio BASIC INC, Canada) as described by the manufacturer.
DNA quantity and quality were performed using a spectrophotometer (Gold S54T, Shanghai) and agarose gel electrophoresis (Sambrook et al., 1989).

| Statistical analyses
Since the number of samples from each locality is unequal, we divided the sample into four groups based on geographic proximity (Beja/Tabarka, Monastir/Sousse, Mahdia/Chebba and Mahres/ Kerkennah).
The level of DNA polymorphism, the haplotype diversity (H d ; Nei, 1987) as well as the nucleotide diversity (π; Nei, 1987;Tajima, 1983), were measured for each group and for the total datasets using DnaSP version 5.10 (Librado & Rozas, 2009). The percentages of GC and AT, the number of variable and parsimony-informative nucleotides sites were calculated with MEGA version 7.0.18 (Kumar et al., 2016).
The demographic history of the Mediterranean population of C. reniformis was investigated. The mismatch distribution test was performed with DnaSP v5.10.01 (Librado & Rozas, 2009) for all datasets and each group. To study the hypothesis of population expansion, additional tests were performed using the total number of mutations: Tajima's D-test (Tajima, 1989), Fu's Fs test (Fu & Li, 1993), raggedness index (rg) and Ramos-Onsins, and Rozas's R2 test (Ramos-Onsins & Rozas, 2002). These analyses were executed using coalescent simulations implemented in DnaSP software, with 1000 simulated re-sampling replicates.
F I G U R E 1 Photo of Chondrosia reniformis specimen collected by Wissem Dallai diver from the Beja locality in September 2020 at 10 depth

| Phylogeographic analysis and Genetic differentiation
To infer the relationships of C. reniformis haplotypes, we used the NETWORK software (Bandelt et al., 1999). Phylogenetic reconstructions were performed using (1)  Bayesian analysis was performed using the HKY + I + G model, as determined by the JModel Test (Posada, 2008) using the model correction based on AIC (Hasegawa et al., 1985). One sequence from Chondrilla nucula (GenBank accession numbers: EF519598.1) was used as outgroup.
The analysis of molecular variance (AMOVA) (Excoffier et al., 1992) was conducted by Arlequin 3.5 software (Excoffier & Lischer, 2010) to assess the level of genetic differentiation of Tunisian Chondrosia reniformis populations. Two supplementary AMOVA tests were carried out: for the first analysis, we tested the genetic variation between the four groups according to geographic proximity: Beja/Tabarka, Monastir/Sousse, Mahdia/Chebba and Mahres/ Kerkennah. The second analysis was performed to evaluate a comparison between the group Mahres/Kerkennah and the other localities. In addition, the genetic differentiation for both eastern and western Mediterranean localities in Tunisia was tested too. All AMOVA analyses were calculated with 10,000 permutations under null distributions.
The extent of genetic differentiation between populations was estimated using the fixation index F ST and the gene flow (N m ) (Hudson et al., 1992). Values were calculated with 1000 data permutations using the software DnaSP v 5.10.01 (Librado & Rozas, 2009).  (Table 2).

| Genetic diversity and molecular evolution
Selective neutrality was estimated by Tajima (1989) and Fu and Li (1993) tests. These statistic tests were negative and insignificant for the three groups (Beja/Tabarka, Monastir/Sousse, and Mahdia/ Chebba) and all datasets (Table 3) We also calculated Ramos-Onsins and Rozas's R2 and the raggedness index under the demographic expansion model for each population. We found that all populations had a nonsignificant raggedness index, which indicates that data has relatively good fit to a model of a population in demographic expansion (Harpending, 1994).

| Phylogeography and genetic differentiation
The haplotype network, as well as the biogeographic trees, were built to discover genealogical relationships between Chondrosia reniformis haplotypes in Tunisia (Figure 4) showed that more than 44% of the variation was between these two groups (ΦCT = 0.441, p < .05 (Table 4)); these haplogroups were suggested by the network and phylogeographic trees. "Mahres" and "Kerkennah" localities are parts from the Gulf of Gabes, this region of Mediterranean is well known to have extreme environmental conditions (Bejaoui et al., 2004;Ghannem et al., 2011), for that, we test the opportunity to have a specific genetic differentiation in this area.
AMOVA results revealed that more than 73% of variation occurred between populations within this group.
The entire pairwise comparisons of groups based on F ST and N m were significant (
Polymorphism results showed also that C. reniformis harbors high haplotype diversity (H d ) throughout Tunisia coasts (H d = 0.939).
Diversity indices H d and π were calculated to estimate the genetic architecture of populations and retrace possible historical events that may have acted on observed genetic diversity. It is generally accepted that small values of π suggest recently diverged populations due to founder effects or/and bottlenecks. Large values of π indicate deep genetic divergences between populations accumulated in isolation over long periods of time. According to Grant and Bowen (1998), the values of π vary from 0 to >0. value for π; H d > 0.5 and 0.5-0.8% < π ≤ 1%). Thus, we can consider that our sponge populations of C. reniformis may have undergone a bottleneck followed by rapid demographic expansion as mentioned by these authors for this category. This suggestion is strongly confirmed by the results of neutrality tests and "mismatch distribution." However, the lower nucleotide diversity recorded could be derived habitats (Deli et al., 2015;Kelly & Palumbi, 2010). However, decreased time that larvae spent in plankton is usually correlated with high differentiation among populations and vice versa (Avise, 1994;Scheltema, 1971). Several studies have reported that the dispersal ability of C. reniformis larvae is very low (Lazoski et al., 2001;Maldonado et al., 2021). The pelagic larval dispersal of C. reniformis is very short and lasts only a few days or even a few hours (Maldonado & Young, 1996;Uriz et al., 1998). Though, using allozyme marker, populations (Lazoski et al., 2001). In these conditions, this unexpected find can be related to anthropogenic transport that had been reported for many marine invertebrate species (Holland, 2000).
Even though the genetic diversity of sequences was low, genetic differentiation was strong. Both haplotype Network and biogeographic trees analysis showed a structured distribution according to the geographic origin. The AMOVA analysis also con-  (Mejri et al., 2009), the green crab Carcinus aestuarii (Deli et al., 2015), the banded Murex Hexaplex trunculus (Marzouk et al., 2016), and the black sea urchin Arbacia lixula (Deli et al., 2017). This pattern seems to be associated with the well-known discontinuous biogeographic zone: the Siculo-Tunisian Strait, which separates two The AMOVA test of Tunisian C. reniformis sponge revealed that 46.47% of the genetic variation was detected between the four studied groups. AMOVA results for the western Mediterranean and the eastern Mediterranean localities showed that more than 44% of the variation was between these two groups. these haplogroups were suggested by the network and phylogeographic trees. "Mahres" and "Kerkennah" localities are parts from the Gulf of Gabes, this region of Mediterranean is well known to have extreme environmental conditions, for that, we test the opportunity to have a specific genetic differentiation in this area. AMOVA results revealed that more than 73% of variation occurred between populations within this group.  (Marzouk et al., 2016). The hydrodynamics was higher in the northern than in the southern Mediterranean coasts (Oueslati, 1993). According to Pinardi and Masetti (2000), the eastern Mediterranean Basin is characterized by very weak circulation.
The Siculo-Tunisian Strait, from Cap-Bon (Tunisia) to Mazara del Vallo (Sicily Island, southern Italy), has been inferred to be an oceanographic and biogeographic barrier between the two major Mediterranean sub-basins (the western and eastern) (Bianchi & Morri, 2000).
On the other hand, Tunisian coasts are distinguished by a difference in temperature and salinity, the Eastern coasts being warmer and more saline than the Western ones (Serena, 2005).
In addition, the Tunisian coastline has different habitat textures varying from the muddy and sandy texture in the East to the rocky texture in the West. Due to the different geographical range of habitats, genetic differentiation between the western and the eastern Mediterranean populations of C. reniformis has been observed along the Tunisian coastline. This genetic differentiation was conformed to the apparent morphology of C. reniformis collected along this coastline. Thus, as shown in the photos (Figure 5), the Western C. reniformis specimens exhibit a light color and flattened shape, which contrasts with the Eastern specimens of dark color and lobed shape.
Additionally, it is noteworthy to mention that the "Mahres/ Kerkennah" group diverged from Eastern ones in a single sub-clade.
This result was expected; indeed the region "Mahres / Kerkennah" belongs to the Gulf of Gabes, which represents a particular marine environment seriously influenced by phosphate industries. In fact, since the industrialization in 1970, the phosphogypsum discharge has been the main cause of the disequilibrium of this ecosystem of this important gulf. Currently, three regions, Sfax, Skhira, and Gabes generate phosphoric acid along the coasts and produce a large amount of phosphogypsum as a waste product (Bejaoui et al., 2004;Ghannem et al., 2011). The degradation of the ecosystem in these places has resulted in a degradation of water quality, a decrease in fish resources and a loss of marine biodiversity (Hamza-Chaffai et al., 2003;Rabaoui et al., 2014;Salem et al., 2015). Indeed pollution and climate change have created large dead zones in oceans; however, sponges are able to self-organize and adapt more than any other species. They develop in the environments to which they have become accustomed over the millions of years of their evolution (Leys & Kahn, 2018;Müller & Müller, 2003).
That versatility may be the key to their biodiversity even in polluted environments.
In summary, Tunisian Chondrosia reniformis evolution was af- The short dispersal of pelagic larvae of C. reniformis and marine biogeographic barrier created high differentiation among populations.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest regarding the publication of this paper. Khaled Said https://orcid.org/0000-0003-4137-7667