The pearl oyster, Pinctada radiata, shows great variation in shell morphology throughout its distribution. This variation can be related to phenotypic plasticity, genetic variability or a combination of both. Using geometric morphometric and microsatellite DNA analyses, two morphologically distinct populations of the pearl oyster were studied in the northern Persian Gulf, i.e. from the Lavan and Hendourabi Islands. Ten landmarks were selected to define the shape of the left shell. In addition, concentration of Zn, Mg, Fe, Cu, Pb, Cd, Mn and Cr of the soft tissues were measured using atomic absorption spectrometry. Six microsatellite loci were used to assess the population genetic structure of the pearl oyster. There were morphometric differences between the populations suggesting the existence of two morphotypes. There was a significant difference between the two populations in concentrations of Fe, Mg, Zn, Cd, Mn and Cr indicating that the specimens from the Lavan Island experience a more stressful environment than those from the Hendourabi Island. Analysis of molecular variance (AMOVA) indicated that the proportion of the genetic variation attributed to differences among populations of the pearl oyster was highly significant for both FST and RST (FST = 0.066, RST = 0.265, P < 0.001). Our findings showed that stressful conditions resulting from heavy metals may have a direct influence on the separation of the populations in Lavan and Hendourabi despite the lack of a physical barrier.

The variability of shell morphology and relative growth of the invasive pearl oyster Pinctada radiata was studied within and among ten populations from coastal Tunisia using discriminant tests. Therefore, 12 morphological characters were examined and 34 metric and weight ratios were defined. In addition to the classic morphological characters, populations were compared by the thickness of the nacreous layer. Results of Duncan's multiple comparison test showed that the most discriminative ratios were the width of nacreous layer of right valve to the inflation of shell, the hinge line length to the maximum width of shell and the nacre thickness to the maximum width of shell. The analysis of variance revealed an important inter-population morphological variability. Both multidimensional scaling analysis and the squared Mahalanobis distances (D2) of metric ratios divided Tunisian P. radiata populations into four biogeographical groupings: the north coast (La Marsa); harbours (Hammamet, Monastir and Zarzis); the Gulf of Gabès (Sfax, Kerkennah Island, Maharès, Skhira and Djerba) and the intertidal area (Ajim). However, the Kerkennah Island population was discriminated by the squared Mahalanobis distances (D2) of weight ratios in an isolated group suggesting particular trophic conditions in this area. The allometric study revealed high linear correlation between shell morphological characters and differences in allometric growth among P. radiata populations. Unlike the morphological discrimination, allometric differentiation shows no clear geographical distinction. This study revealed that the pearl oyster P. radiata exhibited considerable phenotypic plasticity related to differences of environmental and/or ecological conditions along Tunisian coasts and highlighted the discriminative character of the nacreous layer thickness parameter.