Population genetics and growth of the Rabbitfishes Siganus luridus and Siganus rivulatus: investigating factors structuring successful invasion of the Mediterranean Sea

Understanding the factors allowing for or, constraining the successful invasion of a new environment will be vital in predicting and monitoring the future success of invasive species. This is especially true in the context of the Mediterranean Sea where hundreds of species have invaded through the Suez Canal from the Red Sea over the last 140 years. Siganus luridus and Siganus rivulatus, two Lessepsian rabbitfish species, are among the most successful invaders established in the Mediterranean Sea to date, with abundant populations in the eastern, central and western Mediterranean. This project aimed to quantify the effectiveness of two factors for predicting invasion success; population genetic diversity and growth by examining populations of both species across 12 sites, from the eastern to the western Mediterranean basin. Samples from the Red Sea were also included in the phylogenetic analyses to allow for possible identification of founder effects within the introduced range. However, the small Red Sea sample size precluded their use from other analyses and made it impossible to investigate whether there had been an initial genetic bottleneck. To determine population diversity, sequences for two mitochondrial markers (cytochrome oxidase C subunit I (COI) and the control region (CR)) were obtained from 108 individuals of S. luridus and 109 individuals of S. rivulatus. Sequences were used to generate and examine maximum likelihood (ML) trees, haplotype network analyses and comparison of haplotype (h) and genetic diversity (nucleotide diversity (pi) and pairwise distance) between species and populations found along a latitudinal gradient, with increasing distance from Suez Canal. The ML tree and haplotype networks did not demonstrate evidence of geographic segregation of genetic types, but instead indicated high genetic similarity between populations throughout the Mediterranean Sea. The numbers of shared and unique haplotypes demonstrated that movement of invasion is from east to west; however, with increasing distance from the Suez Canal no notable decreases in haplotype or genetic diversity were found. Haplotype diversity was high in both species but lower than previous estimates from the Red Sea, potentially indicating a reduction in haplotype diversity with invasion. Such estimates of reduction were lower than have been found in other invasive populations (e.g., Fistularia commersonii). Nucleotide diversity levels (pi), pairwise distances and F-statistic values revealed close relationships between populations of both species. However, significant levels of moderate to high genetic differentiation between populations of S. luridus were apparent. The results of the neutrality tests (Tajima’s D, Fu’s F and the mismatch distribution) had variable results in both species, though there was evidence to suggest demographic change in Trachila, the Peloponnese for S. rivulatus and sites from Turkey, the Peloponnese and the Pelagie Islands for S. luridus. The haplotype networks indicate this demographic change is population expansion. It can be argued that S. luridus is potentially a more successful invader due to its increased distribution throughout the Mediterranean Sea, higher haplotype diversity and higher genetic structure among populations than S. rivulatus. The higher genetic structure indicates greater invasion success as the greater the diversity within the species gene pool, the more capable the species will be at surviving environmental changes and coping with stressors. These factors indicate that genetic diversity may be a successful indicator of invasion success. To determine whether life history characteristics differed between S. luridus and S. rivulatus across the Mediterranean Sea, using otolith microchemistry the age and growth rate (using width of annuli as a proxy for annual otolith growth) of 32 individuals of S. luridus and 27 of S. rivulatus were determined. When inferring growth rates from otolith growth bands S. rivulatus showed significantly greater growth than S. luridus throughout all sites. Although S. rivulatus populations within the eastern Mediterranean Sea demonstrated greater-age-at-length than populations in the central Mediterranean Sea, this pattern was not apparent for populations of S. luridus. For this species, sites in the western Mediterranean held the largest individuals for their ages. A linear model was conducted to investigate if annulus, site, species or any of the relationships between these factors had a significant impact on the amount of growth observed (interpreted from otolith annulus widths). Site did not have a significant effect, indicating differences in environmental factors (SST, salinity and chlorophyll-a) associated with increasing distance from the Suez Canal were not a substantial factor in determining growth. The first year of growth was significantly greater than subsequent years for both species. Certain sites however, demonstrated significantly higher levels of growth for some annuli over others. Overall, the lack of a significant reduction in growth between sites indicates that growth is not a good predictor of invasion success, as the environmental factors expected to impact growth do not