Population structure and genetic connectivity of the scalloped hammerhead shark (Sphyrna lewini) across nursery grounds from the Eastern Tropical Pacific: Implications for management and conservation

Abstract

Defining demographically independent units and understanding patterns of gene flow between them is essential for managing and conserving exploited populations. The critically endangered scalloped hammerhead shark, Sphyrna lewini, is a coastal semi-oceanic species found worldwide in tropical and subtropical waters. Pregnant females give birth in shallow coastal estuarine habitats that serve as nursery grounds for neonates and small juveniles, whereas adults move offshore and become highly migratory. We evaluated the population structure and connectivity of S. lewini in coastal areas and one oceanic island (Cocos Island) across the Eastern Tropical Pacific (ETP) using both sequences of the mitochondrial DNA control region (mtCR) and 9 nuclear-encoded microsatellite loci. The mtCR defined two genetically discrete groups: one in the Mexican Pacific and another one in the central-southern Eastern Tropical Pacific (Guatemala, Costa Rica, Panama, and Colombia). Overall, the mtCR data showed low levels of haplotype diversity ranging from 0.000 to 0.608, while nucleotide diversity ranged from 0.000 to 0.0015. More fine-grade population structure was detected using microsatellite loci where Guatemala, Costa Rica, and Panama differed significantly. Relatedness analysis revealed that individuals within nursery areas were more closely related than expected by chance, suggesting that S. lewini may exhibit reproductive philopatric behaviour within the ETP. Findings of at least two different management units, and evidence of philopatric behaviour call for intensive conservation actions for this highly threatened species in the ETP.

Fig 1. Location of sampling sites of Sphyrna lewini in the Eastern Tropical Pacific.

Sampling sites: Guatemala (GUA, N = 72), Ojochal (OJO, N = 43), Coyote (COY, N = 34), Cocos Island (ICO, N = 15), Panama (PAN, N = 65), Nayarit (NAY, N = 25), Oaxaca (OAX, N = 8), Michoacan (MCH, N = 17), Baja California (BJC, N = 25), Chiapas (CHP, N = 14), Sinaloa (SIN, N = 36), Port Buenaventura (PTB, N = 22), Sanquianga (SNQ, N = 20), Utria (UTR, N = 21), Malpelo Island (MLP, N = 18). Sampling sites were plotted using base and raster layers from the Natural Earth (public domain) http://www.naturalearthdata.com/ in ArcMap 10.4.

Fig 2. Haplotype network based on mitochondrial control region sequences for Sphyrna lewini.

Each circle represents a unique haplotype (Haplotype 1 through 16); the size of the circle is proportionate to the number of individuals; the colors represent the proportion of individuals from each sampling location; ticks on connecting lines indicate mutational steps between haplotypes. Sampling sites: Guatemala (GUA), Ojochal (OJO), Coyote (COY), Cocos Island (ICO), Panama (PAN), Nayarit (NAY), Oaxaca (OAX), Michoacan (MCH), Baja California (BJC), Chiapas (CHP), Sinaloa (SIN), Port Buenaventura (PTB), Sanquianga (SNQ), Utria (UTR), Malpelo Island (MLP).

Fig 3. Distribution of pairwise genetic relatedness.

Distribution of pairwise genetic relatedness values for simulated pairs of individuals: Full siblings (FS), Half siblings (HS), Parent/Offspring (PO), and for observed pairs of individuals of Sphyrna lewini sampled in coastal areas of the ETP.

Fig 4. Population structure analyses from microsatellite genotypes of Sphryna lewini individuals in sampling sites of the Eastern Tropical Pacific: Guatemala (GUA), Costa Rica (OJO), Panama (PAN), and Cocos Island (ICO).

A) DAPC plot from the first and second components of the nuclear microsatellite genotypes of three coastal areas B) Genetic clusters inferred by STRUCTURE with K = 2, K = 3 and K = 4 of three coastal areas. C) DAPC plot from the first and second components of the nuclear microsatellite genotypes of three coastal areas and an oceanic island. D) Genetic clusters inferred by STRUCTURE with K = 2, K = 3 and K = 4 of three coastal areas and an oceanic island.

Fig 5. Contemporary gene flow estimated from 9 microsatellite loci genotypes with the divMigrate function.

Arrows represent the relative number of migrants and estimated direction of gene flow between three coastal areas: Guatemala (GUA), Costa Rica (OJO), Panama (PAN); and an oceanic island: Cocos Island (ICO). The darker the arrow, the higher the relative number of migrants between sampling locations.

Fig 6. Distribution of pairwise relatedness values of the Wang estimator of Sphyrna lewini individuals within same sampling sites and between different sampling sites.

The mode of each distribution is presented in a black dashed line.