Age and Size Differences in Sexually Mature Thorny Skates: Implications for Fisheries Management in the Gulf of Maine

Primary tabs

Project Type: 
Research
Project Number: 
R/MED-5
Inception Date: 
2005
Completion Date: 
2006
Theme Area: 
Fisheries Resources

Participants:

Paul Tsang UNH - Department of Molecular, Cellular & Biomedical Sciences Principal Investigator
James Sulikowski University of New England Co-Principal Investigator
Patrick Danley University of Maryland Co-Principal Investigator

Students Involved:

Matthew Salomon University of Florida
Jeff Kneebone UNH - Department of Molecular, Cellular & Biomedical Sciences
Dejerianne Ostrow University of Florida
Abstract: 
The primary objective of the proposed study is to investigate a disparity in age and size at sexual maturity observed in a localized population of thorny skates within the western Gulf of Maine. This disparity manifests as two statistically different reproductive groups: a larger group of sexually mature skates that has an average total length of 91.5 cm, an average weight of 16.5 lbs and an average age of 15.1 years, and a smaller group of sexually mature skates that has an average total length of 59 cm, an average weight of 4 lbs and an average age of 9.4 years.  We will address this disparity by first testing a hypothesis to determine the source of this observed phenomenon with the use of genetic techniques. This hypothesis is divided into two alternative sub-hypotheses (1a and 1b). For sub-hypothesis 1a, we will determine if smaller, early-maturing thorny skates are migrating from Canadian waters into the Gulf of Maine, and for sub-hypothesis 1b, we will determine if the large and small size groups of sexually mature thorny skates are reproductively isolated and do not belong to a single genetically cohesive species.
Objectives: 

In a localized population of thorny skates within the western Gulf of Maine, we observed a large and small reproductive group that significantly differed in age and size at sexual maturity. In order to begin to understand the scope and the processes that have led to this phenomenon, our objective is:

1a) To determine if the smaller, early-maturing thorny skates are migrating from Canadian waters into the Gulf of Maine

1b) To determine whether the small and large size groups of sexually mature thorny skates are reproductively isloated and thus do not belong to a single genetically cohesive species

Methodology: 

Skates will be collected from area 132, the 30-minute square located in the western region of the Gulf of Maine, and from three Canadian areas (Scotian Shelf, Grand Bank and Newfoundland Shelf). Gross morphological characteristics and light microscopy (histology) will be used to assess sexual maturity. A polymerase chain reaction-based DNA finger-printing method, Amplified Fragment Length Polymorphism (AFLP), will be used to assess genetic variation and estimate genetic cohesiveness of the large and small size groups.

Rationale: 

Accurate and comprehensive life history studies contribute vital information (e.g., age, growth, genetics, reproductive biology) to the development of a management plan. In the proposed study, we will use AFLP to determine if a single species exists or whether genetic cohesiveness has been lost within the thorny skate population in the Gulf of Maine. This information will be submitted to the New England Fishery Management Council (NEFMC) to facilitate the development of a management plan and its submission to the National Marine Fisheries Service (NMFS) and the Secretary of Commerce for approval and implementation.

Accomplishments: 

There are no microsatellite markers for skates, resulting in a lack of knowledge regarding genetic structure for these elasmobranchs. Consequently, this represents a significant limitation to their proper management. For this project, we successfuly characterized nine polymorphic microsatellite loci in thorny skates from the Gulf of Maine. Currently, we are employing these microsatellite loci to assess thorny skate population structure.

Genetic Component
 
Initial attempts to elucidate genetic population structure within the Gulf of Maine suggested very low levels of diversity in the mitochondrial control region. Because the maternal inheritance of mitochondrial DNA can bias measures of population structure, microsatellite markers were developed to complement the mitochondrial data.
 
Four microsatellite libraries were created (letters in parentheses indicate nucleotide repeats targeted in the library construction): 1) dinucleotide (CA) library from the small phenotype, 2) dinucleotide (CA) library from the large phenotype, 3) dinucleotide (CA) library from a mixed target (DNA from a large phenotype, a small phenotype, and a Canadian individual combined in equal proportions), and 4) trinucleotide (ATG) library from a mixed target (DNA from a large phenotype, a small phenotype, and a Canadian individual combined in equal proportions). (Three additional libraries were created but were not screened extensively because suboptimal levels of microsatellite enrichment were detected.)
 
A total of 384 sequences were generated from libraries 1 and 2. Nine of these sequences contained repeats suitable for designing primers. Four of these loci amplified well (determined by running PCR on a 1.5% agarose gel), however primers were re-designed in an effort to minimize non-specific amplification. New primers for these four loci were fluorescently labeled for capillary electrophoresis on a MegaBACE 1000 DNA Analysis System (Amersham Biosciences/GE Healthcare). Two of these re-designed primers did not amplify consistently enough for capillary electrophoresis genotyping despite efforts to optimize the reactions. There was no polymorphism detected at the remaining two loci when a subset of individuals was genotyped.
 
Due to the low levels of success and the high cost of custom sequencing, a more conservative approach was selected for screening libraries 3 and 4. The protocol of Wang et al. (2007) was followed to avoid the cost of extensive sequencing. To date, 17 sequences have been generated. Four of these sequences contained microsatellite repeats, 3 of which were of adequate quality to generate primers.
 
Additionally, five microsatellite primers identified in Raja clavata L. (Chevolot et al. 2005) were screened for cross amplification. Two of these loci (Rc-B3, Rc-E9) amplified consistently and appeared polymorphic (when PCR products were run on a 2% agarose gel). These primers were fluorescently labeled, and PCR products were run on a MegaBACE 1000 DNA Analysis System (Amersham Biosciences/GE Healthcare). Locus Rc-B3 did not show polymorphism when a subset of individuals was genotyped. Locus Rc-E9 did not amplify consistently enough for capillary electrophoresis genotyping despite efforts to optimize the reaction.
 
Finally, publicly available EST sequences from Leucoraja erinacea were scanned for perfect microsatellite repeats and primers were designed using BatchPrimer3 v1.0 (http://probes.pw.usda.gov/batchprimer/). A total of 17,314 microsatellite repeats were detected in the EST sequences. Initially, 45 primers were selected based on repeat length and motif. Primers were screened to determine amplification reliability and polymorphism (PCR products were run on a 2% agarose gel). Loci that amplified well and indicated polymorphism were then labeled with a fluorescent dye, and individuals were genotyped on an ABI 3730xl DNA Analyzer (Applied Biosystems). To date, a total of 95 individuals (35 representing the CANADA population, 34 representing the large phenotype, and 26 representing the small phenotype) have been genotyped at five loci.
 
Preliminary analysis of microsatellite data (5 loci):
 
The main goal of this study was to determine if the two phenotypic classes, large and small morphologies, of Amblyraja radiata sampled from the western Gulf of Maine constitute two genetically distinct populations or a single panmictic population. Additionally, we investigated the level of genetic exchange between A. radiata from the western Gulf of Maine and CANADA to test if the small phenotypic class represents a founder population from CANADA. Throughout this section each of the Canadian sampling locations and the two Gulf of Maine phenotypes will be referred to as populations for simplicity, though this terminology does not indicate that each locale is genetically distinct.
 
We implemented the AMOVA (analysis of molecular variance) test in Arlequin (Schneider et al. 2000) to determine the hierarchical level within our sampling that contributed the largest variance component. The low among group variance found with the AMOVA suggests that no significant population structuring can be detected between the two phenotypic classes of A. radiata from the Gulf of Maine or between A. radiata samples from the Gulf of Maine and CANADA (Table 1). Overall, 98% of the genetic variation was explained by within population variation, while less than 2% could be attributed to variation among different populations.
 
We then used a model-based clustering algorithm to infer population structure and assign individuals to groups. This approach minimizes the introduction of bias due to the geographic origin of a sample by probabilistically assigning each individual to a group based on allele frequencies at each locus. This algorithm was implemented in the software package STRUCTURE (Pritchard et al. 2000) using a model of admixture among all individuals. STRUCTURE was unable to recover groupings that fit the geographical sampling. When given a parameter set indicating samples were drawn from two populations (the hypothesis being that a Canadian and a Gulf of Maine population exists), the program allocated individuals from the four sampling locations evenly between the two putative populations. When given a parameter set indicating samples were drawn from three populations (the hypothesis being that the two Gulf of Maine phenotypes differ genetically from each other and from Canadian individuals), the program allocated individuals from the four sampling locations evenly between the three putative populations. When given a parameter set indicating samples were drawn from four populations (the hypothesis being that the two sampling locations within Canadian waters differ from each other and from the two genetically distinct Gulf of Maine phenotypes), the program allocated individuals from the four sampling locations evenly between the four putative populations. The inability of the clustering algorithm to determine one parameter set as more likely than the others suggests that there is no genetic structuring among the four sampling locations.
 
Finally, the number of individual migrants (M) per generation between the Gulf of Maine and the CANADA samples was estimated with the software package MIGRATE (Beerli 2008). The estimated number of migrants per generation for each of the pairwise comparisons indicated large amounts of gene flow, suggesting that genetic isolation has not developed between any of the sampled groups.
Literature Cited
 
Beerli, P (2008) Migrate: estimation of population sizes and gene flow using the coalescent. Ver 2.4. School of Computational Science and Department of Biological Sciences, Florida State University.
 
Chevelot, M, Reusch, TBH, Boele-Bos, S, Stam, WT, and Olsen, JL (2005) Characterization and isolation of DNA microsatellite primers in Raja clavata L. (Thornback ray, Rajidae). Molecular Ecology Notes, 5, 427-429.
 
Pritchard, JK, Stephens, M, and Donnelly, P (2000) Inference of population structure using multilocus genotype data. Genetics, 155, 945-949.
 
Schneider, S, Roessli, D, and Excoffier, L (2000) Arlequin: A software for population genetics data analysis. Ver 2.000. Genetics and Biometry Lab, Dept. of Anthropology, University of Geneva.
 
Wang, XW, Trigiano, RN, Windham, MT, Devries, RE, Scheffler, BE, Rinehart, TA, and Spiers, JM (2007) A simple PCR procedure for discovering microsatellites from a small insert libraries. Molecular Ecology Notes, 7, 558-561.

Publications

Available from the National Sea Grant Library (use NHU number to search) or NH Sea Grant

Journal Article

  • Salomon, M., D. Ostrow, A. Clark, J. Kneebone, J. Sulikowski and P. Tsang (2010). Characterization of 9 polymorphic microsatellite loci in the Thorny skate, "Amblyraja radiata," from microsatellite-enriched libraries and EST sequences. Molecular Ecology Resources 10(3):576-579, March 2010.