Allan E. Strand

The usefulness of the polymerase chain reaction with restriction fragment length polymorphism (PCR-RFLP) and partial sequencing of the cytochrome oxidase I (COI) gene was tested regarding the utility of these techniques in unraveling philometrid life cycles and, in particular, to determine putative paratenic host species. Our focus was to study three species of philometrids commonly found in the estuaries of South Carolina: Philometroides paralichthydis and Philometra overstreeti from the southern flounder, Paralichthys lethostigma and Philometra carolinensis from the spotted seatrout, Cynoscion nebulosus. A total of 473 fish in 19 species known to be potential prey of the spotted seatrout and the southern flounder were dissected. Of all nematode larvae found in the mesenteries of 53 fish of 10 species, 21 specimens were determined to be philometrids using PCR. The use of PCR-RFLP allowed the identification of larvae of P. carolinensis and P. overstreeti in the freshwater goby, Ctenogobius shufeldti, and P. overstreeti in mummichog, Fundulus heteroclitus. However, 12 RFLP profiles could not be matched to control species, thus demonstrating the limitation of this technique in areas where diversity of philometrids is not well known and higher than anticipated. Similarly, COI procedures provided unknown sequences that did not match those of nine philometrid species used as controls. We concluded that although both techniques showed some usefulness and promise, at this point, however, they demonstrate the need of increasing our knowledge of marine and estuarine philometrid biodiversity.

Individual‐based, spatially explicit models provide a mechanism to understand distributions of individuals on the landscape; however, few models have been coupled with population genetics. The primary benefits of such a combination is to assess performance of population‐genetic estimators in realistic situations. kernelpop represents a flexible framework to implement almost any arbitrary population‐genetic and demographic model in a spatially explicit context using a variety of dispersal kernels. Estimates of type I error associated with genome scans in metapopulations are provided as an illustration of this software's utility.