Pilkey et al. (2009) present a well-illustrated paper on a great variety of mostly inshore islands that occur in enormous numbers worldwide. My issue is mostly with the term “barrier,” questionably attached to these features. In several key points, this designation conflicts with conventional coastal terminology, indeed with the universal consensus. Some of the cited landforms are actually not islands but mainland beach ridges that fringe mangrove swamps, permafrost-thermokarst tundra (Figure 13H), or glacial sandur and/or fjord-head coastal surfaces along the mainland shore (Figure 13E). Others represent recurved sand spits, also attached to the mainland shore (e.g., Figure 5E). Those that actually are islands, located inside protected bays or offshore, often even lack rudimentary, thin, narrow, at least semicontinuous beach foreshore lithosomes.
A quick sampling of the vast modern coastal literature (e.g., Davis, 1992, 1995; Fisher, 1982; Neuendorf et al., 2005, p. 54; Oertel, 1985) clearly indicates that the defining common denominator of “barrier island” is the position and function of an island as a barrier; a supratidal, elongated, and narrow landform that lies seaward, off the mainland shore. Unlike sand-veneered and/or sand-fringed inshore islands and islets, they represent a sharp boundary between contrasting bathymetric and sedimentary settings. Barrier islands separate shallow brackish inshore basins (bays, elongated lagoons, sounds), usually characterized by muddy bottoms, from sand-floored margins of deep marine basins. They provide physical, biological isolation; protection of lagoons and the mainland coast and, most importantly, mitigate critical marine influences, including sand influx from offshore, high marine salinities, and storm impact.
Even sediment-starved very low islands, detached by subsidence from deltaic marsh plains and with insignificant Gulf beaches, as the drastically reduced Derniere island chain off the central Mississippi delta plain of Louisiana, have been accepted as valid barriers. Similarly, man-made and well-nourished, robust anthropogenic (“anthropic”) islands, such as Sand Island, in the path of abundant westward-directed longshore drift west of Petit Bois Island, Mississippi (Otvos and Carter, 2008) may originate and evolve in actual, true barrier island settings.
Fetch, just as other hydrodynamic factors and various pertinent sediment sources, does play a vital role both in the functions and ongoing development of barrier islands, dubbed “open-ocean” or “classic” barrier islands by Pilkey and his coauthors. However, the use of that designation for a plethora of sundry inshore islands, islets, and a broad range of other landforms, both in paralic and in offshore settings, is objectionable. Lewis, Cooper, and Pilkey (2005) and, rather surprisingly, Davis himself (1994, p.1), did once before employ the word barrier for bay islands that under the influence of short fetch and weak littoral drift may develop minor- to- insignificant beaches, beach ridges, and sand spits. However, the cited crucial “barrier” function is completely absent from these “fetch-limited” inshore and nearshore islands, islets, and other cited coastal landforms. In conclusion, the term fetch is out of place when employed in this context, a confusing and invalid application of the term barrier island.