Does larval duration contribute to population genetic isolation of the Japanese eel Anguilla japonica?

Japanese eel, Anguilla japonica Temminck & Schlegel, is a temperate catadromous fish with a long migratory loop and lengthy leptocehalus stage. Its freshwater distribution ranges from Taiwan, through Mainland China, Korea, and north to Japan (Tesch 2003). The spawning ground of this species is presumed to be in the western Mariana Islands near 14–16 N 142 E (Tsukamoto 1992, 2006). The leptocephalus larvae drift from their spawning grounds with the North Equatorial Current (NEC) followed by Kuroshio Current (KC) to reach the coasts of Northeast Asia (Tsukamoto 1992; Cheng & Tzeng 1996). They then metamorphose into glass eels along continental shelf and enter estuaries and fresh water for growth. In the past, the concept of panmictic populations for Japanese eels was accepted based on evidence from mtDNA sequences (Sang et al. 1994; Ishikawa et al. 2001). However, Tseng et al. (2006) divided the genetic populations of the Japanese eel into low-latitude (South China and Taiwan) and high-latitude (Japan, Korea and North China) groups based on microsatellite DNA. They suggested that most progeny tend to be transported back to similar locations as their ancestors. The mechanism of how Japanese eel is separated into two stable populations remains a puzzle. Japanese eel appears to spawn in a restricted area and the larvae are passively transported by the NEC and KC, which exhibit considerable changes in speed, eddy structure and route at daily, monthly or even annual levels (reviewed in Aoyama 2009). If two genetic populations for the Japanese eel exist, there must be a mechanism for the non-random return of larvae to the place where their parents formerly resided. In the wild, populations of many organisms are composed of a mixture of individuals that reproduce at different times within a reproductive season