Ocean currents take an important role for the transportation of animals inhabiting there. For example, Hernández-León et al.  showed that the trophic scenario of mesozooplankton could be depicted for the oceanic, upwelling and eddy system in the Canary Current System. Namely, the tremendous increase in production in the coastal area oﬀ Northwest Africa, the coupling of this production with the oceanic area through ﬁlaments and eddies topographically formed in the coast or shed by the islands promote a continuous transport of organic matter towards the deep ocean in the Canary Current System  . Pearce et al.  performed the application of an individual-based particle tracking model to the migration of tropical fish larvae along the continental shelf between the Houtman Abrolhos Islands and Rottnest Island (Western Australia). This application has shown that there was a potential for the southwards advection of passive particles/larvae in the Leeuwin Current system.
Several animals have been reported to be transferred by the Kuroshio which takes important roles for them to make life histories and their distributions, serving several resources for life. For example, the diving behavior of the false killer whale, Pseudorca crassidens was recorded using a pop-up archival transmitting (PAT) tag: “The whale reaches the Kuroshio front in the morning on one day in October, as indicated by the drastic change in the temperature gradient. The whale frequently performed deep dives exceeding 200 m during the daytime in both the transition and the Kuroshio front regions”  . The surface waters of the Kuroshio are generally regarded as nitrogen-deficient and oligo-trophic region  . Many economically important pelagic fish species, however, are abundant in these waters such as the dolphinfish, Coryphaena hippurus with annual catches of 9,500 tons by Taiwanese fisheries  . In the early and middle larval stages of round herring Etrumeus teres, seasonal variation in daily growth rates in Tosa Bay which is strongly influenced by the Kuroshio is largely determined by the sea temperatures experienced by hatch-date cohorts in the Pacific coastal waters off southern Japan  . Based on monthly samples collected from January 2001 to December 2004, in total, 2558 mesopelagic fish larvae were sampled also in the Tosa Bay  . Using the output of a high-resolution ocean general circulation model, particle-tracking experiments were performed to infer the distribution of larvae of the Japanese sardine (Sardinops melanostictus) and to detect effects of transport environment on sardine recruitment, and observed data of sardine spawning grounds during 1978-2004. By the 60th day following spawning, 50% of the larvae had been transported to the Kuroshio Extension  .
The only insects that inhabit the open ocean are members of the genus Halobates, commonly known as sea skaters. They are included in the family Gerridae (Heteroptera), which also includes common pond skaters or water striders. Only six species have been reported from open oceans among the 47 species included in the Halobates genus  . The six oceanic species are Halobates micans Eschscholtz, 1822, Halobates germanus White, 1883, Halobates sericeus Eschscholtz, 1822, Halobates splendens Witlaczil, 1886, Halobates sobrinus White, 1883  and Halobates princeps White 1883  which has been reported as a coastal species  in most cases. Another H. sp. is currently in the process of being described (Harada et al. unpublished).
Three species, H. sericeus, H. micans, and H. germanus, dominantly inhabit the tropical and temperate areas of the Pacific Ocean in the northern hemisphere, including the Kuroshio (Black Current), the East China Sea, and the Japan Sea     . Halobates micans is a cosmopolitan oceanic species which is mainly distributed in the latitude range of 20˚N to 20˚S in the Pacific, Indian and Atlantic oceans  . This species has also been found up to 32˚N only in the area south of the southern Japanese coast, perhaps being transported there by the Kuroshio  . The Kuroshio might be a transporting agent of oceanic Halobates, a point that remains to be studied.
Andersen and Cheng  presented a summary of data regarding the population density of oceanic Halobates. This review paper  shows that H. micans has a low population density of 2,000 individuals/km2 in the eastern tropical Pacific Ocean, even at the low latitude of 10˚ - 20˚N. Halobates micans and H. sericeus have average population densities of 3,000 and 7,000 individuals/km2, respectively, in the 13˚ - 20˚N area of the western Pacific Ocean  . Harada et al.  recently reported highly dense populations of H. micans of 10,000 individuals or more per km2 in the Tropical Western Pacific Ocean at 0˚ - 10˚N, 130˚ - 135˚E. However, the eastern region of 147˚ - 156˚E of the Tropical Western Pacific Ocean is less likely to be affected by freshwater flowing down from rivers in the tropical islands. The population density of H. micans there was only around 5,000 individuals/km2, half that at 130˚ - 135˚E  . In the area on the Kuroshio band of the eastern China Sea, all three species of H. micans, H. germanus, and H. sericeus have been collected at 7 sampling points from 27˚10'N, 124˚50'E to 29˚00'N, 129˚00'E, while the 127˚00' - 129˚00'E area and the area north of the Kuroshio was mainly dominated by H. sericeus  . However, few studies have been done on the role of the Kuroshio flowing near the southern shores of Japan islands for the distribution of Halobates.
This study aims to examine the population density of oceanic Halobates in the area inside and outside of the Kuroshio flowing near the southern shore in the direction of 100˚ - 120˚, and to discuss the role of the Kuroshio in the distribution of Halobates.
2. Materials and Methods
During the eight cruises, samplings were performed with one of two kinds of NEUSTON nets (small net: 0.57 m diameter, large net: 1.3 m diameter) and ORI net (1.5 m in diameter) during the dark, between 19:00 hrs and 05:00 hrs. The net was towed for 15 minutes once at the starboard side of the Research Vessel TANSEIMARU. The towing was replicated 2 to 8 times at each station with the ship moving at a speed of 2 - 2.5 knots. Figures 1(a)-1(e) show detailed information about place and season for each of the 8 cruises. The surface area swept by the nets was calculated as an expression: flow-meter value × inner diameter (0.57 m, 1.30 m or 1.50 m) of the nets. Population density of Halobates was
(a) (b) (c) (d)(e)
Figure 1. (a) The number of individuals (square of circles) and species ratio in September and October in relation with the current route of the Kuroshio at each sampling. Characters A, B, C and E correspond with Table 1(a)-A, B, C and E, respectively, in this manuscript. (b) The number of individuals (square of circles) and species ratio in June in relation with the current route of the Kuroshio at each sampling. No sea skaters were collected in D - D’ in December. Characters D and F correspond with Table 1(a)-D and F, respectively, in this manuscript. (c) The number of individuals (square of circles) and species ratio in July and August in relation with the current route of the Kuroshio at each sampling. Characters G and H correspond with Table 1(b)-G and H, respectively, in this manuscript. (d) The number of individuals (square of circles) and species ratio in September and October in relation with the current route of the Kuroshio at each sampling. Characters, I and J correspond with Table 1(b)-I and J, respectively, in this manuscript. (e) The number of individuals (square of circles) and species ratio in September in relation with the current route of the Kuroshio at each sampling. Characters K and L correspond with Table 1(b)-K and L, respectively, in this manuscript.
compared between inside and outside (around) the Kuroshio and different seasons.
Population density was presented as individuals collected per 1 km2 which was calculated as an expression: [the number of individuals collected × 1000000/(the surface area swept)].
Table 1(a) and Table 1(b) show the number of individuals (total of juveniles and adults) and population density of oceanic sea skaters at 12 sampling points during the 8 cruises (Figures 1(a)-1(e)). Population densities of Halobates micans, H. germanus, H. sericeus and H. sp range from 0 to 11,327.8, 0 - 49,338.8, 0 - 85,084.2 and 0 - 503.5 ind. km−2, respectively, inside and outside the Kuroshio (Table 1(a), Table 1(b)). In the area of the Pacific Ocean south of the southern coast of Japan (30˚00'N - 35˚00'N, 130˚25'E - 141˚04'E), H. sericeus dominated 6 of 12 sampling points whereas H. germanus and H. micans did 3 and 0 of the 12 points, respectively (Table 1(a), Table 1(b)). The average population density of H. sericeus was significantly greater inside the Kuroshio than outside the Kuroshio (Mann-Whitney U-test: z = −2.557, p = 0.011) (Table 2). Inside the Kuroshio, H. sericeus was dominant (Kruskal-Wallis test among three species of
Table 1. (a) Total number of individuals (larvae plus adults) and population density of oceanic sea skaters. (b) Total number of individuals (larvae plus adults) and population density of oceanic sea skaters. H.m.: Halobates micans, H. g.: H. germanus, H.s.: H. sericeus; H. sp.: A proposed new species, N: Number of individuals; ID: Individual density (Number/km2); SS: Surface area swept by Neuston nets.
Table 2. Population density (Mean ± SD [n], individual number/1 km2) inside and around the Kuroshio current. H.m.: Halobates micans; H.g.: H. germanus; H. s: H. sericeus; H. sp.: A proposed new species. When a sampling site has a current surface water speed more than 1 knot, this site was estimated as “inside the Kuroshio, when it does less than 1 knot, estimated as “around the Kurosio.
※Kruskal-Wallis test among three species of Halobates micans, H. germanus and H. sericeus: χ2-value, df, p.
H. micans, H. germanus and H. sericeus, χ2-value = 6.442, df = 2, p = 0.04) with a population density of 16,396.4 ind. km−2, whereas the density of H. germanus was 8,581.9 ind. km−2 which did not differ from the population density outside the Kuroshio (5,079.0 ind. km−2) (Mann-Whitney U-test: z = −0.074, p = 0.941) (Table 2).
The population densities of H. sericeus and H. germanus were significantly higher and more than 15,000 ind. km−2 in October inside and around the Kuroshio (Mann-Whitney U-test: H. sericeus, z = −2.328, p =0.020; H. germanus, z = −2.483, p = 0.013) than the other seasons, whereas there was no significant seasonal differences in population density of H. micans (z = −0.770, p = 0.441 between October and the other months) (Table 3). In total, the population density inside the Kuroshio was 27,234.5 ind. km−2 (±71,486.7 ) (Mean ± SD [n]) on average and tended to be higher than the population density in the area outside the Kuroshio (8,661.6 individuals km−2) (±13,343.1 ) (Mann-Whitney U-test: z = −1.763, p = 0.078). At the beginning of December (Table 1(d)), no oceanic Halobates were collected even in the the Kuroshio where the sea surface temperature was 25˚C.
Results of this study suggest that Halobates sericeus may be the dominant rider of the Kuroshio among the oceanic sea skaters studied. Andersen & Cheng  reported that the distribution of this species was limited to 13˚ - 45˚N in the Pacific Ocean. However, estimated suggest was a moderate population density of 2,010 - 7,100 individuals/km2 in the tropical area from 0˚ - 10˚N (including the area around the equator) in the Western Tropical Pacific Ocean (Harada et al. 2010)  . This species may have been actively transferred by several currents, including the Kuroshio, the North Equator Current, the
Table 3. Population density (Mean ± SD [n], individual number/1 km2) in several months inside and around the Kuroshio. H.m.: Halobates micans; H.g.: H. germanus; H. s: H. sericeus; H. sp.: A proposed new species.
#Mann-Whitney U-test between population density in October and that in the other months.
Mindanao Current and the North Equator Counter Current, widely throughout the Western Pacific Ocean. The Kuroshio may be transporting a relatively large number of H. sericeus individuals (with a density of 1,1721.6 individuals/km2) in the relatively northern area south of the Japanese archipelago.
Harada et al. (2010)  estimated a high population density of 13,000 to 60,000 H. germanus individuals/km2 throughout the wide longitudinal area from 130˚ - 156˚E (0˚ - 10˚N) in the Western Tropical Pacific Ocean. In this study, a moderate population density of around 6,000 individuals/km2 of H. germanus was estimated both inside and outside the Kuroshio in the northern Pacific Ocean area south of the southern coast of Japan. The genetic variation of the populations of H. germanus living outside several currents flowing in the wide range of latitudes from the equator to 35˚N might lead to the result of a lack of genetic crossing, because, for example, there are no opportunities for the population north of the Kuroshio to mate with the other populations north of the North Equator Counter Current, for example, excluding the special occasion of the transportation by typhoons moving south to north in the Pacific Ocean.
The species ratio of H. micans decreases greatly in October and November. This phenomenon suggests that H. micans populations that are transported from low latitude areas may show low reproductive activity under low air temperature conditions in the fall, and suggests that this species is actually a tropical species.
For the future study, the interaction between populations inside and outside the Kuroshio and that between populations inhabiting the two currents of the Kuroshio and the North Equator Current should be tested by examining genes such as mtDNA of the individuals from the two populations. If it would be possible, the “catch, mark and release” project would be exciting, releasing many adult sea skaters which have been collected, marked with color paints and released on the North Equator Current and then, catching them on the Kuroshio, for example.
We would like to thank JAMSTEC for permission to conduct these studies during the cruises on the R/V TANSEIMARU. The samplings were also possible due to support from all of the crew (Captains: Mr. Fujio INABA, Mr. Ushio FUJITA, Mr. Shoichi SUZUKI and MR. Takatoshi SEINO) and all the scientists and engineers from MWJ (Marin Work Japan) on these cruises. We would like to give them special thanks. Thanks are also due to Ms Laura Sato who is a professional English editor and checked very carefully this manuscript before submission, and anonymous reviewers for invaluable comments on this study.