Seasonal and diel changes in the vertical distribution of oncaeid copepods in the epipelagic zone of the Kuroshio Extension region

Transcription

1 Plankton Benthos Res 9(1): 1 14, 2014 Plankton & Benthos Research The Plankton Society of Japan Seasonal and diel changes in the vertical distribution of oncaeid copepods in the epipelagic zone of the Kuroshio Extension region HIROSHI ITOH 1, *, KAORU NAKATA 2, KATSUYUKI SASAKI 3, TADAFUMI ICHIKAWA 4 & KIYOTAKA HIDAKA 4 1 Suidosha Co. Ltd., , Ikuta, Tama-ku, Kawasaki, , Japan 2 Fisheries Research Agency, 15F, Queen s Tower B, 2 3 3, Minato Mirai, Nishi-ku, Yokohama, , Japan , Nishi 23, Minami 7, Chuou-ku, Sapporo, , Japan 4 National Research Institute of Fisheries Science, , Fukuura, Kanazawa-ku, Yokohama, , Japan Received 17 June 2013; Accepted 9 November 2013 Abstract: Species composition and vertical distribution of oncaeid copepods, which are potentially important prey for juvenile fish, were investigated in the Kuroshio Extension region, the NW Pacific, in April, August, November 1998 and February Samples were collected from 8 discrete layers in the epipelagic zone (0 200 m depth) using MOCNESS (0.064 mm mesh) during both day and night. Thirty-five oncaeid species were identified. Oncaea (s.l.) zernovi and Spinoncaea ivlevi were numerically the dominant species comprising % and %, respectively, of adult oncaeid copepods in the epipelagic zone. Cluster analysis on all samples revealed that these were separated into three groups with discrete vertical ranges; the first one appearing in the 0 50 m depth surface layer in April and August and consisting mainly of Oncaea (s. str.), the second one located in the deepest layer and composed mostly of O. zernovi and S. ivlevi with some mesopelagic species, and the third one located above the second one and having intermediate species composition. Species-specific vertical distributions indicate that most oncaeid populations shifted downward from August to November, when the thermocline remarkably descended. However, most Oncaea spp. did not show a downward shift with the thermocline, and were positively correlated to appendicularian abundances, suggesting that appendicularian houses, known to be oncaeid habitats and to provide food, were a possible factor affecting their vertical distribution. Niche partitioning, allowing coexistence of congeners, might be explained by differences in body size and distribution layers in Oncaea and by differences in distribution layer in Triconia. Key words: Copepoda, Kuroshio Extension, Oncaeidae, seasonal change, vertical distribution Introduction The Kuroshio Extension region is in the northern part of the subtropical circulation of the NW Pacific, where water temperatures show remarkable seasonal changes in the epipelagic zone due to its location in the temperate zone. This area is known to be important in the reproduction of commercially utilized fish species such as the Japanese sardine Sardinops melanostictus (Temminck et Schlegel) and the Pacific saury Cololabis saira (Brevoort) (Noto & Yasuda 1999, Tian et al. 2004). As the survival success of juvenile * Corresponding author: H. Itoh; , saphirella@nifty.com fish depends strongly on an adequate food supply, especially on the supply of small copepods, spatial and temporal changes in the copepod fauna substantially influence the recruitment of fish populations. The importance of small copepods as food supplies, has been shown by many studies on the larvae and juveniles of Japanese sardine, the Japanese anchovy Engraulis japonicus (Houttuyn), the chub mackerel Scomber japonicus Houttuyn, the spotted mackerel Scomber australasicus Cuvier, and the Pacific saury in the transition area between the Kuroshio Extension and the Oyashio Current (Takagi et al. 2009, Odate 1977). Especially in the case of the Pacific saury, occasionally high relative abundances of oncaeid copepods (>70%

2 2 H. Itoh et al. of food items) in their stomach contents have been reported (Odate 1977). Previous studies on zooplankton in the Kuroshio Extension region revealed a higher zooplankton diversity consisting of warm water species but a low biomass compared to the Oyashio and the transition region (Odate 1962, 1980). However, since these studies used 335 μm mesh nets, through which most small copepods are lost, our knowledge on the food environment provided by smallsized copepods for fish larvae and juveniles is still limited. The present study used nets of 64 μm mesh to sample small oncaeid copepods in the epipelagic zone of the Kuroshio Extension region; the term oncaeid in this study does not mean the genus Oncaea but the family Oncaeidae. This family represents the main component in regard to numerical abundance and species richness of the small copepods in the area. Oncaeidae is a diverse family including more than 100 species (Böttger-Schnack et al. 2004), and more than 40 species have been reported from the Kuro shio Current region of Japan (Nishibe et al. 2009). Some species belonging to this family utilize appendicularian houses as their habitat and food (Ohtsuka et al. 1993), suggesting that the abundance of oncaeid copepods may be related to appendicularian abundances. The vertical distribution of oncaeid copepods has been investigated previously in the Atlantic Ocean (Boxshall 1977), the Mediterranean Sea (Böttger-Schnack 1994, 1997), the Adriatic Sea (Krŝinić 1998), the Red Sea (Böttger 1987, Böttger-Schnack 1988, 1990a, b, 1992, 1995, Böttger-Schnack et al. 2001, 2004, 2008), the Arabian Sea (Böttger-Schnack 1994, 1996), the Indian Ocean (Tsalkina 1972), the South Pacific Ocean (Heron & Bradford-Grieve 1995), and the North Pacific Ocean (Furuhashi 1966, Zalkina 1970, 1977, Nishibe & Ikeda 2004, Nishibe et al. 2009). Most of these studies have dealt with large scale vertical distributions from the surface to the bathypelagic zone. In contrast, the present study investigated seasonal and diel changes in the fine scale vertical distributions of oncaeid copepods in the epipelagic zone of the Kuroshio Extension region, where there has as yet been no studies on the vertical distribution of oncaeid copepods. After the intensive taxonomic revision of oncaeid copepods of the last two decades (e.g. Heron & Bradford-Grieve 1995, Böttger-Schnack 1999, 2001, 2002, 2003, 2005), this is the second report (after Nishibe et al. 2009) of the speciesspecific vertical distribution and the first report of their vertical diel migration in the subtropical Pacific Ocean following the up-to-date taxonomy of oncaeid copepods. The results are discussed with respect to abiotic and biotic environmental factors, mechanisms of coexistence of congeneric species, and their potential significance as food resources for fish larvae and juveniles. Field sampling Materials and Methods Field surveys were conducted on 17 and 18 April (spring), 21 August (summer), 10 November (autumn) 1998 and 18 and 23 February (winter) 2001 at seven stations in the Kuroshio Extension and the adjacent offshore regions, east of Honshu, Japan (Fig. 1) using the R.V. Soyo-Maru, National Research Institute of Fisheries Science. The plankton samples were collected from eight discrete layers (0 30 m, m, m, m, m, m, m, m) by oblique hauls of a MOCNESS (mouth area 0.25 m 2, mesh opening mm) in the daytime ( ) and nighttime ( ); henceforth the day and night samples are referred to by the letters D and N, respectively. Sampling was not conducted at the same stations between daytime and nighttime in the spring, autumn and winter investigations. The samples were immediately fixed and preserved in 5% formalin-seawater solution. The volume of filtered water was estimated from the reading of a flow meter (Tsurumi- Seiki Co. Ltd., Japan) attached to the mouth of the net. The vertical profiles of temperature and salinity were measured with a CTD (SBE 9 plus, Sea-Bird Electronics, Inc., USA) incorporated on a rosette sampler at each station. The CTD data in August were measured only in the daytime since sampling was done at the same station in the daytime and nighttime. Water samples for chlorophyll-a (Chl a) concentration determination were collected by a rosette sampler equipped with Niskin bottles at 0, 10, 30, 50, 75, 100, 125, 150 and 200 m depths. From each depth, a 200 ml water sample was filtered on a Whatman GF/F filter, then transferred to a plastic vial containing 6 ml of dimethylformamide (Suzuki & Ishimaru 1990) and stored at 80 C until Chl a concentration was measured in the laboratory with a fluorometer (Turner 10-AU, Turner Designs, Fig. 1. Location of sampling stations, off south eastern Japan.

3 Vertical distribution of oncaeid copepods 3 Inc., USA). Enumeration and species identification In the laboratory, large organisms such as medusae and chaetognaths were removed from the whole sample before subsampling. A 5% formalin-seawater solution was added to the whole sample to adjust its volume to 50, 100 or 200 ml. Subsampling from the adjusted sample with a 1 ml Stempel pipette and enumeration under a stereomicroscope were repeated until the total number of enumerated copepods exceeded 200 for enumeration of total copepods and appendicularians, and until the total number of adults exceeded 300 for oncaeid copepods. Species identification followed Shmeleva (1969), Gordeyeva (1972), Heron (1977), Krŝinić & Malt (1985), Boxshall & Böttger (1987), Böttger-Schnack & Boxshall (1990), Heron & Bradford-Grieve (1995), Böttger-Schnack (1999, 2001, 2002, 2003, 2005), Huys & Böttger-Schnack (2007) and Wi et al. (2011). Taxonomic classification of the genera followed Boxshall & Halsey (2004), Böttger- Schnack (2001) and Böttger-Schnack & Schnack (2013). The family Oncaeidae currently consists of seven genera, of which three, i.e. Triconia, Monothula and Spinoncaea, have species moved from Oncaea. Although Böttger- Schnack (2001) redefined the diagnostic characters of Oncaea, some species that do not belong to Oncaea sensu Böttger-Schnack (2001) are still classified into the genus Oncaea. As a result, the name Oncaea presently has two meanings, i.e. Oncaea sensu Philippi (1843) and Oncaea sensu Böttger-Schnack (2001); in some papers they are referred to as Oncaea sensu lato (s. l.) and Oncaea sensu stricto (s. str.), respectively. To simplify the genus and species names in the following text and to follow the latest taxonomy of Oncaeidae, we use the name Oncaea only for Oncaea s. str. and Oncaea for Oncaea s. l. instead of adding these Latin phrases to every species name between its generic and specific names. Males of Triconia, except for T. conifera (Giesbrecht, 1891) and T. furcula, were compiled into two groups [T. elongata Böttger-Schnack, allied species and T. minuta Giesbrecht, allied species] because they were easily distinguished at the group level but were difficult to identify to species level with the limited magnification ( 40 or 100) of the stereomicroscope. Specimens identified as Spinoncaea ivlevi (Shmeleva, 1966) and Oncaea zernovi Shmeleva, 1966 may have included their sibling species S. humesi Böttger-Schnack, 2003 and O. bispinosa Böttger-Schnack, 2002, respectively, because of the difficulty in identification under low magnification. These sibling species have recently been reported from Tosa Bay (Nishibe et al. 2009), which is located upstream of the Kuroshio Extension. Individuals referred to as O. vodjanitskii Shmeleva & Delalo, 1965 included O. vodjanitskii and a species similar to O. atlantica Shmeleva, Oncaea sp. 1, an undescribed small species of 0.25 mm body length, is allied to O. vodjanitskii, but could be clearly separated from its congeners by a dorsal projection on the first pedigerous somite. Three forms (large, medium and small forms) of Oncaea venusta Philippi, 1843 occurred, and they were distinguishable from each other by differences in body length, shape of the second thoracic segment and morphology of the genital double-somite (see Böttger-Schnack & Huys 2004). According to a recent molecular analysis (Elvers et al. 2006), the large and small forms of O. venusta represent different genetic lineages, whereas the medium form includes two genetic lineages. To investigate the size-dependent niche partitioning of coexisting species in the same depth layer, body length (from the anterior end of head to the posterior end of the furca) was measured for 20 females of the target species belonging to Oncaea and Triconia collected in August Data analysis The data analyses of vertical distribution were performed for the 23 species for which the lowest count in a subsample exceeded 10 individuals. For the cluster analysis of oncaeid communities, the Bray-Curtis dissimilarity index was calculated using population density values transformed by log (x 1), and ordination on a two dimensional map by non-metric multidimensional scaling (NMDS) was also conducted. Construction of the dendrogram and NMDS were performed using the add-in software packages Cluster analysis Ver. 3.3 (Hayakari 2001) and Systat 11 (HULINKS Inc.), respectively. Relationships between the communities and environmental factors were shown on the NMDS ordination map with arrows indicating gradients of the factors according to Kruskal & Wish (1978). To test the relationship between population density of each species and environmental factors, Spearman s rank correlation coefficient was calculated. For the description of vertical distributions, the depths of 25th, 50th, 75th percentiles of a population (D 25%, D 50%, D 75% ) were calculated for each species. Interspecific comparison of D 50% was done by the Wilcoxon signed-rank test. An analysis of variance (ANOVA) was carried out with the mean population density among communities and body length among species. Tukey s procedure was used as a post hoc test when the ANOVA was significant. Hydrographic conditions Results Water temperature at the sea surface showed a remarkable seasonal change between 18.8 and 29.2 C, whereas it ranged only from C at 200 m depth (Fig. 2). Mean temperature in the surface mixed layer (defined as the layer from the sea surface to just above the seasonal thermocline) was C in April, 29.1 C in August,

4 4 H. Itoh et al. Fig. 2. Vertical profiles of temperature (Temp), salinity (Sal) and chlorophyll-a (Chl a) concentration C in November and 18.7 C in February. A weak thermocline defined as Δt >0.1 C m 1 was found at around m depth in April, whereas a strong thermocline was found at m depth in August and at m depth in November. In February, a weak thermocline was found at m depth in the daytime, whereas the water column was well mixed showing only slight traces of a thermocline at around 190 m depth in the nighttime. Salinity ranged from in the surface layer, and from at 200 m depth (Fig. 2). A remarkable halocline was found in the upper part of the thermocline in August and November due to low salinity (<34.5) above the halocline. In April and February, salinity was high (>34.6) and uniform throughout the m water column. Vertical distribution of chlorophyll-a concentration Chl a concentrations were high in the shallower layers. Layers where the value exceeded 0.2 mg m 3 were shallower than 100 m in April and in August, shallower than 75 m in November, and shallower than 100 m (D) and 125 m (N) in February (Fig. 2). In August, Chl a concentrations peaked (0.41 mg m 3 ) at 100 m depth just below the thermocline, and low concentrations (<0.2 mg m 3 ) were found from the sea surface to 50 m depth. It was <0.1 mg m 3 in the layers below 150 m depth during all seasons except at 200 m depth in February (D). Vertical distribution of appendicularians Abundance and vertical distribution of appendicularians showed an apparent seasonal variation (Fig. 3). Their abundance was higher in August and February than in April and November, with the maximum abundance of ind. m 3 in the surface layer at night in August. They were concentrated in the 0 50 m layer in both daytime and nighttime in April. In August, their abundance was higher at night than in the daytime, and high abundances were observed above 125 m (D) or 150 m depth (N), which was much deeper than the thermocline at m depth. In November, although the abundance was low, they were concentrated in the m (D) or m (N) layer just above the thermocline. In February, the depth limits of high abundance (> ind. m 3 ) were 125 m (D) and 100 m depth (N). Vertical distribution of copepods Abundance of copepods was generally high (> ind. m 3 ) in the surface layer above m depth and was low (< ind. m 3 ) in the layers deeper than 150 m during all seasons (Fig. 3). The highest abundance of ind. m 3 was found in the 0 30 m layer at night in February. The seasonal pattern of the vertical distribution of copepods was similar to that of appendicularians. That is, high abundances of > ind. m 3 were observed in the layers above 50 m in April, 125 m in August, 125 m (D) and 100 m (N) in November, and 150 m (D) and 125 m (N) in February. In these upper layers, Calanoida and Cyclopoida generally accounted for the major portion of the copepod community, whereas oncaeid copepods were >50% in the deeper depths. The seasonal patterns of abundance of oncaeid copepods were largely similar to those of total copepods, but oncaeid copepods were generally more evenly distributed in the m water column than total copepods. The abundance of oncaeid copepods was highest ( ind. m 3 ) in the m depth layer at night in August, while the abundance in the 0 30 m surface layer was higher, especially at night, in April and February than in August and November.

5 Vertical distribution of oncaeid copepods 5 Fig. 3. Vertical distribution of abundance of appendicularians (top) and copepods (lower). Solid lines in the bottom panels indicate proportion (%) of oncaeids with respect to total copepods. Seasonal fluctuation in the number of species and abundance of adult oncaeid copepods A total of 35 species, consisting of one species of Epicalymma, six species and three forms of Oncaea, two species of Spinoncaea, 11 species of Triconia, and 13 species of Oncaea were identified (Table 1). On each sampling occasion, species were collected, of which 24 species were found during all four seasons. The largest number of species (33) was found in August (N), whereas the smallest number (24) occurred in February (N) because of the disappearance of some mesopelagic species such as Oncaea prendeli Shmeleva, 1966 and O. tregoubovi Shmeleva, Abundance of adult oncaeid copepods in the epipelagic zone (0 200 m depth) ranged from ind. m 2, of which the maximum value was found in August (N). A high abundance (> ind. m 2 ) was also found in February (D, N). Abundance was smallest in November (< ind. m 2 in both D and N). There were no remarkable seasonal or diel variations in species compositions in the epipelagic zone. Species that accounted for more than 5% of adult oncaeid copepods at least once in a sampling occasion were Oncaea zernovi (maximum: 48.2%), Spinoncaea ivlevi (26.8%), Oncaea media Giesbrecht, 1891 (16.6%), O. scottodicarloi Heron & Bradford-Grieve, 1995 (12.4%), S. tenuis Böttger-Schnack, 2003 (12.0%), O. vodjanitskii (9.9%), O. mediterranea (Claus, 1863) (9.4%) and Triconia elongata Böttger- Schnack, 1999 (6.4%). Among them, Oncaea zernovi and Spinoncaea ivlevi were the most and the second most abundant species on all sampling occasions, comprising % and %, respectively. The most abundant species after these two species were Oncaea media and O. scottodicarloi, which comprised 16.6% in August (N) and 12.4% in February (N), respectively, when they were most abundant. Community analysis From a total of 64 samples collected in the present study, 63 samples could be allocated to one of three groups, defined as Communities A1, A2 and B, which were separated by a Bray-Curtis dissimilarity index of 0.43 (Fig. 4A). Only the sample from m at night in February did not cluster with any other sample. In the ordination by NMDS, Communities A1 and A2 overlapped slightly with each other, but were distinguishable from Community B, supporting the result of the cluster analysis (Fig. 4B). Community A1, consisting mainly of species of Oncaea (especially O. media), had the lowest number of species compared with the other communities (Table 2). Community A2, which included Spinoncaea ivlevi, Oncaea zernovi with species of Oncaea (mostly O. scottodicarloi, O. media and O. waldemari), had the highest abundance compared with the other communities. Community B was formed by small oncaeid copepods (<0.4 mm in body length) such as S. ivlevi, S. tenuis O. vodjanitskii and O. zernovi with some mesopelagic species (e.g. O. crypta and O. parabathyalis). This community occupied numerically the highest proportion of the whole copepod community in the lower epipelagic layer. The directions of the arrows in the NMDS plot indicate

6 6 H. Itoh et al. Table 1. Abundance (n, ind. m 2 ) and numerical composition (%) of adult oncaeid copepods in the m water column in the Kuroshio Extension. The abundances of small species marked with an * may include those of their congeneric or sibling species (see text). Species Sex Apr Aug Nov Feb Day Night Day Night Day Night Day Night n % n % n % n % n % n % n % n % Epicalymma E. bulbosa Böttger-Schnack, 2009 F, M < Oncaea s. str. O. clevei Früchti, 1923 F, M 68 < < O. media Giesbrecht, 1891 F, M O. mediterranea (Claus, 1863) F, M O. scottodicarloi Heron & Bradford-Grieve, 1995 F, M O. venusta Philippi, 1843 large form F, M < O. venusta Philippi, 1843 medium form F, M O. venusta Philippi, 1843 small form F, M O. waldemari Bersano & Boxshall, 1994 F, M Spinoncaea S. ivlevi (Shmeleva, 1966)* F, M S. tenuis Böttger-Schnack, 2003 F, M Triconia T. conifera (Giesbrecht, 1891) F, M T. denticula (Wi, Shin & Soh, 2011) F < <0.1 T. dentipes (Giesbrecht, 1891) F < < <0.1 T. derivata (Heron & Bradford-Grieve, 1995) F < < < <0.1 T. elongata Böttger-Schnack, 1999 F T. furcula (Farran, 1936) F, M < < < T. giesbrechti Böttger-Schnack, 1999 F 20 < < T. minuta Giesbrecht, 1892 F T. redacta (Heron & Bradford-Grieve, 1995) F 25 < < < <0.1 T. similis (Sars, 1918) F 46 < T. umerus (Böttger-Schnack & Boxshall, 1990) F T. elongata+allied species M T. minuta+allied species M Oncaea O. crypta Böttger-Schnack, 2005 F, M O. longipes Shmeleva, 1968 F, M 13 < <0.1 O. longiseta Shmeleva, 1968 M 39 <0.1 O. minima Shmeleva, 1968 F, M <0.1 O. parabathyalis Böttger-Schnack, 2005 F, M O. parila Heron, 1977 F, M O. platysetosa Boxshall & Böttger, 1987 F, M 39 < < < < < <0.1 O. prendeli Shmeleva, 1966 F, M < <0.1 O. shmelevi Gordejeva, 1972 F, M 13 < < <0.1 O. tregoubovi Shmeleva, 1968 F, M < O. vodjanitskii* F, M O. zernovi* F, M O. sp. 1 F Total number of taxa Total abundance

7 Vertical distribution of oncaeid copepods 7 Fig. 4. Cluster (A) and non-metric multidimensional scaling ordination (B) analyses for grouping of communities (samples) and vertical distributions of each community (C). Sample names in (A) were composed of month (Ap, April; Au, August; N, November; F, February), time (D, daytime; N, nighttime) and depth of sampling layer (m). Arrows in (B) indicate gradients of environmental parameters from low to high. that temperature, Chl a concentration and appendicularian abundance were more positively correlated with Communities A1 and A2, and that depth, salinity, and σ t were more positively correlated with Community B (Fig. 4B). Among these factors, appendicularian abundance differed greatly between the communities, i.e. their abundances in communities A1 and A2 were about 10 times higher than in Community B (Table 2). Community A1 appeared only in April (D) and August (D, N) and was distributed in the uppermost (0 30 or 0 50 m) layer. Community A2 was located above 50 m depth in April and m depths in the other seasons. Community B was always distributed below Community A2. The seasonal pattern of the community distribution was similar to that of the depth of the thermocline (Figs. 2, 4C). Vertical distribution of oncaeid copepods The layer between D 25% and D 75% (henceforth D 25 75% layer ) of the species/forms of Oncaea were located in the thermocline and/or surface mixing zone and they usually overlapped with each other (Fig. 5). However, the interspecific differences of mean D 50% throughout the four seasons were statistically significant between some species/forms, e.g. O. venusta large form was shallower than O. mediterranea, O. media was shallower than O. scottodicarloi and O. venusta small form was shallower than O. meditteranea etc. ( p<0.05). The vertical distribution of each species/ form changed seasonally. The D 50% in April (D, N) and in February (N) was shallower than 50 m in all species/forms of Oncaea, whereas the D 50% in the daytime was deeper than 50 m in November in O. clevei Früchtl, 1923 and O. venusta large form, and in August and November for all the other species/forms. Although the seasonal thermocline shifted downward by ca. 50 m from August to November, D 50% did not show a downward shift parallelling the thermocline, except for O. clevei, which apparently shifted downward in distribution from August to November. For O. waldemari, O. venusta medium form, O. venusta small form, O. media, O. mediterranea and O. scottodicarloi, the diel change of D 50% in each season indicated an upward nocturnal migration as described for oncaeid copepods in previous investigations (e.g. Zalkina 1970,

8 8 H. Itoh et al. Table 2. Mean abundance (ind. m 3 ) of adults of the analyzed species, total oncaeid copepods including immature copepodids, and appendicularians in the three distinct communities. p is significance probability in one-way ANOVA for differences among the three communities (***: p<0.001, **: p<0.01, *: p<0.05). Underlines indicate maximum values and values with no difference (p>0.05) from maximum values accoding to Tukey s post-hoc test. Species Community A1 A2 B Oncaea clevei *** O. media *** O. mediterranea *** O. scottodicarloi *** O. venusta large form *** O. venusta medium form *** O. venusta small form *** O. waldemari *** Spinoncaea ivlevi *** S. tenuis *** Triconia conifera T. denticula *** T. elongata T. furcula * T. giesbrechti *** T. minuta * T. umerus * Oncaea crypta *** O. minima *** O. parabathyalis *** O. tregoubovi * O. vodjanitskii *** O. zernovi ** Number of Species * Abundance of oncaeid copepods (ind. m 3 ) *** Proportion of oncaeid copepods in whole copepod community (%) *** Abundance of appendicularians (ind. m 3 ) *** p 1977, Boxshall 1977, Böttger-Schnack 1990a, b). Of the genus Spinoncaea, the D 25 75% layer was found above 150 m depth in S. ivlevi and below 150 m depth in S. tenuis. The D 25 75% layer of the former species showed a downward shift from April to August and no difference between August and November. It expanded in February and shifted considerably shallower at nighttime. The latter species exhibited no seasonal change in the depth of the D 25 75% layer. The D 25 75% layer of Triconia conifera and T. furcula, belonging to the conifer-subgroup (Böttger-Schnack 1999), was mostly below the thermocline in the daytime, but, contrastingly, was in the thermocline and/or the surface mixed layer in the nighttime. The D 50% of the latter species was deeper than that of the former species (p<0.05) and the D 25 75% layer of the latter species expanded in the daytime. Triconia conifera is well-known as a nocturnal upward migrant (e.g. Zalkina 1970, 1977, Boxshall 1977, Böttger- Schnack 1990a, b). Triconia giesbrechti Böttger-Schnack, 1999 and T. elongata Böttger-Schnack, 1999, belonging to the dentipes-subgroup (Böttger-Schnack 1999), exhibited D 25 75% layers mostly in the surface mixed layer and in the thermocline, respectively. The D 25 75% layer of the latter species shifted downward with the thermocline from April to November and expanded in February. Regarding Triconia minuta Giesbrecht, 1892, T. umerus (Böttger-Schnack & Boxshall, 1990) and T. denticula Wi et al., 2011 belonging to the similis-subgroup (Böttger-Schnack 1999), the D 25 75% layer of the former two species was in or around the thermocline and showed a downward shift with the thermocline from August to November. Triconia denticula was always distributed below the thermocline. Although the D 25 75% layer of the former two species partially overlapped each other, the D 50% of T. minuta was shallower than that of T. umerus (p<0.05). The D 25 75% layers expanded in February and shifted upward especially in the nighttime. As for the species of Oncaea, the D 25 75% layers of O. zernovi and O. vodjanitskii were found in or below the thermocline and shifted downward from April to November followed by an upward shift at nighttime in February. In the remaining four species, O. tregoubovi, O. minima

9 Vertical distribution of oncaeid copepods 9 Fig. 5. Vertical distribution of oncaeid copepods in April (Ap), August (Au), November (N) 1998 and February (F) Solid vertical bars indicate the layer between the depths of the 25th and 75th percentile population (D 25 75%, layer in the text). Open and closed circles indicate depths of 50 th percentile population (D 50% in the text) in the daytime and nighttime, respectively. Broken lines indicate range of thermocline (>0.1 C m 1 ). Shmeleva, 1968, O. parabathyalis Böttger-Schnack, 2005 and O. crypta Böttger-Schnack, 2005, the D 25 75% layer was deeper than in O. zernovi and O. vodjanitskii and exhibited a downward shift or entirely disappeared from the epipelagic zone in February. Relationship between population density and environmental factors in oncaeid copepods Population densities of Oncaea spp. and Triconia giesbrechti were significantly positively correlated (p<0.01) to Chl a concentration and appendicularian abundance and, except for O. scottodicarloi, to water temperature (p<0.05) (Table 3). In contrast, population densities of Spinoncaea tenuis, T. denticula and most species of Oncaea, excluding O. zernovi, were not positively correlated to Chl a concentration and appendicularian abundance, but their densities were significantly positively correlated to depth, salinity and σ t. Body length of females of Oncaea and Triconia Mean body length of females of Oncaea spp. ranged from 0.48 mm (O. waldemari) to 1.19 mm (O. venusta large form) (Table 4). The interspecific differences in their body length were significant (p<0.001) within the genus, except for the differences between O. clevei and O. scottodicarloi and between O. venusta medium form and O. mediterranea (p>0.05). Mean body length of females of Triconia ranged from 0.45 mm (T. giesbrechti) to 1.14 mm (T. conifera). The interspecific differences were significant ( p<0.001), except for the differences between T. gies-

10 10 H. Itoh et al. Table 3. Spearman s rank correlation coefficients between abundance and evironmental parameters for the analyzed species. Significance probability: **, p<0.01; *, p<0.05. species Depth Water temperature Salinity σ t Chlorophyll-a Abundance of appendicularians Oncaea clevei 0.373** 0.567** 0.542** 0.362** 0.600** 0.539** O. media 0.705** 0.527** 0.450** 0.569** 0.653** 0.516** O. mediterranea 0.393** 0.305* * 0.373** 0.433** O. scottodicarloi 0.463** ** 0.680** O. venusta large form 0.396** 0.623** 0.388** 0.342** 0.647** 0.683** O. venusta medium form 0.754** 0.510** 0.443** 0.522** 0.727** 0.588** O. venusta small form 0.678** 0.577** 0.460** 0.572** 0.671** 0.780** O. waldemari 0.476** 0.278* 0.424** 0.307** 0.663** 0.666** Spinoncaea ivlevi * * S. tenuis 0.650** 0.529** 0.430** 0.608** 0.536** 0.594** Triconia conifera ** T. denticula 0.701** 0.374** 0.522** 0.632** 0.466** 0.513** T. elongata ** T. furcula * 0.321** T. giesbrechti * ** 0.628** T. minuta T. umerus 0.271* 0.272* 0.421** 0.306* * Oncaea crypta 0.723** ** 0.606** O. minima 0.784** 0.281* 0.485** 0.518** 0.532** 0.524** O. parabathyalis 0.824** 0.393** 0.459** 0.637** 0.556** 0.552** O. tregoubovi 0.624** ** 0.435** O. vodjanitskii 0.473** ** 0.285* 0.487** 0.322** O. zernovi ** ** * Table 4. Body length of females of Oncaea and Triconia collected from the Kuroshio Extension in August Species n Body Length (mm) Range Average SD Oncaea waldemari O. scottodicarloi O. clevei O. media O. venusta small form O. venusta medium form O. mediterranea O. venusta large form Triconia giesbrechti T. elongata T. minuta T. denticula T. umerus T. furcula T. conifera brechti and T. elongata, and between T. minuta and T. denticula (p>0.05). These two exceptions belong to the same subgroups (dentipes-subgroup and similis-subgroup), respectively. The relationships between D 50% and body length were examined for Oncaea and Triconia using scatter diagrams based on the data in August. The results showed that the variation of D 50% among the species/forms of Oncaea, excluding O. venusta large form, was greater in the nighttime than in the daytime (Fig. 6A). No correlation was found between D 50% and body length in these species/forms. However, it is notable that two pairs of species/forms hav-

11 Vertical distribution of oncaeid copepods 11 Fig. 6. Scatter diagrams of body length vs. depth of 50th percentile population of Oncaea (A) and Triconia (B) in August, Open circles and cross symbols represent daytime and nighttime samples, respectively. Abbreviations of the species/form names are the same as in Fig. 5. ing similar body lengths, viz. O. clevei and O. scottodicarloi of ca. 0.6 mm and O. venusta medium form and O. mediterranea of ca. 1.0 mm, had very different nighttime D 50% s, and that three pairs of species/forms with similar nighttime D 50% s, viz. O. media and O. venusta medium form with ca. 40 m D 50%, O. waldemari and O. venusta small form with ca. 60 m D 50%, and O. scottodicarloi and O. mediterranea with ca. 75 m D 50%, differed considerably in body length within each pair. The differences in body length between species/forms of these pairs were 1.4 times between O. media and O. venusta medium form, 1.8 times between O. waldemari and O. venusta small form, and 1.6 times between O. scottodicarloi and O. mediterranea. Although some pairs such as O. waldemari and O. scottodicarloi had similar body lengths and similar D 50% s, the scatter diagram of Oncaea indicates the trend that species/ forms of Oncaea with similar sizes were distributed in different layers at night and those distributed in the same layer at night had different sizes from each other. Regarding species of Triconia, body lengths were distinctly different among the subgroups but similar within a subgroup (Fig. 6B). The diel migration patterns were different among subgroups. Although D 50% s were distinctly different among species belonging to the same subgroup, the relationship between body length and D 50% that was observed for Oncaea was not apparent. Discussion Although the vertical distribution of oncaeid copepods has been investigated in various places around the world studies focusing on their fine scale distribution in the epipelagic zone have been very limited. Nishibe et al. (2009) investigated the vertical distribution of oncaeid copepods in the epipelagic and upper mesopelagic zone in Tosa Bay on the Pacific coast of Japan in August and November using a fine (63 μm) mesh net and reported 45 species and three forms. Although the number of species in the present study was only 35 because of combining some small species during the analysis and limitation of the sampling layer to the epipelagic zone, 33 of the species were common to Tosa Bay. Of the nine dominant species in Tosa Bay (Nishibe et al. 2009), Oncaea media, O. scottodicarloi, Spinoncaea ivlevi, S. tenuis and Oncaea zernovi were also the dominant oncaeid copepods in the Kuroshio Extension. The three communities (A1, A2 and B) detected in the present study were largely the same as the three communities reported in Tosa Bay, i.e. the upper epipelagic community predominated by O. venusta, O. media and O. waldemari, the lower epipelagic community dominated by O. scottodicarloi, O. zernovi, O. bispinosa, Spinoncaea ivlevi and S. tenuis and the mesopelagic community dominated by O. crypta, O. parabathyalis, O. minima and O. tregoubovi etc. The coincidence of oncaeid communities in the Kuroshio Extension region and upstream in Tosa Bay suggests that an oncaeid community structure characterized by three communities inhabiting different depth ranges is widespread in the epipelagic zone of temperate waters along the Kuroshio Current. The present study revealed that there was a positive correlation between the population density and temperature in most species of Oncaea, which were the main component of Communities A1 and A2, and a negative correlation in Spinoncaea tenuis, Triconia denticula and three species of Oncaea, which were the main components characterizing Community B. This indicates that the vertical distribution of temperature is an important factor related to their vertical distribution. However, most Oncaea stayed in the surface mixed layer, despite the downward shift of the thermocline from August to November. Utilization of appendicularian houses as a habitat and food by oncaeid copepods has been demonstrated by previous studies (Alldredge 1972, Ohtsuka & Kubo 1991, Ohtsuka et al. 1993, Steinberg et al. 1994, Ohtsuka et al. 1996). Enumeration of appendicularians in the present study was not based on their houses but rather on their bodies. However, it is natu-

12 12 H. Itoh et al. rally considered that the abundance of appendicularian houses would be in high in waters where the abundance of their bodies is high. Nishibe et al. (2009) observed that in Tosa Bay the highest abundances of oncaeid copepods occurred at the depths where appendicularians were abundant. They suggested that oncaeid copepods depend strongly on appendicularian houses. In the present study, highly significant positive correlations were observed particularly between the population densities of Oncaea, the main component of Community A2, and the abundance of appendicularians. This result mirrors that of Nishibe et al. (2009) in Tosa Bay, and suggests that dependence on appendicularian houses is strong, especially in Oncaea. In the present study, six species and three forms of Oncaea and 11 species of Triconia were recorded in the epipelagic zone of the Kuroshio Extension. Such coexistence of many congeneric species of oceanic copepods has also been reported in the genera Euaugaptilus (Matsuura et al. 2010) and Clausocalanus (Peralba & Mazzocchi 2004), the families Scolecitrichidae (Kuriyama & Nishida 2006) and Oithonidae (Nishida & Marumo 1982), etc. Mechanisms allowing their coexistence in oceanic waters have been proposed to involve alleviation of interspecific competition by partition of niches represented by differences in distribution layer, body size, morphs of mouthparts, etc. Some mechanisms for niche partitioning are considered with respect to the coexistence of species/forms of Oncaea in the present study. The relationship between their body length and D 50% s in August showed the trend that in the nighttime, species/forms with similar sizes were distributed in different layers and those in the same layer had different sizes. According to Hutchinson (1959), a difference in the size of an animal by 1.3 times respective body lengths allows alleviation of competition between two species in the same guild. For the present case of Oncaea, the species/ forms with similar D 50% s had body lengths times co-occurring species suggesting that competition between species/forms of each pair could be alleviated by size-dependent niche partitioning, perhaps signifying different feeding habits. The present result that similar-sized species/forms of Oncaea, such as O. clevei and O. scottodicarloi, had very different D 50% s from each other indicates another mechanism of coexistence, i.e. depth partitioning. Thus it is likely for Oncaea that the coexistence of species/ forms in the epipelagic zone is achieved by a combination of body size and depth partitioning. On the other hand, the relationships between body size and D 50% were not as apparent in Triconia. However, interspecific differences of D 50% were significant within the subgroup in dentipes- and similis-subgroups, neither of which have remarkable diel migration. In these subgroups, the coexistence of congeners might be explained by niche partitioning related to differences in distribution depth. The present study revealed that more than 35 species of oncaeid copepods, including mesopelagic species and unidentified species, inhabited the epipelagic zone in the Kuroshio Extension and the D 25 75% of the epipelagic species were shallow in spring, shifted downward in summer or autumn, and expanded (or shifted upward) in winter. The seasonal change in vertical distribution resulted in fluctuations in their seasonal abundance in the surface layer shallower than 50 m, where the abundance of oncaeid copepods was higher in spring and winter than in summer and autumn. Since larvae and juveniles of Japanese sardine, Japanese anchovy and Pacific saury inhabit mainly the 0 50 m depth layer in spring (Tsukamoto et al. 2001, Yatsu et al. 2005), the abundance of small copepods, including oncaeid copepods, in that layer is critically important for them as prey. Yamazi (1957) reported a nocturnally dense concentration of Oncaea spp. at the sea surface in Wakayama Harbor on the Pacific coast of Japan. Extremely high abundance (ca. 150 ind. m 3 ) and high dominancy (ca. 40% of total copepods) of Oncaea venusta were observed in the surface layer (0 0.5 m) of Suruga Bay on the Pacific coast of Japan in September 1978 (Itoh unpublished). These aggregations of oncaeid copepods, as well as the seasonal pattern of their abundance in the surface layer where they were abundant especially at night in winter and spring, are probably important as food for fish larvae and juveniles in the surface layer, especially for Pacific saury, which have a neustonic ecology with nighttime feeding (Tsukamoto et al. 2001, Odate 1977) and occasionally have high abundances of oncaeid copepods in their stomach contents (Odate 1977). Acknowledgements We thank Dr R. Böttger-Schnack for critical comments on an earlier draft of this paper. We also thank the editor and two anonymous referees for constructive comments on the manuscript. We are grateful to the captain and crew of the R.V. Soyo-Maru, National Research Institute of Fisheries Science, for their help in sampling at sea. This work was supported by the Pacific Ocean Fishing resources (VENFISH) and study for the prediction and control of population outbreaks of marine life in relation to environmental change (POML) program from the Ministry of Agriculture, Forestry, and Fisheries, Japan. References Alldredge AL (1972) Abandoned larvacean houses: A unique food source in the pelagic environment. Science 177: Böttger R (1987) The vertical distribution of micro- and small mesozooplankton in the central Red Sea. Biol Oceanogr 4: Böttger-Schnack R (1988) Observations on the taxonomic composition and vertical distribution of cyclopoid copepods in the central Red Sea. Hydrobiologia 167/168: Böttger-Schnack R (1990a) Community structure and vertical distribution of cyclopoid copepods in the Red Sea. I. Central Red Sea, autumn Mar Biol 106:

13 Vertical distribution of oncaeid copepods 13 Böttger-Schnack R (1990b) Community structure and vertical distribution of cyclopoid copepods in the Red Sea. II. Aspects of seasonal and regional differences. Mar Biol 106: Böttger-Schnack R (1992) Community structure and vertical distribution of cyclopoid and poecilostomatoid copepods in the Red Sea. III. Re-evaluation for separating a new species of Oncaea. Mar Ecol Prog Ser 80: Böttger-Schnack R (1994) The microcopepod fauna in the Eastern Mediterranean and Arabian Seas: a comparison with the Red Sea fauna. Hydrobiologia 292/293: Böttger-Schnack R (1995) Summer distribution of micro- and small meso-zooplankton in the Red Sea and Gulf of Aden, with special reference to non-calanoid copepods. Mar Ecol Prog Ser 118: Böttger-Schnack R (1996) Vertical structure of small metazoan plankton, especially non-calanoid copepods. I. Deep Arabian Sea. J Plankton Res 18: Böttger-Schnack R (1997) Vertical structure of small metazoan plankton, especially non-calanoid copepods. II. Deep Eastern Mediterranean (Levantine Sea). Oceanol Acta 20: Böttger-Schnack R (1999) Taxonomy of Oncaeidae (Copepoda, Poecilostomatoida) from the Red Sea. I. 11 species of Triconia gen. nov. and a redescription of T. similis (Sars) from Norwegian waters. Mitt hamb zool Mus Inst 96: Böttger-Schnack R (2001) Taxonomy of Oncaeidae (Copepoda, Poecilostomatoida) from the Red Sea. II. Seven species of Oncaea s. str. Bull nat Hist Mus Lond (Zool) 67: Böttger-Schnack R (2002) Taxonomy of Oncaeidae (Copepoda, Poecilostomatoida) from the Red Sea. VI. Morphology and zoogeography of Oncaea bispinosa sp. nov., a sister taxon of O. zernovi Shmeleva. J Plankton Res 24: Böttger-Schnack R (2003) Taxonomy of Oncaeidae (Copepoda, Poecilostomatoida) from the Red Sea. V. Three species of Spinoncaea gen. nov. (ivlevi-group), with notes on zoogeographical distribution. Zool J Linn Soc 137: Böttger-Schnack R (2005) Taxonomy of Oncaeidae (Copepoda, Poecilostomatoida) from the Red Sea. VII. Oncaea cristata, a new species related to the ovalis-complex, and a revision of O. ovalis Shmeleva and O. bathyalis Shmeleva from the Mediterranean. Cah Biol Mar 46: Böttger-Schnack R, Boxshall G A (1990) Two new Oncaea species (Copepoda: Poecilostomatoida) from the Red Sea. J Plankton Res 12: Böttger-Schnack R, Hagen W, Schnack-Schiel S B (2001) The microcopepod fauna in the Gulf of Aqaba, northern Red Sea: species diversity and distribution of Oncaeidae (Poecilostomatoida). J Plankton Res 23: Böttger-Schnack R, Huys R (1998) Species groups within the genus Oncaea Philippi, 1843 (Copepoda: Poecilostomatoida). J Mar Syst 15: Böttger-Schnack R, Huys R (2004) Size polymorphism in Oncaea venusta Philippi, 1843 and the validity of O. frosti Heron, 2002: a commentary. Hydrobiologia 513: 1 5. Böttger-Schnack R, Lenz J, Weikert H (2004) Are taxonomic details of relevance to ecologists? An example from oncaeid microcopepods of the Red Sea. Mar Biol 144: Böttger-Schnack R, Schnack D (2013) Definition of species groups of Oncaeidae (Copepoda: Cyclopoida) as basis for a worldwide identification key. J Nat Hist 47: Böttger-Schnack R, Schnack D, Hagen W (2008) Microcopepod community structure in the Gulf of Aqaba and northern Red Sea, with special reference to Oncaeidae. J Plankton Res 30: Boxshall GA (1977) The depth distributions and community organization of the planktonic cyclopoids (Crustacea: Copepoda) of the Cape Verde Islands region. J Mar Biol Ass U K 57: Boxshall GA, Böttger R (1987) Two new species of Oncaea (Copepoda: Poecilostomatoida) from the Red Sea and a redescription of O. atlantica Shmeleva. J Plankton Res 9: Boxshall GA, Halsey SH (2004) An introduction to copepod diversity. The Ray Society, London. 966 pp. Elvers D, Böttger-Schnack R, Blohm D, Hagen W (2006) Sympatric size variants of the microcopepod Oncaea venusta exhibit distinct lineages in DNA sequences. Mar Biol 149: Furuhashi K (1966) Studies on the vertical distribution of copepods in the Oyashio region east of Japan and in the Kuroshio region south of Japan. Publ Seto Mar Biol Lab 14: Gordeyeva KT (1972) New species of the genus Oncaea (Copepoda, Cyclopoida) from the tropical zone of the Atlantic Ocean. Zool Zh 51: (in Russian with English abstract) Hayakari S (2001) Statistical analysis add-in Cluster Analysis Ver Available at: shfirst.htm (accessed on 20 Dec. 2001) Heron GA (1977) Twenty-six species of Oncaeidae (Copepoda: Cyclopoida) from the southwest Pacific Antarctic Area. In: Biology of the Antarctic Seas VI, (ed. Pawson DL). American Geophysical Union, Washington, pp Heron GA, Bradford-Grieve JM (1995) The marine fauna of New Zealand: Pelagic Copepoda: Poecilostomatoida: Oncaeidae. NZOI Mem 104: Hutchinson GE (1959) Homage to Santa Roselia or why are there so many kinds of animals? Am Nat 93: Huys R, Böttger-Schnack R (2007) Taxonomy of Oncaeidae (Copepoda, Cyclopoida) from the Red Sea. VIII. Morphology and phylogenetic position of Oncaea tregoubovi Sehmeleva, 1968 and the closely related O. prendeli Shmeleva, 1966 from the Mediterranean Sea. Mitt hamb zoo Mus Inst 104: Krŝinić F (1998) Vertical distribution of protozoan and microcopepod communities in the South Adriatic Pit. J Plankton Res 20: Krŝinić F, Malt S J (1985) Little known species of small Oncaeidae (Cyclopoida) from the South Adriatic. J Plankton Res 7: Kruskal JB, Wish M (1978) Multidimensional Scaling: Quantitative applications in the social sciences 11. Sage publications, London. 93 pp. Kuriyama M, Nishida S (2006) Species diversity and niche-partitioning in the pelagic copepods of the family Scolecitrichidae (Calanoida). Crustaceana 79: Matsuura H, Nishida S, Nishikawa J (2010) Species diversity and vertical distribution of the deep-sea copepods of the genus Euaugaptilus in the Sulu and Celebes Seas. Deep-Sea Res Part II 57: Nishibe Y, Ikeda T (2004) Vertical distribution, abundance and