ABSTRACT: The present study dealt with seasonal and interannual variations in the abundance and biomass, and spatio-temporal distributions of the portunid crab Charybdis bimaculata dominant in Ise Bay, central Japan. The abundance and biomass of the crab decreased in summer when the oxygen-poor water developed in central or inner parts of the bay, and then increased through new recruits from autumn (October–November) to the following spring (March–May) when the oxygen-poor water disappeared. Berried females were collected mainly from spring to autumn. Recruits were collected in any season. Particularly in winter, most recruits were located in the innermost part of the bay. According to the cohort separation based on size frequency distribution in carapace width of the crab specimens, the cohorts that were derived from spawning in spring to summer largely contributed to establishing and maintaining the benthic populations for the following year in the bay, whereas those from other seasons failed to recruit because of serious damage caused by the oxygen-poor water. Most crab individuals one year post hatch contributed to spawning and then died by the winter of the same year.
The portunid crab Charybdis bimaculata is commonly found and dominant among benthic crustaceans in Japanese coastal waters such as Tokyo Bay,1,2 Ise Bay,3–6 Osaka Bay,7,8 Yuya Bay,9 Beppu Bay10 and the Seto Inland Sea.11 Although many individuals of the crab are collected by small trawl nets in these coastal waters, little ecologic information is known to date because the commercial value of the crab is very low for trawl fisheries. The crab has an important role for the production of marine fishery resources because it is an important food organism for the frog flounder Pleuronichthys cornutus12 and the conger eel Conger myriaster13 in the above coastal waters. It is important to know the population dynamics of this crab for successful use of fishery resources in order to clarify the megabenthos community structure in Ise Bay, central Japan. Based on short-term observations, seasonal variations in the abundance and biomass, spawning season, sex ratio2,11 and food habits of the crab14 have been reported, whereas those based on long-term observations have been reported only by Hossain5 and Narita et al.6 According to these studies,5,6 the abundance and biomass of the crab decreased in summer when oxygen-poor water developed in the central or inner parts of the bay, and then increased from autumn to the following spring when the oxygen-poor water disappeared. To clarify the population dynamics of the crab in the bay, it is necessary to examine its migration or larval recruitment events in relation to the development/disappearance of the oxygen-poor water in the bay, based on detailed observations through the life cycle of the crab.
The aims of the present study were to clarify the population dynamics of the portunid crab Charybdis bimaculata in Ise Bay, based on long-term observation during the 10 years from 1993 to 2002. Specific aims were: (i) to examine seasonal and interannual variations in the abundance and biomass, the occurrence of berried females, and spatio-temporal distributions of the crab; (ii) to separate different cohorts for examining the fates of these cohorts in relation to the development/disappearance of the oxygen-poor water in the bay; and (iii) to make clear when and where new recruits are located within the bay.
MATERIALS AND METHODS
Ise Bay is a semi-enclosed bay with a surface area of 1738 km2 and a mean depth of 19.5 m, located along the Pacific coast of central Japan (Fig. 1). The bay water is exchanged with the open coastal waters through the Irago Channel (width 13 km, depth 73 m). The Kiso Rivers (Kiso, Nagara and Ibi rivers), three of the largest rivers in Japan, flow into the innermost part of the bay and influence the oceanographic conditions in the bay.15,16 Sediment in the northern and central parts of the bay is silt–clay, while it is muddy sand or sandy gravel15 in the entrance and south-eastern parts (Fig. 1).
Recently, the bay water has become rich in nutrients and highly turbid from freshwater discharge and sewage effluent from cities situated on the western and northern coasts of the bay. With the recent progress of eutrophication in the bay, red tides have often been observed every summer and autumn, which is coupled with the oxygen-poor water widely developed in the bay.16
Sampling procedures and data processing
We analyzed variations in environmental factors (i.e. water temperature, salinity and dissolved oxygen content in the bottom waters) of Ise Bay, based on data from the Mie Prefectural Institute of Fisheries Technology,17 in which observations were carried out at 14 stations within the bay (Fig. 1). In the present study, ‘oxygen-poor water’ is defined as when the oxygen content is less than 3 p.p.m. in waters 1 m above the bottom, following our previous studies.5
Crab samples were collected at 18 stations in the bay on board the T/V Seisui Maru of Mie University, three or four times every year from April 1993 to December 2002. Samples were collected at 15 stations in September 1993, 17 stations in April 1993 and June 1999, and 19 stations in July 1993, April 1994 and September 1994. At three stations (stns 17′, 18 and 35), samples were collected once during the study period. We defined the area including stns 2′−8 as the northern part, the area including stns 9–17 as the central part, and the area including stations 18–34 as the southern part (Fig. 1).
To collect juvenile and adult crab samples, large gears were towed for 900 m along the bottom, usually for 15 min at a ship speed of 2 knots (1 m/s). Each tow covered a bottom area of approximately 1500 m2 per station. However, large gears with a 2.7-cm mesh opening3,5 were not able to collect individuals with a carapace width less than 12.7 mm. In the present study, individuals with a carapace width more than 12.7 mm were defined as juveniles or adults, while those with less than 12.7 mm were defined as recruits. On the other hand, in order to collect new recruits, we used two types of gear (small gears a and b) with 0.6- and 1.0-cm mesh openings, respectively. The samples of recruits were collected at 15 stations (except for stations 32, 33 and 34 where there was sand or sandy gravel sediment) from May 2000 to November 2001 using the small gear a, and from January 2002 to December 2002 using the small gear b. Mesh openings of the small net gears were too small to avoid net damage by bottom gravels at the above three stations, so the gear was towed approximately 180 m along the bottom usually for 3 min at a ship speed of 2 knots. Each tow covered a bottom area of approximately 200 m2 per station.
Megabenthos samples from each tow were fixed immediately with 20% formalin–sea water. In the laboratory, C. bimaculata specimens were sorted from these samples to measure wet weights and numbers of specimens. The abundance and biomass of the crab were indicated as individual number (inds/stn) and wet weight (kg/stn), respectively. Spearman's correlation coefficient was calculated for examining the relationship between environmental factors and the abundance and biomass of the crab. To examine seasonal and interannual variations in the mean abundance (inds/stn) of the crab, all samples were divided into the following five periods: spring (March–May), early summer (June–July), late summer (September), autumn (October–November) and winter (January–December). Differences in the abundance and biomass of the crab were examined between years and between seasons using a Kruskal–Wallis test. In cases where the Kruskal–Wallis test indicated significant differences, their means were compared using the Steel–Dwass test.
Then, differences in spatial (except for stns 17′ and 18, which were observed only in one sampling) and temporal (sampling times) distributions of the crab abundances were examined using a Freidman test. To examine relationships between bottom sediment types and the abundance of juvenile and adult crabs, bottom sediments at sampling stations were divided into two types (stns 2′, 6–16 with silt–clay sediments; and stns 5, 17, 17′ 30, 32–34 and 35 with sand and coarse to fine sand sediments) and their average abundance calculated for each sediment type was compared using a Wilcoxon signed-ranks test.
Carapace widths of the crab specimens were measured to the nearest 0.1 mm using a digital caliper after classifying sex. Size frequency distributions in carapace width for females and males were made. Based on the samples collected from April 1993 to December 2002, different cohorts of the crab were separated by applying Akamine's method,18 which separates a polymodal length distribution into two or more normal distributions calculated by using MS-Excel (Microsoft, Redmond, WA, USA) as referred to Aizawa and Takiguchi.19
Seasonal and interannual variations in environmental factors
Variations in environmental factors in bottom waters of Ise Bay are shown in Figure 2. Water temperature showed a clear annual cycle from 1993 to 2002, reaching the highest in September or October and the lowest in February or March every year. No consistent pattern in the annual cycle of salinity was detected. The lowest salinity was more than 30 except for June to July 1993, May 1994 and April 1995 with salinity less than 27. The dissolved oxygen content showed a clear annual cycle for the same years, reaching the highest in February or March every year while the lowest was recorded in August or September. A negative correlation was detected between water temperature and the dissolved oxygen content (Spearman's correlation coefficient by rank test, rs = −0.70, P < 0.0001). The oxygen-poor water usually developed from summer (June or July) to autumn (September–November) every year. Monthly mean dissolved oxygen contents showed remarkable interannual variations. The mean dissolved oxygen contents were not less than 3 p.p.m. in 1995, while the other were 2 p.p.m. or less for several months in summer.
Seasonal and interannual variations in abundance and biomass of juveniles and adults of Charybdis bimaculata
Seasonal and interannual variations in the abundance and biomass of the crab are shown in Figure 2. The crab abundance decreased in summer when the oxygen-poor water developed in the central or inner parts of the bay, and then increased from autumn to the following spring when the oxygen-poor water disappeared. There was a significant difference in the crab abundance between seasons (Kruskal–Wallis test, H = 16.85, P < 0.01). Higher abundance was detected in summer (June and July), and lower in late summer (September) and/or autumn (Steel–Dwass test, P < 0.01). There was no significant difference in the crab abundances between years (Kruskal–Wallis test, H = 12.05, P = 0.21). However, the highest mean abundance through the study period was 56.13 inds/stn in 2002 and the lowest was 2.65 inds/stn in 1997. There was no significant correlation between the crab abundance and monthly means of environmental factors for each survey (Spearman's correlation coefficient by rank test, P > 0.05).
The crab biomass usually decreased from summer (June and July) or late summer (September) toward autumn and then increased from winter toward the following spring. There was a significant difference in the crab biomass between seasons (Kruskal–Wallis test, H = 17.55, P < 0.01). Higher biomass was detected in summer and lower in late summer and/or autumn (Steel–Dwass test, P < 0.01). There was no significant difference in the crab biomasses between years (Kruskal–Wallis test, H = 11.12, P = 0.27). However, the highest mean biomass through the study period was 160.08 g/stn in 2002 and lowest was 8.69 g/stn in 1997. There was no significant correlation between the crab biomass and monthly means of environmental factors for each survey (Spearman's correlation coefficient by rank test, P > 0.05).
Spatio-temporal distributions of juveniles and adults of Charybdis bimaculata
Figure 3 shows spatio-temporal distributions of the juveniles and adults of the crab, based on 41 samplings from April 1993 to December 2002. Their abundance was significantly different between stations (Friedman test, χ2 = 68.41, P < 0.0001) and also between sampling times (Friedman test, χ2 = 345.6, P < 0.0001). As indicated in Figure 2, the highest average abundance of 139.0 inds/stn was recorded in June 1999 while the lowest was 0.0 inds/stn in November 1996 and October 1998. However, the highest average abundance through the study period was observed at stn 11 (51.5 inds), but the lowest one was observed at stn 9 (15.2 inds).
High densities of the juvenile and adult crabs were mainly found in the northern and central parts of the bay where bottom sediment is occupied by silt–clay. The mean densities of the juveniles and adults were significantly higher in silt–clay sediments (stns 2′, 6–16) (mean ± standard deviation [SD] 32.27 ± 45.53) than at sand and coarse to fine sand sediments (stns 5, 17, 17′, 18, 30, 32–34 and 35) (mean ± SD 19.74 ± 28.89) (Wilcoxon signed-ranks test, P < 0.0001).
Spatio-temporal distributions of Charybdis bimaculata recruits
Figure 4 shows spatio-temporal distributions of the recruits of the crab, based on 13 samplings from May 2000 to December 2002. In 2000, few recruits were only collected in autumn (October). In 2001, a few recruits were collected only at local sites in winter (January) to spring (May) and autumn (November). In 2002, recruits appeared in summer (June) and in winter (January and December). In January 2002, the highest density of recruits through the study period was found in the central part where bottom sediment is occupied by silt–clay.
Size frequency distribution in carapace width and cohort separation of Charybdis bimaculata population
Carapace width distribution was divided into one or two modes of normal distribution by applying Akamine's method (Fig. 5). Each of the female and male populations were composed of one size group from spring to summer, and then individuals with a mean carapace width 15–20 mm appeared after autumn. From the following spring to summer , individuals of the above size group grew to approximately 30 mm in mean carapace width. Most females were berried from spring to autumn. These individuals decreased in number and/or disappeared in autumn when the oxygen-poor water disappeared.
Thus, the growth curves for each of females and males were identified as a total of 11 cohorts (A–K) (Fig. 6). The periods of occurrence of these cohorts for both females and males were approximately 1 year. Cohort E of both females and males disappeared once in summer when the oxygen-poor water developed, and was detected again in autumn when the oxygen-poor water disappeared. Maximum mean carapace width of the crab was 31.3 mm for females and 33.6 mm for males.
Recruits with mean carapace width approximately 7.0 mm for females and 5.0 mm for males in October 2000 grew to individuals with mean carapace width 7.7 mm for females and 11.3 mm for males in January 2002. However, these cohorts were not collected in summer in 2002, indicating juvenile and adult populations were not successfully established. In December 2002, recruits which may be derived from spawning in autumn were collected with a mean carapace width approximately 8.5 mm for females and 8.2 mm for males.
In Ise Bay, C. bimaculata was mainly found in the northern and central parts of the bay where bottom sediment is occupied by silt–clay or sand and coarse to fine sand, although a higher density of the crab was detected in silt–clay sediments (Fig. 3). This is also true for Tokyo Bay,1,2 suggesting that the crab prefers to inhabit silt–clay sediments.
In the present study, there was no significant correlation between the mean of abundance and biomass of C. bimaculata and the mean of environmental factors (water temperature, salinity and dissolved oxygen content). There may be too large a bias on abundance/biomass of crab due to remarkable seasonal variation for each year. The occurrence of the oxygen-poor water is thought to be the most important factor affecting crab abundance and biomass. The abundance and biomass of the crabs decreased from summer to autumn, corresponding to the period of the oxygen-poor water events, and increased from winter to the following spring, corresponding to the periodof increased dissolved oxygen content (Fig. 2). Further, higher densities of the crab were not found in areas deeper than 20 m depth in the central part of the bay in summer where the oxygen-poor water developed (Fig. 3). These findings were revealed by the present study as well as in previous studies dealing with other dominant megabenthos species in Ise Bay.4,5,20–22 The degree of suffering caused by the oxygen-poor water differs depending on benthic species with different tolerances to the lower oxygen concentration and also with different mobilities.23 Mass mortality or localized spatial distributions have often been reported for immobile benthic organisms from oxygen-poor water in coastal regions of the world.3–8,12,24–28 However, it is well known that mobile benthic organisms, particularly some species of Pisces and crustaceans, can escape from the oxygen-poor water by migration.25,26,28–34
According to Bell et al.32 the blue crab Callinectes sapidus exhibited only weak orientation toward the areas with higher dissolved oxygen content, sometimes being within oxygen-poor areas (2 mg/L as 2 p.p.m. of dissolved oxygen content) for several hours. Bell and Eggleston33 also reported that two species of paralichthid flounder Paralichthys dentatus and P. lethosigma gradually decreased in abundance in oxygen-poor water of 2–4 mg/L, whereas the abundance of the above crab dropped sharply at 2 mg/L. Therefore, they concluded that the blue crab is more likely to be in oxygen-poor water than paralichthid flounder. According Ariyama et al.8 many individuals of C. bimaculata died out in Osaka Bay in summer because of lower mobility and tolerance to the oxygen-poor water. In the present study, the abundance of both females and males of the crab decreased after the oxygen-poor water disappeared, simultaneously with the death of most cohorts (Figs 5 and 6). Although the higher densities of the crab were found in the central part of the bay (Fig. 3), many individuals may have died out because of lower mobility, lower tolerance to the oxygen-poor water, and weak orientation toward the areas with the higher dissolved oxygen content, as revealed for the blue crab.
Berried females of C. bimaculata were collected for most months except for January and December (winter) in Ise Bay (the present study), although they were collected in December to the following February in Tokyo Bay,2 and in most months except for January and February in Seto Inland Sea.11 According to the above results, the abundance peak of berried females was detected from summer to autumn, similar to previous studies, and many new recruits were collected from autumn to the following spring (April), particularly in January, April, and December 2002, indicating planktonic larvae of the crab constantly settled in the bay through its long spawning period. Judging from carapace width histograms of the crab and the cohort separation (Figs 3 and 6), the recruits with carapace widths less than 12.7 mm, which appeared from winter to the following spring, failed to recruit into the juvenile and adult populations because growing individuals of these cohorts were not collected in summer when the water temperature was higher and the growth rates of crabs would be increased. On the other hand, individuals with carapace widths 15–25 mm, which appeared from autumn to the following spring, and were derived from spawning from spring to summer, were successful in recruiting to establishing their benthic populations. These findings suggest that C. bimaculata populations in the bay were established and maintained by the recruits derived from spawning in spring to summer during the oxygen-poor water development. The period of maximum occurrence of these cohorts was approximately 1 year for cohort A, with 10 months for females and 13 months for males. Most cohorts disappeared from summer to autumn (October) 1 year after their larvae settled, or from winter (January) to spring (April) 2 years after their larvae settled (Fig. 6). In Tokyo Bay,2 the abundance of 1+ year individuals decreased from late summer (September) or autumn (November) to winter (December–February), and alternatively the 0+ year individuals smaller than the above individuals increased. In the Seto Inland Sea,11 many two-year-old individuals died in winter (February) after larval releasing by spawning. Thus, the life span of the crab is estimated to be 18 months in these waters, assuming larvae for most individuals were released in September when most females with eggs were collected. In Ise Bay, the duration of cohorts was approximately same as in Tokyo Bay, but shorter than in Seto Inland Sea, which indicates that the disappearance of cohorts due to the oxygen-poor water in summer occurred more easily in Ise Bay than in the Seto Inland Sea.
In Ise Bay, population dynamics of two dominant megabenthos species, the dragonet Repomucenus valenciennei and mantis shrimp Oratosquilla oratoria, were affected by the oxygen-poor water.20,21 According to these studies, populations of the dragonet in the bay were established and maintained by individuals able to avoid the oxygen-poor water and also by the new recruits derived from larvae released by spawning in autumn. Two cohorts of recruits were detected for the mantis shrimp, i.e. one in summer and the other in autumn. Usually, populations of the mantis shrimp were largely maintained by summer cohorts of recruits. In the present study, the oxygen-poor water in summer 1995 developed in the narrow central part with the mean dissolved oxygen contents higher than 3 p.p.m. in the bay, whereas the oxygen-poor water (<2 p.p.m.) in the other years developed widely in the central part and continued for several months (Fig. 2).17 Therefore, many individuals of R. valenciennei and O. oratoria were able to avoid the oxygen-poor water in 1995 compared to other years. On the other hand, the abundance of C. bimaculata decreased from summer (June) to autumn (November) in 1995 as well as in the other years, and increased from winter to the following spring (March). As mentioned above, many crab individuals were not able to avoid the oxygen-poor water probably because of weak mobility inferior to R. valenciennei and O. oratoria. The surviving crabs died out in winter because of the shorter life duration of their cohorts than R. valenciennei and O. oratoria, so that the abundance of crab recruits largely increased from winter to the following spring.
Recruits of C. bimaculata with carapace width less than 12.7 mm appeared from winter (January and December) to the following spring (April and May) were not successful in establishing juvenile and adult populations. These recruits were derived from spawning in autumn when the oxygen-poor water disappeared, since planktonic larvae of many crustaceans are estimated to settle after the planktonic stage of 1 month. Then, those individuals were not able to grow beyond a carapace width of 12.7 mm from winter to the following spring, and died out in summer because of the oxygen-poor water development. On the other hand, recruits grew beyond a carapace width of 12.7 mm from winter to the following spring, and were successful in establishing juvenile and adult populations. These cohorts were derived from spawning in summer and also from their larvae settled in late summer to autumn in a shallow area in the bay, because a few recruits were collected at stations deeper than 10 m depth. According to Tankersley and Wieber,34 tolerances to hypoxia and anoxia (from the oxygen-poor water) of the blue crab Callinectes sapidus juveniles (as recruits) were inferior to megalopa larvae when the oxygen content was less than 20%. Juveniles died out faster than megalopa larvae when the oxygen content was 0%, and megalopa larvae may avoid the oxygen-poor water then metamorphose to juveniles when they visit optimal sedimentary areas because metamorphosis may be delayed when the oxygen content is 40–60%. In Ise Bay, planktonic larvae of C. bimaculata released in summer could not survive because of the oxygen-poor water when they settled in areas deeper than 10 m. The juvenile and adult populations may be established by recruits that settled in areas shallower than 10 m depth where severe oxygen-poor water did not develop, or by recruits that migrated from the open coastal waters through the Irago Channel. In future studies, we must make clear whether recruits appeared in water shallower than 10 m depth, and how the recruits that migrated and settled in the bay contribute to the population of this species in Ise Bay.
We express sincere thanks to colleagues, particularly Mr T Amakawa, Marine Ecology Laboratory of Mie University, for encouragement and help with sampling throughout the study. Thanks to the captain and crew of the T/V Seisui Maru of Mie University for facilitating sampling on board the vessel.