Journal of Oceanography, Vol. 62, pp. 551 to 558, 2006




   Ecosystem Shift Resulting from Loss of Eelgrass and
   Other Submerged Aquatic Vegetation in Two Estuarine
   Lagoons, Lake Nakaumi and Lake Shinji, Japan
   M ASUMI YAMAMURO1*, JUN-ICHI HIRATSUKA2, YU ISHITOBI3, SHINYA HOSOKAWA4
   and YOSHIYUKI NAKAMURA4
   1
    Geological Survey of Japan, AIST, Central 7, Higashi, Tsukuba, Ibaraki 305-8567, Japan
   2
    Shimane Research Group of Wildlife, c/o Dr. Iwao Sakamoto, Shimane University School of Medicine,
    Enyamachi, Izumo, Shimane 693-8501, Japan
   3
    Shimane Prefectural Institute of Public Health and Environmental Science,
    Nishihamasada, Matsue, Shimane 690-0122, Japan
   4
    Marine Environment and Engineering Department, Port and Airport Research Institute,
    Nagase, Yokosuka, Kanagawa 239-0826, Japan

   (Received 11 October 2005; in revised form 10 February 2006; accepted 12 April 2006)

   Zostera marina L. was intensively harvested until the early 1950s in Lake Nakaumi, a          Keywords:
   eutrophic estuarine lagoon. We have estimated the amount of nitrogen (N) and phos-           ⋅ Eelgrass harvest-
   phorus (P) removed from the lagoon through Z. marina harvesting. Lake Nakaumi              ing,
   lies in Tottori and Shimane prefectures, and the annual harvest of Z. marina in the          ⋅ Zostera marina,
   late 1940s in Tottori was recorded as at least 56,250 t wet weight. The nutrient con-         ⋅ 2,4-D,
                                                      ⋅ seagrass,
   tent of 56,250 t of Z. marina was calculated to be 61.9 t of N and 12.9 t of P, which is
                                                      ⋅ SAV,
   equivalent to 5.3% and 11%, respectively, of present annual nutrient loads to the           ⋅ Musculista
   lake. The nutrients formerly used by Z. marina were likely used by phytoplankton             senhausia,
   after the Z. marina started to decline in the mid-1950s at Lake Nakaumi. This shift in         ⋅ Corbicula
   the chief primary producer, from benthic macrophytes to phytoplankton, caused a             japonica.
   subsequent shift in secondary producers. Benthic fish and crustacean populations
   decreased and the non-commercial filter-feeding bivalve, Musculus senhausia, in-
   creased in Lake Nakaumi after the decline of seagrass beds. This affected the local
   economy, inducing not only eutrophication but also the collapse of local fisheries. On
   the other hand, at adjacent Lake Shinji, loss of submerged aquatic vegetation in-
   duced an increase of the commercial filter-feeding bivalve, Corbicula japonica, which
   doubled the fishery yield in the lake.



1. Introduction                            being correlated with increasing nutrient and sediment
  Submerged aquatic vegetation (SAV) includes both          inputs from development of the surrounding watershed
marine and freshwater rooted vegetation that grows un-         (Kemp et al., 1983). The most severe naturally occurring
derwater in shallow zones where light penetrates. SAV         change affecting seagrasses occurred in the 1930s, when
improves water quality by filtering nutrients and contami-       almost the entire North Atlantic population of eelgrass
nants from the water, stabilizing sediments, and damping        (Zostera marina L.) was destroyed by a wasting disease
wave action. It also provides food and shelter for water-       (Rasmussen, 1977).
fowl, fish, and shellfish (Kemp et al., 2004).               In Japan, Z. marina beds began to diminish around
  Both natural and human-induced disturbances have          1950 (Aioi et al., 2001). Quantitative data regarding the
decreased SAV populations. All SAV species in             timing, causes, and extent of the loss of eelgrass beds
Chesapeake Bay, USA, declined dramatically in the late         have not been recorded for all of Japan. Z. marina was
1960s and 1970s (Orth and Moore, 1983), the decline          commonly used as a green mulch and fertilizer in Japan
                                    until the 1950s (Aioi, 2004). Before the decline of the
                                    beds, harvesting of Z. marina probably protected these
* Corresponding author. E-mail: m-yamamuro@aist.go.jp         habitats from eutrophication because it removed the ni-
Copyright©The Oceanographic Society of Japan/TERRAPUB/Springer     trogen (N) and phosphorus (P) contained in the


                                 551
                                Thus, the loss of eelgrass beds should have caused eco-
                                nomic losses associated with ecosystem degradation
                                (Kahn and Kemp, 1985).
                                  Eelgrass beds previously extended all along the shal-
                                low shoreline of Lake Nakaumi in southwestern Japan.
                                The eelgrass was formerly harvested as fertilizer, but no
                                Z. marina currently survives in Lake Nakaumi. If details
                                regarding the eelgrass harvest are clarified, including the
                                amount harvested annually, it may be possible to quanti-
                                tatively evaluate the impact of the loss of eelgrass beds
                                on water quality, habitat opportunity for animals, and
                                human economics in Lake Nakaumi and its adjacent area.
                                We studied the issue from the perspectives of natural sci-
                                ence (including chemical analyses) as well as social sci-
                                ence (we conducted interviews and collected statistics and
                                historical documents). Lake Shinji, an adjacent
                                oligohaline lagoon, has also been studied for compari-
                                son.

                                2. Materials and Methods

                                2.1 Study area
                                  Lake Nakaumi (area 92.1 km2, mean depth 5.4 m)
                                and Lake Shinji (area 81.8 km2, mean depth 4.5 m) are
                                eutrophic coastal lagoons (Fig. 1A). Lake Nakaumi lies
                                in both Tottori and Shimane Prefectures (the Yumigahama
                                Peninsula is in Tottori Prefecture and the rest of the shore-
                                line lies in Shimane Prefecture). Lake Shinji lies entirely
                                in Shimane Prefecture. Seawater enters Lake Nakaumi
                                through the Sakai Channel and the Nakaura gate, while
                                oligohaline water from Lake Shinji is supplied by the
                                Ohashi River. The combination makes Lake Nakaumi
                                polyhaline. More than 70% of freshwater entering Lake
                                Shinji is discharged by the Hii River. The average con-
                                centrations of nitrogen (N) and phosphorous (P) in the
                                surface water of Lake Nakaumi are 444 and 44 µg L–1,
                                respectively; levels at Lake Shinji are 442 (TN) and 45
                                µg L–1 (TP) (Goto et al., 2004). These amounts exceed
                                the legal limits, which are 400 (TN) and 30 µg L–1 (TP)
                                (Law Concerning Special Measures for Preservation of
                                Lake Water Quality, Ministry of the Environment). Pre-
                                fectures are responsible for lowering the TN and TP lev-
                                els. The average chlorophyll a concentration in the sur-
Fig. 1. A: Lake Nakaumi and the surrounding area at present.  face water is 13.0 µ g L –1 at Lake Nakaumi and 16.8
  B: Bathymetry (m) and morphology of Lake Nakaumi in    µg L–1 at Lake Shinji (Goto et al., 2004). Monthly envi-
  1994. C: Bathymetry (m) and morphology of Lake Nakaumi   ronmental monitoring of the lakes began in 1976 for
  in 1954.                          chemical parameters (i.e., dissolved oxygen and nutrients)
                                and in 1981 for phytoplankton and zooplankton.

Z. marina biomass from the marine ecosystem. Disap-      2.2 Historical observations of eelgrass harvesting and
pearance of the eelgrass beds resulted in a loss not only     the lake environment
of nutrient control, but also of habitat and epiphytic food    We interviewed local inhabitants around Lake
sources for those creatures that used the Z. marina beds    Nakaumi, who are now in their 70s and 80s. Ten lived in
(which possibly included commercial fish and shellfish).    Shimane Prefecture and five in Tottori Prefecture. They


552  M. Yamamuro et al.
had all harvested Z. marina for use as fertilizer. We asked
about the duration, place, method, and amount of the har-
vest, the species of plant harvested, and the price at which
it was sold. We asked about the equipment and boats used
for the harvest, as well as how the harvested Z. marina
was processed. We also asked when and why the Z.
marina bed disappeared from Lake Nakaumi. Historical
documents and statistics describing seagrass harvesting
and previous water quality were collected at several city
and prefectural libraries. In addition, we interviewed six
elderly inhabitants around Lake Shinji to confirm whether
the ecological changes observed in Lake Nakaumi had
also occurred in the adjacent lagoon.

2.3 Chemical analysis of eelgrass
   Because there is no Z. marina in Lake Nakaumi to-
day, eelgrass was sampled at Akkeshi Bay (ca. lat 43°N,    Fig. 2. Harvesting Zostera marina with a rake especially de-
long 145°E) in Hokkaido, Japan, in August 2002 and         signed for seagrass. Photo from Sakaiminato City (1986).
analyzed for its nutrient content. The eelgrass sample was
pulled up from where it grew (in water about 1.5 m deep)
by people in a boat, in the same way as it had once been    Sargassum thunbergii (Mertens) O. Kuntze, which grew
harvested. The sampled eelgrass was stored in a plastic    on rocky bottoms. The area of Lake Nakaumi shallower
bag with ice and delivered to the laboratory three days    than 3 m, as measured by planimeter on a map published
after the sampling. The Z. marina sample, including sedi-   in 1954 (Fig. 1C), was 2012 ha. About 20% of that was
ment, detritus, and epiphytic organisms, was weighed to    rocky bottom, supporting S. thunbergii, meaning that the
determine total wet weight (ww). It was sorted, separat-    Z. marina bed at Lake Nakaumi in the early 1950s cov-
ing vital leaves, decaying leaves, and underground parts    ered about 1600 ha.
(roots and rhizomes). All parts were cleaned with a tooth-      The S. thunbergii harvest was less than one-sixth that
brush and weighed again, both before and after oven dry-    of Z. marina (Sakamoto, 1962), but this species was re-
ing (at 50°C), and homogenized separately.           garded as higher quality than Z. marina for use as ferti-
   The N content of Z. marina was determined with an     lizer. Macrophytes with attached Asian mussels (e.g.,
elemental analyzer (model MT-5, Yanagimoto, Kyoto,       Musculista senhausia (Benson)), were also regarded as
Japan) by the method of Yamamuro and Kayanne (1995),      fertilizer of higher quality.
and total P was determined colorimetrically with a          Because the tidal range in this area is less than 30
Technicon Auto-Analyzer (Model AACS-II,            cm, Z. marina was always submerged. A specially de-
BRAN+LUEBBE, Tokyo, Japan) after digestion, using       signed rake was used as a dredge to harvest the seagrass
the method described by Ohtsuki (1982).            from a wooden boat (Fig. 2). Although seagrass beds act
                                as a nursery for some commercial fish, local fishermen
3. Results                           did not oppose Z. marina harvesting. There are several
                                probable reasons: fishermen liked to collect the edible
3.1 Aquatic macrophytes use at Lake Nakaumi and Lake      red algae, Gracilaria verrucosa (Hudson) Papenfuss,
   Shinji                          which grew on the sandy bottoms created by the harvest-
   According to our informants, the collectors of aquatic  ing of Z. marina; in addition, Z. marina caused problems
macrophytes in Lake Nakaumi were usually farmers,       for fishermen where it grew adjacent to their boat slips.
rather than fishermen, who harvested aquatic macrophytes    The Japanese cockle, Scapharca subcrenata (Lischke),
to use on their own fields. However, a few professionals    was a chief target for fishermen, but was collected in
also harvested aquatic macrophytes. Such specialists      waters deeper than the Z. marina bed.
owned larger boats with engines, and they harvested         At Lake Nakaumi, Z. marina was harvested year-
aquatic macrophytes not only at Lake Nakaumi but also     round, but it was easier to harvest in summer, when it
along the coast of the Sea of Japan.              flowered and could be easily pulled out from the bottom.
   All the aquatic macrophytes that grew in Lake       S. thunbergii was harvested only in spring and early sum-
Nakaumi were harvested for fertilizer, but two species     mer.
were dominant: eelgrass, Z. marina, which grew on soft        Aquatic macrophytes were sold commercially by wet
bottoms at depths of around 3 m; and the brown alga      weight, especially along the eastern shore of Lake


                                         Ecosystem Shift by Loss of Eelgrass Bed  553
Table 1. Measured weight and nutrient content of Zostera marina samples, and estimated total quantity of nitrogen and phospho-
  rous removed annually from Lake Nakaumi through eelgrass harvesting in the late 1940s.

        Parts         Portion  Dry/Wet   Nitrogen  Phosphorus   Removed   Removed
                   (w/w %)  (w/w %)   (w/w %)   (w/w %)    N (t)    P (t)
        Vital leaf       57.8     8.1    1.85    0.353     48.6     9.3
        Decaying leaf      7.0     6.2    1.15    0.150     2.8     0.4
        Root and rhizome    19.9    13.3    0.70    0.219     10.4     3.3
        Attached materials   15.2     —      —      —      —      —

        Total                                  61.9    12.9



Nakaumi (the Yumigahama Peninsula), where the farm-       abundant before the mid-1950s, because the shallow bot-
ers specialized in growing cotton. Our informants said     tom now inhabited by C. japonica was covered with SAV.
that cotton requires more potassium than other crops, and    The SAV started to decline around the beginning of the
Z. marina was considered a superior fertilizer for cotton    1950s and had completely disappeared by 1960. Five of
because it contains more potassium than terrestrial plants.   the six Lake Shinji informants believed that the use of
Aquatic macrophytes were also used as fertilizer for taro,   herbicides was the cause of the SAV die-off.
potato, sweet potato, vegetables, and mulberry (for silk-
worm culture). On a nutrient basis, aquatic macrophytes     3.2 Nutrient removal through Zostera marina harvest-
were sold for only about 6% of the price of dried fish,        ing in Lake Nakaumi
which was also used as fertilizer. Therefore, cotton could      The wet weight of the Z. marina sampled at Akkeshi
be produced at far less cost where aquatic macrophytes     Bay, including sediment, detritus, and epiphytic organ-
were available for fertilizer than in areas where they were   isms, was 956.2 g. After cleaning, the wet weights of vi-
not.                              tal leaves, decaying leaves, and underground parts (roots
   The harvesting of aquatic macrophytes at Lake       and rhizomes) were 552.6, 67.4, and 190.7 g, respectively
Nakaumi began to decline in the mid-1950s. At that time,    (a loss of 145.5 g). The loss was attributed to the attached
farmers on the Yumigahama Peninsula gradually began       materials, and not to the Z. marina itself (Table 1). The
to change from growing cotton to tobacco, which has a      wet/dry weights (dw) varied from 6.2% to 13.3%, depend-
low salt tolerance. Chemical fertilizers with potassium     ing on the part of the plant being weighed. N and P con-
also came into common use about the same time. The       centration also varied considerably (Table 1). Therefore,
decline of the Z. marina bed also began in the mid-1950s.    we estimated the N and P content in the harvested Z.
Subsequently, the S. thunbergii and G. verrucosa biomass    marina by plant part.
also decreased, because they no longer grew in the deeper       In Tottori Prefecture, the mean annual harvest of
waters but only in the shallower waters along the shore.    aquatic macrophytes from Lake Nakaumi in 1948 and
Most aquatic macrophytes had disappeared by the end of     1949 was estimated by the local fisheries station as at
the 1960s. The remaining shallow beds were finally de-     least 56,250 t ww, most of which was Z. marina (Sokuri,
stroyed by land filling and reclamation projects begun in    1955). Based on our analytical results, the amount of N
1968. Why aquatic macrophytes originally began to de-      and P annually removed through the harvesting of Z.
cline in the mid-1950s is still unknown. Our informants     marina from the lake in Tottori Prefecture was 61.9 t and
said that it coincided with the beginning of the extensive   12.9 t, respectively (Table 1).
use of herbicides in the area.                    Annual production of the aboveground portion of Z.
   The use of SAV as fertilizer was not common around     marina (i.e., the leaves) at Otsuchi Bay, northeastern
Lake Shinji. Our informants who knew the SAV in Lake      Honshu, was estimated to be 1100–2400 g m–2 dw yr–1
Shinji were therefore all fishermen. Two of the six in-     (Iizumi, 1996). If comparable productivity rates were typi-
formants said that they used to harvest SAV in Lake Shinji   cal of the entire 1600-ha Z. marina bed in Lake Nakaumi,
for their own use as fertilizer. At present, one fisherman   total annual leaf production would be 17,600–38,400
mainly targets fish and the five others target the Asiatic   t yr–1 (dw). Applying the 8% dw/ww ratio of the leaves
clam, Corbicula japonica Prime. This clam is presently     (Table 1), 22–48 × 104 t yr–1 ww of Z. marina would have
the dominant macrobenthic fauna in Lake Shinji, account-    been produced during the early 1950s. Because 42% of
ing for 97% of the wet weight of all macrobenthos, in-     annual production in the lake was estimated to have been
cluding those without shells (Nakamura et al., 1988).      harvested from Lake Nakaumi each year in early 1960’s
However, all informants said that C. japonica was not as    (Sakamoto, 1962), the annual Z. marina harvest of 56,250


554  M. Yamamuro et al.
t from Tottori Prefecture in the late 1940s, estimated by
the local fisheries station, is consistent with these calcu-
lations.
   Decayed Z. marina in Lake Nakaumi sometimes
caused objectionable smells in the summer, and local fish-
eries officers consequently tried to use Z. marina as raw
material for sandpaper. They harvested 54 t ha–1 of Z.
marina in five days in June and nine days in August 1933
(Shimane Prefectural Fisheries Experimental Station,
1935). If the same rate is applicable to the entire 1600 ha
of the former Z. marina bed in Lake Nakaumi, 86,400 t
could have been harvested in one year. This is circum-
stantial evidence that Z. marina was once abundant
enough in Lake Nakaumi for an annual harvest of 56,250
t.
                                Fig. 3. Chlorophyll a concentration (mg L–1) in the surface
   The nutrient content of 56,250 t of Z. marina would      water of Lake Nakaumi in 1961. Data compiled from Ondoh
be 61.9 t of N and 12.9 t of P, while the present annual      (1962).
nutrient load to Lake Nakaumi is 1164 t N and 116 t P
(Shimane Prefecture, Report for FY 1998). Therefore, the
former Z. marina harvest removed an equivalent of 5.3%
and 11% of the present N and P loads to Lake Nakaumi.     times the European and 7.2 times the United States’ us-
Because we have excluded from consideration nutrients     age (Aioi, 2003). Sixty-seven percent of herbicides and
removed by the Z. marina harvested from Shimane Pre-      pesticides used in Japan are used on rice paddies, and
fecture, and because the lake is probably more eutrophic    35% of that amount is herbicides (Aioi, 2003).
now than it was in the late 1940s, these percentages should     In Japan, herbicides are registered at the Ministry of
be considered minimum estimates.                Agriculture, Forestry and Fisheries; once registered, the
                                herbicides are used widely. Herbicides are used inten-
4. Discussion                         sively in rice paddies for 30 to 50 days after transplanta-
                                tion of seedlings. Thus, it is during the rainy season, be-
4.1 Water quality change after eelgrass bed disappear-     tween June and July, that the annual dose of herbicides is
   ance                           sprayed in the rice paddies. During this period, the herbi-
   Loss of SAV is often attributed to increases in tur-   cides drain to rivers and downstream to eelgrass beds.
bidity due to eutrophication (Moore and Wetzel, 2000;        2,4-Dichlorophenoxyacetic acid (2,4-D), which in-
Scheffer et al., 2001). The chlorophyll a threshold, be-    hibits the growth of the temperate seagrass, Posidonia
yond which SAV is not present, is statistically derived as   oceanica (Balestri et al., 1998), and the Eurasian
15 µg L–1 in estuarine environments (Kemp et al., 2004).    watermilfoil, Myriophyllum spicatum (Sprecher et al.,
In Lake Nakaumi, maximum chlorophyll a concentration      1998), was registered in Japan in 1950 and subsequently
in 1961 (by which time most SAV had disappeared) was      used widely. The decline of the Z. marina beds in Lake
6 µg L–1 (Fig. 3), or well below the threshold. Therefore,   Nakaumi began in the mid-1950s, apparently coinciding
the decline of the eelgrass bed at Lake Nakaumi was not    with the beginning of the widespread use of herbicides in
likely to have been caused by eutrophication. Rather, it    the area. The Shimane Agricultural Experimental Station
is the loss of the eelgrass bed that contributed to the    (SAES) records showed that 2,4-D may have been used
eutrophication (Mitchell, 1989). The annual Z. marina     as early as 1950; the first record of the amount of 2,4-D
harvest in the late 1940s from Tottori Prefecture contained  used was 12.99 t in 1957. Because 2,4-D scarcely degrades
N and P amounts equivalent to 5% and 11% of the present    in soil and water (Chang et al., 1998; Toräng et al., 2003),
N and P loads, respectively, in all of Lake Nakaumi. The    a significant portion of the 2,4-D spread around Lake
nutrients previously used by Z. marina to sustain growth    Nakaumi may have entered the lake.
would have been used by phytoplankton instead, thus
causing further decline of the Z. marina bed by reducing    4.2 Changes in secondary production after the disap-
the amount of light reaching the bed.                pearance of eelgrass beds
   Aioi (2003) suggested that one of the causes of the      In 1961, when the eelgrass bed in Lake Nakaumi was
nationwide decline of Z. marina beds is the widespread     estimated to be only 89 ha (Sakamoto, 1962), total an-
use of herbicides in Japan. At present, 14.3 kg ha–1yr –1 of  nual production of Z. marina and phytoplankton in Lake
herbicides and pesticides are used in Japan, which is 3.2   Nakaumi were estimated to be 640 and 65,700 t yr–1 dw,


                                         Ecosystem Shift by Loss of Eelgrass Bed  555
respectively (Sakamoto, 1962). In the early 1950s, the Z.
marina in Lake Nakaumi may have covered 1600 ha, with
an annual production of 17,600–38,400 t yr–1 dw. Because
phytoplankton production in the early 1950s may have
been less than in 1961 (due to competition for nutrients
between Z. marina and the phytoplankton), primary pro-
duction by Z. marina may have been as significant as
phytoplankton in the early 1950s.
   The epiphytic zoobenthos abundance in the 89-ha Z.
marina bed of 1961 was estimated to be 1.5–4.8 t dw,       Fig. 4. Changes in the fisheries yield in Lake Nakaumi and
based on counts in sampled quadrats (Kikuchi, 1962), or        Lake Shinji. Fisheries yield in 1959 is from Kawanabe
3.3 t dw, determined by towing a plankton net over the        (1962); yield in 1996 is from the website of Shimane Fish-
bed (Sakamoto et al., 1962). At the same time, the abun-       eries Station of Inland Water (http://www2.pref.shimane.jp/
dance of zooplankton in all of Lake Nakaumi was esti-         naisuisi/).
mated to be 53.4 t (Yamaji, 1962). If the same relative
abundance is applied to the former 1600-ha bed, the abun-
dance of zooplankton and epiphytic zoobenthos would
have been 53.4 and 27–86 t dw, respectively. These esti-     trients absorbed by eelgrass were instead consumed by
mates suggest that epiphytic secondary producers were       phytoplankton, which thus increased, causing an increase
as important as planktonic secondary producers when the      in plankton feeders. The physical absence of eelgrass beds
Z. marina bed covered 1600 ha.                  may also cause a decrease in bottom dwellers. Before the
   Epiphytic primary production often exceeds seagrass     loss of the eelgrass beds, shrimp (i.e., Metapenaeus ensis),
production (Moncreiff et al., 1992; Pollard and Kogure,      bottom fish (mentioned above), edible seaweed
1993), and the epiphytic zoobenthos generally does not      (Gracilaria verrucosa), and cockles (Scapharca
consume seagrasses but feeds on epiphytes (Jernakoff et      subcrenata) were chief targets of commercial fisheries in
al., 1996). The production rate of epiphytes in the former    Lake Nakaumi. Comparing the fisheries yield in Lake
Z. marina bed of Lake Nakaumi is not available, but the      Nakaumi in 1958 and 1996, the seaweed and cockle fish-
equivalent abundance of zooplankton and epiphytic         eries were completely destroyed, while the total catch
zoobenthos indicates a high level of epiphyte production     decreased by 50% (Fig. 4).
in the Z. marina bed. The epiphytic and macrobenthic          In contrast, the neighboring oligohaline lake, Lake
primary producers would have been important organic        Shinji, shows an increase in shell fisheries. Due to the
producers in Lake Nakaumi before the decline of the Z.      difficulty of osmotic regulation in oligohaline water,
marina bed.                            Corbicula japonica (one of the most popular commercial
   The zooplankton fauna and the epiphytic zoobenthos     bivalves) is the only bivalve that can stably inhabit the
of the Z. marina bed were distinctively different. The      lake. As in Lake Nakaumi, SAV disappeared from Lake
dominant zooplankton taxa were Copepoda, Cladocera,        Shinji during the mid 1950s. Thereafter, the abundance
and Ciliata (Yamaji, 1962), whereas the epiphytic         of C. japonica increased, and the number of shellfish fish-
zoobenthos was dominated by Isopoda, Amphipoda, and        ermen increased accordingly. Because C. japonica feeds
Neomysis (Kikuchi, 1962; Sakamoto et al., 1962). Ow-       on phytoplankton, the loss of SAV and increase in
ing to these differences, it is possible to identify the fishes  phytoplankton may result in an increase in habitat and
that were dependent on the Z. marina bed for their food      food for C. japonica. It now inhabits the shallow sandy
sources.                             bottom (up to 3 m deep) that is barely reached by anoxic
   The contents of the digestive tracts of fishes from     water, with a population of 1000 ind m–2 (Nakamura et
Lake Nakaumi examined during the late 1950s and early       al., 1988).
1960s suggest that many of the bottom-dwelling fish de-         A similar event occurred in Lake Nakaumi. Although
pended directly on the epiphytic zoobenthos of the Z.       the number of cockles decreased, Musculus senhausia, a
marina bed for their food. For example, one rockfish,       fast-growing, opportunistic bivalve, increased and became
Sebastes inermis Cuvier, ate chiefly Amphipoda and other     the most dominant benthos in Lake Nakaumi, with a den-
epiphytic zoobenthos (Harada, 1962). Another rockfish,      sity of more than 5000 ind m–2 (Yamamuro et al., 2000).
Sebastes schlegeli Hilgendorf, lived in the Z. marina bed     Because M. senhausia is not viewed as a commercial spe-
and ate small epiphytic crustaceans (Harada, 1962). Japa-     cies, fisheries statistics showed a decline in shellfish har-
nese whiting, Sillago japonica Temminck et Schlegel, ate     vested (Fig. 4). The increase in filter-feeding bivalves is
Amphipoda and Neomysis (Kawanabe and Asano, 1962).        likely linked to the increase in phytoplankton. Since they
   After the eelgrass decline in Lake Nakaumi, the nu-     breed almost year-round (Yamamuro et al., 2000), M.


556  M. Yamamuro et al.
senhausia populations recover quickly at the end of an-    Aioi, K. (2004): Conservations of Zostera beds of the environ-
oxia. We assume that the cockles could not adapt to peri-     mental functions. Aquabiology, 26, 303–308 (in Japanese).
odic anoxia, induced by eutrophication after the decline    Aioi, K., T. Akimichi and Y. Omori (2001): Ethnoecological
of the seagrass bed.                        studies on the utilization of eelgrass Zostera marina and
                                  ecological study on its global extinction. Annual Report of
                                  the Interdisciplinary Research Institute of Environmental
5. Conclusion
                                  Sciences, 20, 13–17 (in Japanese).
   Our results showed that bottom-dwelling fish and     Balestri, E., L. Piazz and F. Cinelli (1998): In vitro germina-
crustaceans decreased and opportunistic filter-feeding       tion and seedling development of Posidonia oceanica.
bivalves increased in Lake Nakaumi during the years be-      Aquatic Botany, 60, 83–93.
tween the mid-1950s and present. One of the causes of     Chang, B. V., J. Y. Liu and S. Y. Yuan (1998): Dechlorination
this faunal change in Lake Nakaumi was probably the        of 2,4-dichlorophenoxyacetic acid and 2,4,5-
change in the primary producers, from benthic           trichlorophenoxyacetic acid in soil. The Science of the To-
macrophytes and epiphytes to planktonic microalgae, in       tal Environment, 215, 1–8.
those areas formerly occupied by Z. marina. In the early    Chesworth, J. C., M. E. Donkin and M. T. Brown (2004): The
1950s, the epiphytic zoobenthos of the bed was an im-       interactive effects of the antifouling herbicides Irgarol 1051
                                  and Diuron on the seagrass Zostera marina (L.). Aquatic
portant source of food for some bottom-dwelling fishes.
                                  Toxicology, 66, 293–305.
Since bottom-dwelling fishes and shrimp usually have
                                Goto, M., Y. Karino, H. Kamiya and Y. Ishitobi (2004): Water
high commercial value, the decline of the seagrass beds      quality of Lake Shinji and Lake Nakaumi. Report of the
may have reduced the economic value of this water area,      Shimane Prefectural Institute of Public Health and Envi-
not only in terms of water quality, but also in terms of      ronmental Science, 45, 112–116 (in Japanese).
fisheries productivity.                    Harada, E. (1962): Sebastes. p. 163–169. In Report on the Eco-
   Specifying the cause of the seagrass decline is indis-    logical Study for Fishes in Relation to the Nakaumi Recla-
pensable for the restoration of the beds. This study mod-     mation and Freshening Project, ed. by D. Miyadi (in Japa-
estly suggests the possibility that not only eutrophication    nese).
and construction but also herbicides may be the cause of    Iizumi, H. (1996): Temporal and spatial variability of leaf pro-
the loss of SAV. Some forms of 2,4-D are still used in       duction of Zostera marina L. at Otsuchi Bay, northern Ja-
                                  pan. p. 143–148. In Seagrass Biology: Proceedings of an
Japan. Pentachlorophenol (PCP), which is also used for
                                  International Workshop, Rottnest Island, Western Australia,
rice paddies, was registered in 1956. The EC50 value for
                                  25–29 January 1996, ed. by J. Kuo, R. C. Phillips, D. I.
the seagrass Thalassia testudinum, after 40 h exposure to     Walker and H. Kirkman, Faculty of Science, The Univer-
PCP, is 0.74 ppm (Walsh et al., 1982). Sediments of Lake      sity of Western Australia, Nedlands.
Shinji accumulated during 1945–1948 contained PCP, and     Jernakoff, P., A. Brearley and J. Nielsen (1996): Factors affect-
PCP is still detected in the surface sediments (Masunaga      ing grazer-epiphyte interactions in temperate seagrass mead-
et al., 2001). Diuron, which affected the growth of Z.       ows. Oceanography and Marine Biology Annual Review,
marina at concentrations greater than 0.5 µ g L –1         34, 109–162.
(Chesworth et al., 2004), was registered in 1960 in Ja-    Kahn, J. R. and W. M. Kemp (1985): Economic losses associ-
pan. Okamura et al. (2003) measured 142 water samples       ated with the degradation of an ecosystem: The case of sub-
collected from fishery harbours, marinas, and small ports     merged aquatic vegetation in Chesapeake Bay. Journal of
                                  Environmental Economics and Management, 12, 246–263.
along the coast of western Japan, and 19 waters showed
                                Kawanabe, H. (1962): Abundance of fishes. p. 113–122. In
Diuron concentrations greater than 0.5 µg L–1. If any of
                                  Report on the Ecological Study for Fishes in Relation to
these chemicals inhibit the growth of Z. marina, their       the Nakaumi Reclamation and Freshening Project, ed. by
continued use may interfere with the re-establishment of      D. Miyadi (in Japanese).
seagrass beds, even if light availability increases due to   Kawanabe, H. and H. Asano (1962): Other fishes. p. 170–179.
reduction in nutrient inflow.                   In Report on the Ecological Study for Fishes in Relation to
                                  the Nakaumi Reclamation and Freshening Project, ed. by
Acknowledgements                          D. Miyadi (in Japanese).
   We thank Dr. Hiroshi Mukai for sampling Z. marina    Kemp, W. M., R. R. Twilley, J. C. Stevenson, W. R. Boynton
at Akkeshi Bay. We also thank Mrs. Shin Chou and Mrs.       and J. C. Menas (1983): The decline of submerged vascular
Sayuri Umeda for their assistance with chemical analy-       plants in upper Chesapeake Bay—Summary of results con-
                                  cerning possible causes. Marine Technology Society Jour-
sis. This study was financially supported by the National
                                  nal, 17, 78–89.
Institute of Advanced Industrial Science and Technology
                                Kemp, W. M., R. Batiuk, R. Bartleson, P. Bergstrom, V. Carter,
(AIST).                              C. L. Gallego, W. Hunley, L. Karrh, E. W. Koch, J. M.
                                  Landwehr, K. M. Moore, L. Murray, M. Naylor, N. B.
             References                Rybicki, J. C. Stevenson and D. J. Wilcox (2004): Habitat
Aioi, K. (2003): Ecosystem of macrophyte beds and global en-    requirements for submerged aquatic vegetation in
  vironment. Iden, 57, 53–58 (in Japanese).

                                          Ecosystem Shift by Loss of Eelgrass Bed  557
  Chesapeake Bay: Water quality, light regime, and physical-     141–154.
  chemical factors. Estuaries, 27, 363–377.            Rasmussen, E. (1977): The wasting disease of eelgrass (Zostera
Kikuchi, T. (1962): Macro zoobenthos. p. 99–105. In Report       marina) and its effects on environmental factors and fauna.
  on the Ecological Study for Fishes in Relation to the       p. 1–51. In Seagrass Ecosystems, ed. by C. P. McRoy and
  Nakaumi Reclamation and Freshening Project, ed. by D.       C. Helfferich, Marcel Dekker, New York.
  Miyadi (in Japanese).                      Sakaiminato City (1986): History of Sakaiminato City.
Masunaga, S., Y. Yao, I. Ogura, S. Nakai, Y. Kanai, M.       Sakamoto, M. (1962): Production of benthic macrophyte. p. 84–
  Yamamuro and J. Nakanishi (2001): Identifying sources and     88. In Report on the Ecological Study for Fishes in Rela-
  mass balance of dioxin pollution in Lake Shinji basin, Ja-     tion to the Nakaumi Reclamation and Freshening Project,
  pan. Environmental Science and Technology, 35, 1967–        ed. by D. Miyadi (in Japanese).
  1973.                              Sakamoto, M., E. Harada and H. Kawanabe (1962): Produc-
Mitchell, S. F. (1989): Primary production in a shallow eutrophic   tion. p. 135–147. In Report on the Ecological Study for
  lake dominated alternately by phytoplankton and by sub-      Fishes in Relation to the Nakaumi Reclamation and Fresh-
  merged macrophytes. Aquatic Botany, 33, 101–110.          ening Project, ed. by D. Miyadi (in Japanese).
Moncreiff, C. A., M. Sullivan and A. E. Daehnick (1992): Pri-    Scheffer, M., S. Carpenter, J. A. Foley, C. Folke and B. Walker
  mary production dynamics in seagrass beds of Mississippi      (2001): Catastrophic shifts in ecosystems. Nature, 413, 591–
  Sound: The contributions of seagrass, epiphytic algae, sand    596.
  microflora, and phytoplankton. Marine Ecology Progress     Shimane Prefectural Fisheries Experimental Station (1935): In
  Series, 87, 161–171.                        situ experiment of harvesting Zostera marina using a har-
Moore, K. A. and R. L. Wetzel (2000): Seasonal variations in      vester with engine. Bulletin of Shimane Prefectural Fisher-
  eelgrass (Zostera marina L.) responses to nutrient enrich-     ies Experimental Station, FY 1933, 31–35 (in Japanese).
  ment and reduced light availability in experimental ecosys-   Sokuri, M. (1955): Fisheries in the Nakaumi. Annual Report of
  tems. Journal of Experimental Marine Biology and Ecol-       Tottori Prefectural Fisheries Experimental Station, 35, 1–
  ogy, 244, 1–28.                          15 (in Japanese).
Nakamura, M., M. Yamamuro, M. Ishikawa and H. Nishimura       Sprecher, S. L., K. D. Getsinger and A. B. Stewart (1998): Se-
  (1988): Role of the bivalve Corbicula japonica in the ni-     lective effects of aquatic herbicides on sago pondweed.
  trogen cycle in a mesohaline lagoon. Marine Biology, 99,      Journal of Aquatic Plant Management, 36, 64–68.
  369–374.                            Toräng, L., N. Nyholm and H. J. Albrechtsen (2003): Shifts in
Ohtsuki, A. (1982): Chemical methods for lake sediment analy-     biodegradation kinetics of the herbicides MCPP and 2,4-D
  sis. p. 147–151. In Guide for Environmental Research on      at low concentrations in aerobic aquifer materials. Envi-
  Lakes, ed. by Japan Society on Water Pollution Research,      ronmental Science and Technology, 37, 3095–3103.
  Association for Controlling Environmental Pollution, To-    Walsh, G. E., D. L. Hansen and D. A. Lawrence (1982): A flow-
  kyo (in Japanese).                         through system for exposure of seagrass to pollutants. Ma-
Okamura, H., I. Aoyama, Y. Ono and T. Nishida (2003): Anti-      rine Environmental Research, 7, 1–11.
  fouling herbicides in the coastal waters of western Japan.   Yamaji, I. (1962): Zooplankton. p. 89–96. In Report on the
  Marine Pollution Bulletin, 47, 59–67.               Ecological Study for Fishes in Relation to the Nakaumi
Ondoh, Y. (1962): Chlorophyll. p. 65–77. In Report on the Eco-     Reclamation and Freshening Project, ed. by D. Miyadi (in
  logical Study for Fishes in Relation to the Nakaumi Recla-     Japanese).
  mation and Freshening Project, ed. by D. Miyadi (in Japa-    Yamamuro, M. and H. Kayanne (1995): Rapid direct determi-
  nese).                               nation of organic carbon and nitrogen in carbonate-bearing
Orth, R. J. and K. A. Moore (1983): Chesapeake Bay: An un-       sediments using Yanaco MT-5 CHN analyzer. Limnol.
  precedented decline in submerged aquatic vegetation. Sci-     Oceanogr., 40, 1001–1005.
  ence, 222, 51–53.                        Yamamuro, M., J. Hiratsuka and Y. Ishitobi (2000): Seasonal
Pollard, P. C. and K. Kogure (1993): The role of epiphytic and     change in a filter-feeding bivalve Musculista senhousia
  epibenthic algal productivity in a tropical seagrass,       population of a eutrophic estuarine lagoon. J. Mar. Syst.,
  Syringodium isoetifolium (Aschers.) Dandy, community.       26, 117–126.
  Australian Journal of Marine and Freshwater Research, 44,




558  M. Yamamuro et al.