JIRCAS 2005 International Symposium in Ishigaki, 97-102

Red Soil Runoff from the Miyara River, and an Environmental Problem on Ishigaki Island

Kenji Banzai* and Ken Nakamura**

ABSTRACT

A large volume of sediments has entered the coastal areas of the Ryukyu (Nansei) Islands, particularly

Okinawa Island, as a result of large-scale agricultural land preparation and development projects. This has

resulted in damage to the valuable ecosystem of corrals and seaweed beds and to the coastal landscape and fisheries. A water-quality survey was conducted over a four-year period at the Miyara River to measure

suspended solids, nitrogen and other elements, and the river flow rate, so as to provide important information

for planning load reduction measures and management methodologies for the catchment area. The volume of red soil and the total nitrogen loads transported by the rivers were determined, taking into account the

frequent soil runoff and increases in nitrogen emissions associated with rainfall. The volume of fertilizer

input was also estimated from existing data.

Keywords: red soil flow, flow observation, L-Q equation, SS, nitrogen

1. Introduction

A large number of public works projects were carried out in the islands of Okinawa, centering on the main island of Okinawa, following the return of the islands to Japan from the United States of America.

In relation to agricultural land development and field development, large-scale projects involving the

elimination of slopes were implemented to create farmland suitable for mechanized farming. One of the widely used methods for turning slopes into farmland is field reclamation at improved slope, which involves

leveling the vertically irregular topography of slopes by cutting and filling to create flat land suitable for

mechanized farming.

Since this method results in relocation of large amounts of soil and the extensive exposure of immature soil, various types of damage, such as soil erosion by heavy rainfall and slope failure, often occur for a few

years following the completion of the work. These types of damage caused widespread devastation to large­

scale land-improvement projects in the Okinawa islands, including the main island of Okinawa and Ishigaki Island. The soil of the islands of Okinawa and Nansei is categorized as "red-yellow soil" (Kunigami Maaji),

which is highly susceptible to structural collapse as it is scattered by raindrops and rainwater. Land in

tropical areas typically has thin layers of humus and is subject to decomposition, ablation, and eluviation of

organic materials once the covering vegetation is removed; this is likely to create unstable soil conditions as a result of the loss of the water-retention or drainage capacity of soil humus.

* National Institute for Agro-Environmental Sciences Kannondai, Tsukuba, Ibaraki 305-8604

** Okinawa Subtropical Station, JIRCAS Maezato Kawarabaru, Ishigaki, Okinawa 907-0002

97

Red Soil Runoff from the Miyara River, and an Environmental Problem on Ishigaki Island - Kenji:Eanzai:@Ji::if,~IHIII_a]smnum

2. Monitoring of Flow Rate and Water Quality in the River Basins

Soil runoff from farmland is always associated with a deterioration of the quality of river waters. On

islands, where there is less flatland and therefore more farming on hills and slopes, there is an intimate

relationship between soil runoff and river water quality. We focused on the main rivers along which

agriculture and stock-raising are actively pursued. We made long-term hydrological measurements, such as

flow-rate measurements and we monitored water-quality parameters, such as the turbidity and suspended

solids (SS), to calculate rough values of water-quality loads associated with rainfall.

(1) Observation Points and Methodology We had six fixed observation points along the

Miyara River basin (Regina Weir, Kawahara Bridge,

Futamata Weir, Shinko Bridge, Nakamizu Bridge,

and Todoroki Bridge in Fig. I). A water gauge and an

automatic continuous water sampler were installed at

the Regina Bridge point in the downstream area of the

Miraya River. The water gauge was kept permanently

operational, whereas the water sampler was activated

mainly during rainfall or flood. Sample containers filled

with water samples were taken back to the laboratory for

analysis. In addition to the Miyara River watershed, the

other river basin of the Todoroki River was chosen for

comparison of the-measurement results. The properties

of the sampled water that were measured were the pH, Fig. 1. Basins of Miyara River and Todoroki River

electrical conductivity (EC), turbidity, and water level in

the river. Samples taken back to the laboratory were analyzed for total nitrogen (T-N), total phosphorus (T-P),

SS, etc.

(2) Land Use The watershed area of the Miyara River is 27 .34 km2 and it is mainly composed of the catchment areas

of the Maesato Dam and the Sokobaru Dam (forests) covering an area of 11.15 km2 (36%), paddy fields

covering an area of about 1.0 km2 (3.7% ), and upland fields (sugar cane, pineapple, pasture, etc.) covering

about 8.23 km2 (30% ). The area of the Todoroki River watershed is 13.23 km 2, consisting of 4.94 km2 of

forests and residential area (37%), 0.45 km2 of paddy fields (3.4%), and 3.88 km2 of upland fields (sugar

cane, pineapple, pasture, etc.) (29% ).

3. Monthly Observation Results

According to the Red Soil Prevention Ordinance of the Okinawa Prefecture, the SS level of the river

water should be below 200 mg/L. Observation of the river basin for over a year allowed us to identify

changes, periodicity, and long-term trends. It is not surprising to note that the results, as shown in Fig. 2,

often exceeded the limit values stipulated in the ordinance.

In terms of the relationship between turbidity and the measurement points, the lev_el of turbidity

increased (as shown in logarithmic representation) with increasing daily rainfall on the sampling days; such

results for rainy days showed an increase of about 100 times compared with those on days without rain. This

is probably the result of runoff of soil in a suspended state from the upland fields in the river basin (Fig.

2). Particularly high values were observed at the Shinko Bridge point, located in farmland in the upstream

98

JIRCAS 2005 International Symposium in l shigaki

area of the Miyara River, and the Todoroki Bridge point, along the Todoroki River, a river surrounded by

agricultural land.

10000

1000

~ 100

" E

"' (// 10

2001 2002

100

E E

200 • !If w

300

Rainfall mm -Hegin;;i W -e-

Hegina B. -.­Futamat W.

----Nakamlzu B. ~

Shinko B. -+­

Todoroki B

----

Fig. 2. SS(suspended solid) values at each observation point of the basins at 2001-2002

4. Annual SS at Two Points

SS measurements at two of the seven observation points , the Regina Bridge point and Todoroki Bridge

point, located in the downstream areas, are shown by year in Fig. 3. The values increased during years with

many rainy days and typhoons, and remained low in

years with less rainfall. The average SS values for

those two points were less than the regulatory limit

of 200 mg/L in red soil SS, but the maximum values

reached almost 1000 mg/L. The annual highs showed

a decline for the past four years, but as it is uncertain

whether the SS will continue to decline in the future,

it will still be necessary to continue soil-preservation

efforts on farmland. The river water under the

Todoroki Bridge in normal times is clear, in stark

contrast to the dark water observed during rainfall.

10000

1000

-' ';;.

• Hegina W. • - average " Todoroki B.

1-----..-,ilo-------~- I:,. - average

• E 100 1-----W-----z---- 1-T------'._..---i

(/) (/)

1 .__ __ __..__ ___ _______ ...... _ __J

1998 1999 2000 2001 2002 2003 year

Fig. 3. SS values by the year of 2 observation points

5. Changes of SS and T-N from the Estuary to the Upriver Point

Fig. 4 shows measured values for the observation points along the Miyara River , going upstream

from the estuary. The SS value at the Sokobaru Dam point is low, as the dam stores runoff water from the

mountainous area of Mount Omotodake and the sediments that flow into the river settle in the bottom of

the dam reservoir. A little way downstream of the dam, higher SS values are found, but no decline in SS is

observed on going further downstream. Once the SS value rises, the elevated value continues down to the

sea. Fig. 5 shows the values of the nitrogen content at each observation point. The nitrogen concentration

at the point of the Sokobaru Dam, where water flows down from sources in the mountains and settles, is the

lowest, and the value rises in the direction of the estuary. The value at the Shinko Bridge point, located in

the tributary area, is excessively high because of its proximity to pasture and cow barns. The value during

rainfall is particularly high at this observation point. The nitrogen concentration along the main river below

the point of confluence with this tributary is lower than that at the Shinko Bridge point as a result of dilution

of the water.

99

Red Soil Runoff from the Miyara River, and an Environmental Problem on Jshigaki Island - Kenji Banzai and K:en Nakamura===i

10000

1000

I •

_J 100

" bl E 0 (/) 10

0 . 1 ..i:: ..., ;:! 0

::E .... a.,

.::: 0:::

a Mean values at each site

• . : .. •• • ••

,_, Q)

bD a., 't

is ·;::: .:Q o:l

0 ro tu) 0 Q) bD

:::i:: a,

:::i::

••

• Q)

·= 8 bD a., 't

is o:l

0 .... o:l .:Q ..., ;:!

;:! o:l .... ro N

8 ..c rn 0 8 ..., ..!G rn

;:! 0 ~ J'.r.. U) z

••

• Q)

bD 't .... .0 0

..!G 0

:.2 U)

7~-----------------~ • Mo..-1 values at each site

61-------- ------ -----.- --, • • 51-- --- ----------- -Jl'---.J

~ 4

~ 3t-..-~ ~ ..-~~--.r-~~~-----,=--. .... -----i f",

0 ..c:

,_ .;, s G)

"' .., "' "' 0.0

" -d

~ "' -d 0.0 0 ~ ,_ Q ·;:; -d

~ "' co ~ " co

,.. ,.. "' "' "' ;. co

::,

"' "" "' s ..Q N 0

.:: "' 0.0 "' 0 -"' ~ :c "' -- -"' s " :r:: ::, 0 "' ..c:

I:., (/) -"' (/)

z Fig. 4. SS values from river mouth to the upper reaches observation point

Fig. 5. T-N values from river mouth to the upper reaches observation point

6. Overflow Volume of the Miyara River

The flow rate of the river water at the Hegina Weir and Hegina Bridge points was also measured at

designated observation times and during rainfall by means of an electromagnetic flowmeter. As a water

gauge was also installed at the Hegina Bridge point, we were able to chart the water level-flow rate curve,

which we refer to as the "first curve." At the Hegina Weir, river water is pumped up for distribution to other

areas as irrigation water. The water level decreases, except on rainy days, because of the pumping of water.

The pump operation/water level-flow rate curve is a chart that represents the relationship between the pump

discharge and the water level changes. Those two water level-flow rate curves, parts of which are shown in

Fig. 6, were used to calculate the river runoff. This Figure also shows the amount of rainfall and the water

levels. The total annual flow rate, calculated as above, between July 2000 and June 2001 is shown in Table 1:

for this one-year record, the rainfall was 2601 mm whereas the overflow was 1955 mm.

20

E 15 E

~ 10 C: ·ro 5 Cl'.

0 1.E+04

l I I -i L ~ ' I ' l _I 1 ... I • ~ • •• I

5.0

4 .0

1.E+03 3.0 VJ

' 2 .0 E M E 1.E+02 Q)

e :;: 1.E+01 0

U::

1.0 Q)

> 0.0 ~ ... - 1.0 Q)

--.; -2 .0 ~

1.E+OO - 3.0

-4.0

1.E-01 -5.0 co "' N a, "' M 0 r-- M 0 r-- "' N a, CD M 0 r--

""" co "' N 0 r--

' ' N N '-. N N ' N ' '-. N M ' N N ' N N r-- r-- ' ' ' co '-. '-. ' a, '-. '-. ' 0 0 ' ' '-. '-. '-. ' N ' ' ' ' '-. r-- r-- r-- '-. co co co ' a, a, a, 0 0 0 N N N 0 0 ' '-. ' 0 ' '-. ' 0 ' ' ' ' ' ~ ' ' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ' ' ' 0 ' ' ' 0 ' ' ' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N N 0 0 0 N 0 0 0 N 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

N N N N N N N N N N N 0 0 0 N 0 0 0 N 0 0 0 N N N N N N N N N

Fig. 6. Relationships among rainfall, water level and river flow rate

100

JIRCAS 2005 International Symposium in Ishigaki

7. Relationship between SS, T-N and Flow Rate

The annual soil runoff and the loads of nitrogen and phosphorus are calculated by multiplying the

river flow rate by the concentration of each element. Because the concentration values are not continuously

generated, an equation is necessary for estimating the annual loads from the discontinuous measurements. In this case, regression equations for flow rates and loads (observation values) were used to estimate the loads.

The regression equations (LQ equations) of the flow rate and loads are shown in Fig. 7 and Fig. 8. SS

values are represented in two equations, taking into account the rainfall intensity. Some of the calculations of

the loads using those regression equations are shown in Fig. 6. It is shown that the loads fluctuate with large

changes in the flow rate.

"' '-"" :'!"?

0 "'

-0

" -0 C:

" C.

"' :, (/)

1000000 1000.0

100000 I---- y = 65.1 06x 12698

____ --F.,_- ----j + T- N mg/L

• T- N g/s

10000

1000

100

10

0.1

0.01 0.1

R2 = 0.8456

--Y = 5.5765x0.9636

R2 = 0.8681

• Rain intensity above 5.5mm/ day

• 11 below 5.5mm/ day

10 100 1000 10000 River flow rate m3

/ s

z I

f-

100.0

10.0

1.0

0.1

0.0 0.0

- Regression curve

•

- .-----.--.,--Y = 0.7867x10,a3

0.1 1.0 10.0

Flow rate m3/ s

R2 = 0.9092

100.0 1000.0

Fig. 7. The relationship flow rate and SS load Fig. 8. The relationship flow rate and T-N load

1.00E+06

1.00E+OS

1.00E+04

1.00E+03

1.00E+02

1.00E+01

1.00E+OO

1.00E-01

1.00E- 02

' r-' 0 0 0

"'

---calculated SSkg/ h - --- ------....--- ------ ----- ------

• observed SSkg/ h • · calculatedT-Nkg/h •

• observedT-Nkg/ h

co LO "' 0 <O M CX) ..;- C') <.O M co LO C') <.O M co LO M 0

' "' C") ' "' "' ' "' ' "' ' ' "' ' ' "' C") r- '- ' ' co ' '- ' C') ' ' ' 0 ' '- ' ' ' "' "' ' ' '-' r- r- r- ' co co co ' C') C') C') 0 0 0 "' "' "' 0 '- ' ' 0 ' '- ' 0 ' ' ' ' .:::: ' ' .:::: ' ' 0 0 0 0 0 0 0 0 0 0 0 0 0 ' ' 0 0 ' 0 0 ' '- ' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

"' 0 0 0 "' 0 0 0 "' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

"' "' "' "' "' "' "' "' "' "' 0 0 0 "' "' 0 0 N N 0 0 0 N N N N N "' N N

Fig. 9. Relationships among SS load, T-N load and time

8. Annual Material Balances for SS, T-N, and T-P

The annual loads of SS, nitrogen, and phosphorus carried to the sea are 1882 metric tons, 68 metric

tons, and 7 metric tons, respectively. Assuming that the suspended solids are all made up of soil particles

and that the specific gravity of river basin subsoil is 1.3, by dividing the volume of particles by the total area

of fields, the average depth of soil erosion can be estimated to be 0.2 mm (Table 1).

Annual nitrogen and phosphorus inputs from domestic livestock excreta and fertilizers were calculated

on the basis of the Agricultural and Forestry Statistics for Ishigaki City to be 276 metric tons and 113 metric

tons, respectively (Table 2). The proportion of these inputs carried to the sea are therefore calculated to be

101

Red Soil Runoff from the Miyara River, and an Environmental Problem on Ishigaki Island -Kenji-~anzai:mi41Ji,:n,Nakarwm•

25% for nitrogen and 6% for phosphorus, respectively, which indicates that the outflow of phosphorus is

much smaller than that of nitrogen. It is probable that much of the phosphorus is absorbed by soil or plants.

Table. 1. Rainfall, river flow, suspended solids, Total N outflow and Total P from July 2000 to Jun 2001

Rainfall

(mm)

2601

River flow

(mm)

1995

Suspended solid T-N outflow volume

(ton) (ton)

1882 68

T-P outflow volume

(ton)

7

Table. 2. Total amounts of nitrogen and phosphorus in agriculture fertilizer and from livestock excretions in one year

T-N T-P (ton/year) (ton/year)

Livestock excretions*l 100 12

Agriculture fertilizer *2 176 101

Total amount 276 113

*1 It presumes based on the livestock numbers 2120 bred in a basin. *2 From the planted area according to crop kind, and the amount of their standard fertilization to presumption

9. Conclusion

Red soil runoff, a problem for the Okinawa prefecture, is an environmental pollutant originating from

development and preparation of farmland. It has caused a heavy accumulation of red soil in coastal areas of

the island, particularly near shallows where corrals grow, leading to destruction of corrals and other major

colonies of marine life. We investigated and measured the amount of red soil transported by rivers, the soil

runoff associated with the increased water overflow due to rainfall, and increased loads of nitrogen and

phosphorus during four years. The results, although incomplete, revealed the current status of soil runoff and

its associated problems. The authors would like to thank the staff of the Service Division, Okinawa Subtropical Station of

the Japan International Research Center for Agricultural Sciences (JIRCAS), for their assistance in the

installation of observation equipment and the investigation. We also thank Ms. Momoyo Shimabukuro, of

the Island Environment Management Laboratory, Okinawa Subtropical Station, JIRCAS, for her assistance

in sample analysis.

Acknowledgement: About this research, it is thankful to a member of Department of Field Management

in Okinawa Subtropical Station JIRCAS which installation of observation equipment was obtained and

offered support of investigation. In analysis of the sample water, it appreciates that Mrs. Shimabukuro, M. of

Island Environment Management Laboratory in the same station cooperated.

References

Banzai, K. and Nakamura K. 2003 "Soil particles runoff from cultivat.ed land water quality loading in river basin" Applied

Hydrology (JSIDRE), 16: 38-45 (in Japanese)

Banzai, K., Sugahara, K. and Nakamura K. 2004 "Red soil runoff and its control in Island Basins" In: Participatory strategy for

soil and water conservation, ERECON: 49-54

102