- Email: [email protected]
Evaluation of suitable hydraulic conditions for spawning of ayu with horizontal 2D numerical simulation and PHABSIM
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/ecolmodel
Evaluation of suitable hydraulic conditions for spawning of ayu with horizontal 2D numerical simulation and PHABSIM Takayuki Nagaya a,∗ , Yoshiki Shiraishi a , Kouki Onitsuka b , Makoto Higashino c , Tohru Takami c , Noriharu Otsuka d , Juichiro Akiyama b , Hiroaki Ozeki e a
CTI Engineering Co. Ltd., Daimyo, Fukuoka, Japan Department of Civil Engineering, Kyushu Institute of Technology, Tobata-ku, Kitakyushu, Japan c Department of Civil Engineering, Oita National College of Technology, Oita, Japan d Kyusyu Regional Development Bureau, Ministry of Land, Infrastructure and Transport, Nobeoka, Japan e Toyota Production Engineering Corporation, Chuo-ku, Fukuoka, Japan b
a r t i c l e
i n f o
a b s t r a c t
Article history:
Hydraulic structures such as a dam, weir and water gate affect on the flow depth, velocity,
Published on line 2 June 2008
bed topology, water quality and so on, so that the hydraulic structures may affect on the water environment for plants and aquatic lives in and around rivers. Therefore, before con-
Keywords:
struction of such hydraulic structures, it is necessary to conduct the environmental assess-
Ayu
ment. However, the accuracy of the environmental assessment is not so high at present,
Spawning
because the preference curves of the flow depth, velocity and so on are not clear. Recently,
PHABSIM
Nagaya et al. [Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Otsuka, N., Akiyama, J.,
Suitability index
Matsumoto, K., 2004. Evaluation of weight of parameter for spawning of ayu, Plecoglossus
2D numerical simulation
altivelis, and prediction of spawn density per unit area. In: Lee, J.H.W., Lam, K.M. (Eds.), Environmental Hydraulics and Sustainable Water Management, vol. 2. AA. Balkema Publishers, UK, pp. 947–954] pointed out that the velocity, flow depth and substrate affect on the suitability of spawning for ayu, Plecoglosses altivelis altivelis on the basis of statistical investigation. Preference curves of velocity and flow depth were made clear by Onitsuka et al. [Onitsuka, K., Nagaya, T., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Ozeki, H., Matsumoto, K., Shiraishi, Y., 2005. Suitable flow depth and velocity for spawning of ayu, Plecoglossus altivelis. In: Proceedings of the 31th Congress of IAHR, Korea, pp. 1850–1858]. On the other hand, the preference curve of substrate has not been investigated. In this study, the preference curve of substrate was investigated on the basis of field survey. Further, a numerical simulation by making use of a horizontal 2D numerical model was carried out with changing the discharge at the Gokasegawa River due to a lot of ayu inhabits in the Gokasegawa River. The numerical model is based on the model which was suggested by Nagata et al. [Nagata, N., Hosoda, T., Muramoto, Y., 2000. Numerical analysis of river channel processes with bank erosion. Journal of Hydraulic Engineering, ASCE 126 (4), 243–252]. The suitability of spawning for ayu is predicted with the preference curves of the flow depth, velocity and substrate. © 2008 Published by Elsevier B.V.
1.
∗
Corresponding author. E-mail address: [email protected] (T. Nagaya). 0304-3800/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.ecolmodel.2008.02.043
Introduction
Ayu, Plecoglossus altivelis altivelis, is the migratory fish which inhabits in Japan, Korean peninsula and the east of the continent of China (see Seki et al., 1988). They spawn at the lower reach of the river in autumn and drift downstream to the sea
134
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Fig. 1 – Life cycle of ayu, Plecoglossus altivelis altivelis.
as soon as hatching (see Ishida, 1959, 1976; Tsukamoto et al., 1987, 1988; Tsukamoto and Uchida, 1992; Uchida et al., 1990). After about half year, they migrate to the upper and middle reach of the river and feed on attached algae on rocks. In autumn, mature ayu spawns and then dies (see Fig. 1). The ayu is one of the most important fish for fishery from a point of view of economy in Japan. Further the environmental assessment law was established in 1997. One of the most famous environmental assessment methods is PHABSIM. However, the environmental assessment with PHABSIM has not been conducted quantitatively. This is because what kind of parameters affect on the suitability of spawning has not been investigated. Nagaya et al. (2004) conducted this research with a statistical analysis and found that the flow depth, velocity and substrate are the most effective parameters which affect on the suitability of spawning. Onitsuka et al. (2005) investigated on the preference curves of flow depth and velocity on the basis of field surveys. As a result, the preference curves of flow depth and velocity were proposed. On the other hand, the preference curve of substrate has not been investigated. Therefore, the following investigations are necessary to establish the environmental assessment. At first, the preference curve of substrate must be investigated. Second, the accuracy of the PHABSIM must be verified on the basis of field data. At this phase, the spawning suitability of ayu can be predicted after calculation of hydraulics such as flow depth, velocity. Finally, an evaluation of environment when the discharge changes can be conducted to compare the hydraulics of each discharge case.
In this study, a preference curve of substrate is investigated on the basis of field survey. A numerical simulation by making use of a horizontal 2D numerical model was carried out with changing the discharge at the Gokasegawa River. Further, an evaluation of environment when the discharge changes was conducted to compare the hydraulics of each discharge case. Such an examination was firstly conducted.
2.
PHABSIM
One of the most famous environmental assessment methods is PHABSIM. The available physical habitat is evaluated from the composite suitability index CSI. CSI is calculated from the suitability index such as flow depth SI(h), velocity SI(Um ), substrate SI(s), cover SI(c) and so on. CSI = SI(h) × SI(Um ) × SI(s) × SI(c) × . . .
(1)
The parameters on the right hand side in Eq. (1) are chosen according to the fish species. The suitability of spawning for ayu can be also evaluated from Eq. (1). The cover may not affect on the suitability of spawning for ayu, due to the area of the spawning bed of ayu is located downstream of the river. Unfortunately, what kinds of parameters are effective on the suitability of spawning of ayu. The effective parameters for spawning of ayu were selected from the several values such as water temperature, velocity, flow depth, channel slope, BOD, COD, pH, SS and so on, by making use of a principal compo-
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
nent regression analysis on the basis of the field data obtained from the Gokasegawa River, Japan, by Nagaya et al. (2004). As a result, it was found that the most effective parameters are water temperature, velocity and flow depth. Therefore, the suitability of spawning for ayu is calculated by the following equation: CSI(suitability of spawning for ayu) = SI(h) × SI(Um ) × SI(s) (2) Suzuki (1958) found that the ayu can spawn even if the flow depth is the same as the body height of the ayu. Onitsuka et al. (2005) also pointed out that the effects of the flow depth SI(h) on the suitability of spawning for ayu are quite small on the basis of their field survey and the previous data which was obtained by Ishida (1976). The preference curve of the flow depth SI(h) is described as follows (Shiraishi et al., 2007): SI(h) = 33.3h (0 ≤ h(m) < 0.03) SI(h) = 1 (0.03 ≤ h(m))
(3)
The velocity strongly affects on the suitability of spawning for ayu. The preference curve of the velocity for spawning of ayu SI(Um ) is suggested by Onitsuka et al. (2005) as shown by the following equation: SI(Um ) = 0 (0 ≤ Um (m/s) < 0.3) SI(Um ) = 3.3Um − 1 (0.3 ≤ Um (m/s) < 0.6) SI(Um ) = 1 (0.6 ≤ Um (m/s) < 1.0)
(4)
SI(Um ) = −1.4Um + 2.4 (1.0 ≤ Um (m/s) < 1.7) SI(Um ) = 0 (1.7 ≤ Um (m/s)) The preference curve of the substrate SI(s) is not cleared at present. Ishida (1976) pointed out that the bed condition must be “soft bed” to spawn. Soft bed condition means that the water is running through the aperture between the bed materials. On the other hand, “firm bed” means that the fine sediments are caught between the bed materials, so that the
135
water cannot run through the aperture between the bed materials. When the bed condition is soft bed, the ayu only can spawn, because the spawns of the ayu needs a lot of oxygen in the water. Fishermen pointed out that if someone walks on the soft bed, the bed materials around foots start to move. In contrast, if someone walks on the firm bed, the bed materials around foots do not start to move. Unfortunately, the soft bed and also firm bed is not made clear quantitatively. Therefore, the prediction of spawning suitability for ayu with PHABSIM is not able at present.
3.
Field surveys
Field surveys were conducted to obtain the preference curve of the substrate SI(s). Fig. 2 shows the map of Nobeoka city where is located at the middle part of Kyushu Island, Japan. The Gokasegawa River, Hourigawa River and Kitagawa River are running through Nobeoka city. Those rivers belong to Class A river. The administrator of the Class A river is Ministry of Land, Infrastructure and Transport (MLIT). The Gokasegawa River has a branch, i.e., Osegawa River. Those rivers reach to the Pacific Ocean. The length and watershed catchment area of the Gokasegawa River are 106 and 1820 km2 , respectively. Nobeoka city is frequently damaged by the flood, i.e., August 1993, September 1997 and September 2005. Fig. 3 shows the riverside situation of the Gokasegawa River in September 1997 when the flood occurs. A lot of cars sink into the Gokasegawa River. Therefore, the river improvement must be conducted as soon as possible. The Gokasegawa River has three spawning beds of ayu, i.e., Hyakken, Misu and Agata Rapid (see Nagaya et al., 2004). Those rapids were also described as the black circles in Fig. 2. The ayu spawns at the lower reach of the river in autumn. The field surveys were conducted in the Agata Rapid which is located at 4 km from the river mouth 5th January in 2005. Fig. 4 shows the measurement points. Measurement points from R1 to R16 are located at the spawning bed so that the bed
Fig. 2 – There are three spawning beds of ayu, i.e., Hyakken, Misu and Agata Rapid, in the Gokasegawa River.
136
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Fig. 3 – A sample of Flood damage in the Gokasegawa River (September 1997).
condition of almost these areas is soft bed. However, the bed condition of some area is firm bed. On the other hand, measurement points from P1 to P2 are not located at the spawning bed, so that the bed condition is firm bed. The mark time experiments were conducted. A man marks time on the soft bed and also on the firm bed and if bed materials start to move, he must stop the marking time experiments and count his cumulative steps. The bed materials which move were caught by a net and the weight of those were measured. The velocity measurements of three components were conducted with an electromagnetic current meter. The probe of the current meter was traversed in the vertical direction about 20–25 points. The measurement frequency and measurements time at each point are 10 Hz and 51.2 s, respectively.
Fig. 5 – The number of cumulative steps of each bed conditions, i.e., firm bed, semi-soft bed and soft bed.
Further, bed materials at measurement points were sampled within 0.1 m depth. The same surveys were conducted in Kitagawa River in 27th December in 2005.
4.
Preference curve of substrate
Fig. 5 shows the number of cumulative steps and its bed condition. There is no relationship between the number of cumulative steps and the bed condition. Fig. 6 shows the relationship between total weight of moved bed material and its
Fig. 4 – Measurement points in Gokasegawa River. R means the rapid and P means the pool.
137
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
depends on the velocity near the bed. The velocity near the bed concerns with the friction velocity. The friction velocity is calculated from the log-law: 1 y + y U = ln + Ar U∗ ks
(5)
In which y is the reference bed level, is the Karman constant and Ar is the integration constant. In the case that the grain size is same and the shape of grain is sphere, the reference bed level y is equal to 0.25 ks , the Karman constant is 0.412 and the integration constant Ar is 8.5 (Nezu and Rodi, 1986; Nezu and Nakagawa, 1993). The equivalent sand roughness ks can be obtained from the grain size distribution as follows: ks = d65
Fig. 6 – Relationship between total weight of removed bed material and bed conditions, i.e., firm bed, semi-soft bed and soft bed.
(6)
As a result, it was found that the ranges of the friction velocity at the soft beds are 0.05–0.12 m/s. Therefore, the following preference curve is suggested: SI(s) = 0 (0.05 > U∗ (m/s), 0.12 ≤ U∗ (m/s)) SI(s) = 1 (0.05 ≤ U∗ (m/s) < 0.12)
bed condition. The total weights of almost data in soft bed conditions are large. However, the weights of some data in soft bed conditions are small, so that both conditions are cannot be distinguished by the total weight with high accuracy. Fig. 7 shows the weights of moved bed material by one step and its bed condition. It was found that the soft bed condition is satisfied if the weights of the removed bed materials by one step are larger than 0.1 kg when someone marks time on the river bed. In contrast, this result is not useful to make a river channel planning, because flow characteristics such as the velocity, flow depth and so on are necessary to make it. Therefore, the relationship between the bed condition and hydraulics must be investigated quantitatively. The fine sediments between each large bed materials are washed away by the flow. The strength of this function
5.
(7)
2D model and calculation condition
A horizontal 2D numerical model in Cartesian coordinates system was described as following equations: ∂h ∂M ∂N + + =0 ∂t ∂x ∂z
(8)
∂M ∂Um M ∂Wm M ∂H 0x ∂U U ∂U W + + + gh =− −h m m −h m m ∂t ∂x ∂z ∂x ∂x ∂z (9)
W ∂N ∂Um N ∂Wm N ∂H 0z ∂U W ∂Wm m + + + gh =− −h m m −h ∂t ∂x ∂z ∂z ∂x ∂z (10)
U = 2D −Um h m
W = D −Um h m
∂U
∂U
W = 2D −Wm h m
Dh = 0.3hU∗
Fig. 7 – Relationship between weight of removed bed material by one step and bed conditions, i.e., firm bed, semi-soft bed and soft bed.
m
∂x m
∂y
+
−
∂Wm ∂x
∂W m
∂y
2 k 3
−
(11.a)
2 k 3
(11.b)
(11.c) (12)
In which x and z are the streamwise and spanwise coordinate, respectively, t is the time, g is the gravitational acceleration, is the density of water, H is the water level, 0 is the bed shear stress, Dh is coefficient of eddy viscosity, k is the depth averaged turbulence energy, Wm are the depth
138
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Table 1 – Calculation condition Upstream
Inlet discharge was given as following discharge Case A. 15 m3 /s: corresponding to 335-day discharge Case B. 25 m3 /s: corresponding to 185-day discharge Case C. 35 m3 /s: corresponding to discharge between 185-and 75-day discharge
Downstream
Predicted flow depth at Mitsuse water level observation station, where is located in the calculation region, was given. Predicted flow depth is calculated from H–Q curve 0.033 s/m1/3
Manning’s coefficient of roughness Time interval Judge of convergence of calculation
0.1 s When flow becomes steady
and W are the turbuaveraged velocity in the z direction, Um m lence fluctuation in the x and z directions, respectively. M and N are the discharge flux in the x and z directions, respectively. The generalized coordinate system proposed by Nagata et al. (2000) was used in order to apply to the complicated landform of the actual river. Therefore, Eqs. (8)–(12) were changed to the generalized coordinate system (see Nagaya and Matsuo, 2002). It is better to use small mesh and also large area when the calculation was conducted. Unfortunately, computer power and calculation cost are limited, so that the mesh sizes of the streamwise and spanwise directions were set to be 100 and 15 m, respectively. These data is obtained by the interpolation of the cross-sectional bed level data measured each 200 m pitch by Ministry of Land, Infrastructure and Transport (MLIT). The calculation domain is shown in Fig. 8. Table 1 shows the calculation condition. Three sets of discharges at the inlet edge of calculation domain were given as 15.0 m3 /s (corresponding to 335-day discharge), 25.0 m3 /s (corresponding to 185-day discharge) and 35.0 m3 /s (corresponding to discharge
between 185- and 75-day discharge). The predicted flow depth at Mitsuse water level observation station, where is located in the calculation domain, was given. The predicted flow depth is calculated from H to Q (flow depth and discharge relationship) curve. Manning’s coefficient of roughness n and time interval t were set to be 0.033 s/m1/3 and 0.1 s, respectively. The judge of convergence of calculation was conducted whether the flow becomes steady or not (Fig. 8).
6.
Results and discussions
6.1.
Velocity
Fig. 9 shows the velocity vector for each case (case A, B and C). In the case B, high velocity area is seen near the sand bank which is located at the upstream area of the calculation domain. This area is corresponding to the Hyakken Rapid. High velocity is also seen at the upstream region of the corner where is located at the center of this figure. This area is corresponding to the Misu Rapid. These imply that the accuracy of this calculation is high. It was found that the water surface area increases with an increase of the discharge. A small branch is generated inside of the inner bank of the corner when the discharge is 25 and 35 m3 /s. In particular, the side cavity is generated at the left side of the sand bank when the discharge is 25 m3 /s. There is a possibility that the small fish and crab can hide in such a branch, due to the velocity is lower than that in the main channel.
6.2. ayu
Composite suitability index CSI of spawning for
Flow characteristics such as flow depth, velocity and also friction velocity are already obtained from the 2D numerical calculation. The suitability index of flow depth, velocity and friction velocity of spawning for ayu is calculated from Eqs. (3), (4) and (7). Therefore, the composite suitability index CSI of spawning is calculated from Eq. (2). Fig. 10 shows the contour of composite suitability index CSI of ayu spawning for each case (case A, B and C). CSI takes high
Fig. 8 – Calculation domain includes three spawning bed of ayu.
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
139
Fig. 9 – (a) Velocity vector when discharge is 15 m3 /s (case A), (b) velocity vector when discharge is 25 m3 /s (case B) and (c) velocity vector when discharge is 35 m3 /s (case C).
value near the Hyakken and Misu Rapid when the discharge is 15 and 25 m3 /s. In contrast, CSI takes high value in almost area when the discharge is 35 m3 /s. The former result is corresponding to the field data (see Nagaya et al., 2005). However, the latter result is not corresponding to the field data. There are two kind of swimming speed such as burst speed and cruising one. The burst speed depends on fish species,
surrounding velocity and body length BL of fish. Unfortunately, such a relationship has not been investigated up to now. At present, the burst speed of the ayu UBS is estimated roughly from the following formula (Nakamura, 1995):
UBS = 12.4BL (cm/s)
(13)
140
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Fig. 10 – (a) Contour of composite suitability index CSI of spawning for ayu when discharge is 15 m3 /s (case A), (b) contour of composite suitability index CSI of spawning for ayu when discharge is 25 m3 /s (case B) and (c) contour of composite suitability index CSI of spawning for ayu when discharge is 35 m3 /s (case C).
Fish can maintain the burst speed shorter than a few second. On the other hand, fish can maintain the cruising speed more than 1 h. The cruising speed UCS is estimated from the following formula (Nakamura, 1995):
UCS = (2 − 4)BL (cm/s)
(14)
The body length of the ayu in the spawning season is almost 15 cm, so that the cruising speed UCS is 30 cm/s. The velocity in almost area when the discharge is 35 m3 /s is larger than the cruising speed UCS (30 cm/s). In such a situation, ayu cannot spawn. This implies that PHABSIM has fault. PHABSIM does not include the time dependent, because there is no time term in Eqs. (1) and (2). The time depen-
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
141
Fig. 12 – Variation of water surface area with changing discharge.
dent will be included in the PHABSIM near future by the authors.
6.3.
Biodiversity
There are various lives in the river such as fish, crab, shell, insect, plankton and so on. Suitable environment for each live is different from each other. Therefore, biodiversity is necessary in the river. The biodiversity is satisfied if the various velocity areas exist in the river. Fig. 11 shows the present occupying area for each velocity range. In the case of 15 m3 /s, almost velocities are in the range between 0.0 and 0.2. High velocity area is few. This implies that it is difficult to inhabit for lives that prefer high velocity area. In contrast, there are
Fig. 11 – (a) Present occupying area for each velocity range when discharge is 15 m3 /s (case A), (b) present occupying area for each velocity range when discharge is 25 m3 /s (case B) and (c) present occupying area for each velocity range when discharge is 35 m3 /s (case C). Fig. 13 – Variation of water edge length with changing discharge.
142
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
a lot of high velocity area in the cases of 25 and 35 m3 /s. Therefore, biodiversity may increases with an increase of the discharge. Biodiversity depends not only on the velocity but also on the water surface area and length of water edge. Fig. 12 shows the variation of the water surface area with changing discharge. Fig. 13 shows the variation of the length of water edge with changing discharge. The water surface area almost increases with an increase of the discharge. In contrast, the length of water edge increases with an increase of the discharge in the condition that the discharge is less than 25 m3 /s. The length of water edge decreases in the condition that the discharge is in the range between 25 and 35 m3 /s. The side cavity is seen when the discharge is 25 m3 /s (see Fig. 9(b)). On the other hand, such a side cavity disappears when the discharge is 35 m3 /s (see Fig. 9(c)). Nezu and Onitsuka (2001, 2002) pointed out that a lot of river lives inhabit in the side cavity and also near the vegetation. Therefore, it can be said that the biodiversity is high when the discharge is 25 m3 /s.
7.
Conclusion
The preference curve of substrate was investigated in the Gokasegawa Rivear and Kitagawa River and also a numerical simulation by making use of a horizontal 2D numerical model was carried out with changing the discharge at the Gokasegawa River. A new finding is described as follows: (1) The spawning suitability for ayu depends on flow depth, velocity and substrate. The preference curves of flow depth and velocity are already suggested by Onitsuka et al. (2005) and Shiraishi et al. (2005). In contrast, the preference curve of substrate has not been investigated at all. The preference curve of substrate was investigated on the basis of field surveys and that was suggested (Eq. (7)). (2) Evaluations of spawning suitability for ayu with a 2D numerical simulation and also PHABSIM may be conducted for the first time. (3) The results of composite suitability index (CSI) do not correspond to the field data. In particular, the accuracy becomes lower with an increase the flow discharge. This implies that PHABSIM is not always valid, due to the time dependent term is not included. PHABSIM must be included time dependent near future. (4) Biodiversity was also investigated from viewpoints of velocity range, length of water edge and water surface area. It was found that biodiversity is high when the discharge is 25 m3 /s in the case of the Gokasegawa River.
Acknowledgements The authors are thankful to M. Suda, H. Kudoh, M. Kai and S. Tsuchida who belong to Nobeoka Gokasegawa fisherman’s cooperative association and also K. Nagase who is a manager of Kitagawa fisherman’s cooperative association. The authors are also thankful to H. Takao, H. Kai and S. Aoki who belong
to Nobeoka Work Office, Ministry of Land, Infrastructure and Transport.
references
Ishida, R., 1959. On the spawning of the ayu, Plecoglossus altivelis T. & S.-I. Structure of the spawning shoal and spawning behaviour. Bulletin of the Japanese Society of Scientific Fisheries 25 (4), 259–268 (in Japanese). Ishida, R., 1976. An investigation of spawning behavior of ayu. Anima (43), 12–20 (in Japanese). Nagata, N., Hosoda, T., Muramoto, Y., 2000. Numerical analysis of river channel processes with bank erosion. Journal of Hydraulic Engineering, ASCE 126 (4), 243–252. Nagaya, T., Matsuo, K., 2002. Evaluation of fish habitat using two-dimensional numerical analysis. Advances in River Engineering, JSCE 8, 73–78 (in Japanese). Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Matsumoto, K., 2004. Evaluation of weight of parameter for spawning of ayu, Plecoglossus altivelis, and prediction of spawn density per unit area. In: Lee, J.H.W., Lam, K.M. (Eds.), Environmental Hydraulics and Sustainable Water Management, vol. 2. AA. Balkema Publishers, UK, pp. 947–954. Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Shiraishi, Y., 2005. Prediction of spawning suitability for Ayu, Plecoglossus altivelis, with horizontal 2-D numerical simulation. In: Proceedings of 31th Congress of IAHR, Korea, pp. 1920–1928. Nakamura, S., 1995. Topics of Fishways. Sankaido Publishing Co. Ltd (in Japanese). Nezu, I., Nakagawa, H., 1993. Turbulence in Open-Channel Flows, IAHR/AIRH Monograph. Balkema Publishers. Nezu, I., Onitsuka, K., 2001. Turbulent structures in partly vegetated open-channel flows with LDA and PIV measurements. Journal of Hydraulic Research, IAHR 39 (6), 629–642. Nezu, I., Onitsuka, K., 2002. PIV Measurements of side-cavity open-channel flows—Wando model in rivers. Journal of Visualization, The Visualization Society of Japan 5 (1), 77–84. Nezu, I., Rodi, W., 1986. Open-channel flow measurements with a laser doppler anemometer. Journal of Hydraulic Engineering, ASCE 112 (5), 335–355. Onitsuka, K., Nagaya, T., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Ozeki, H., Matsumoto, K., Shiraishi, Y., 2005. Suitable flow depth and velocity for spawning of Ayu, Plecoglossus altivelis. In: Proceedings of 31th Congress of IAHR, Korea, pp. 1850–1858. Seki, S., Taniguchi, N., Jeon, S., 1988. Genetic divergence among natural populations of ayu from Japan and Korea. Nippon Suisan Gakkaishi 54 (4), 559–568 (in Japanese). Shiraishi, Y., Onitsuka, K., Nagaya, T., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Ozeki, H., 2005. Field survey on suitable condition of habitat for ayu, Plecoglossus altivelis. In: Proceedings of 31th Congress of IAHR, Korea, pp. 1929–1938. Shiraishi, Y., Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Matoba, S., Akiyama, J., 2007. A proposal of river improvement project considering spawning of ayu, Plecoglossus altivelis altivelis, in Osegawa River. In: Proceedings of 32th Congress of IAHR, Italy, p. 10 (on CD-ROM). Suzuki, K., 1958. Suisanzosyoku 5 (3), 21–26. Tsukamoto, K., Aida, K., Otake, T., 1988. Plasma thyroxine concentration and upstream migratory behavior of juvenile ayu. Nippon Suisan Gakkaishi 54 (10), 1687–1693. Tsukamoto, K., Ishida, R., Naka, K., Kajihara, T., 1987. Switching of size and migratory pattern in successive generations of
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
landlocked ayu. American Fisheries Society Symposium 1, 492–506. Tsukamoto, K., Uchida, K., 1992. Migration mechanism of the ayu. In: Ilyichev, V.I., Anikiev, V.V. (Eds.), Oceanic and Anthropogenic Controls of Life in the Pacific Ocean. Kluwer Academic Publishers, pp. 145–172.
143
Uchida, K., Tsukamoto, K., Kajihara, T., 1990. Effects of environmental factors on jumping behaviour of the juvenile ayu Plecoglossus altivelis with special reference to their upstream migration. Nippon Suisan Gakkaishi 56 (9), 1393–1399.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/ecolmodel
Evaluation of suitable hydraulic conditions for spawning of ayu with horizontal 2D numerical simulation and PHABSIM Takayuki Nagaya a,∗ , Yoshiki Shiraishi a , Kouki Onitsuka b , Makoto Higashino c , Tohru Takami c , Noriharu Otsuka d , Juichiro Akiyama b , Hiroaki Ozeki e a
CTI Engineering Co. Ltd., Daimyo, Fukuoka, Japan Department of Civil Engineering, Kyushu Institute of Technology, Tobata-ku, Kitakyushu, Japan c Department of Civil Engineering, Oita National College of Technology, Oita, Japan d Kyusyu Regional Development Bureau, Ministry of Land, Infrastructure and Transport, Nobeoka, Japan e Toyota Production Engineering Corporation, Chuo-ku, Fukuoka, Japan b
a r t i c l e
i n f o
a b s t r a c t
Article history:
Hydraulic structures such as a dam, weir and water gate affect on the flow depth, velocity,
Published on line 2 June 2008
bed topology, water quality and so on, so that the hydraulic structures may affect on the water environment for plants and aquatic lives in and around rivers. Therefore, before con-
Keywords:
struction of such hydraulic structures, it is necessary to conduct the environmental assess-
Ayu
ment. However, the accuracy of the environmental assessment is not so high at present,
Spawning
because the preference curves of the flow depth, velocity and so on are not clear. Recently,
PHABSIM
Nagaya et al. [Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Otsuka, N., Akiyama, J.,
Suitability index
Matsumoto, K., 2004. Evaluation of weight of parameter for spawning of ayu, Plecoglossus
2D numerical simulation
altivelis, and prediction of spawn density per unit area. In: Lee, J.H.W., Lam, K.M. (Eds.), Environmental Hydraulics and Sustainable Water Management, vol. 2. AA. Balkema Publishers, UK, pp. 947–954] pointed out that the velocity, flow depth and substrate affect on the suitability of spawning for ayu, Plecoglosses altivelis altivelis on the basis of statistical investigation. Preference curves of velocity and flow depth were made clear by Onitsuka et al. [Onitsuka, K., Nagaya, T., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Ozeki, H., Matsumoto, K., Shiraishi, Y., 2005. Suitable flow depth and velocity for spawning of ayu, Plecoglossus altivelis. In: Proceedings of the 31th Congress of IAHR, Korea, pp. 1850–1858]. On the other hand, the preference curve of substrate has not been investigated. In this study, the preference curve of substrate was investigated on the basis of field survey. Further, a numerical simulation by making use of a horizontal 2D numerical model was carried out with changing the discharge at the Gokasegawa River due to a lot of ayu inhabits in the Gokasegawa River. The numerical model is based on the model which was suggested by Nagata et al. [Nagata, N., Hosoda, T., Muramoto, Y., 2000. Numerical analysis of river channel processes with bank erosion. Journal of Hydraulic Engineering, ASCE 126 (4), 243–252]. The suitability of spawning for ayu is predicted with the preference curves of the flow depth, velocity and substrate. © 2008 Published by Elsevier B.V.
1.
∗
Corresponding author. E-mail address: [email protected] (T. Nagaya). 0304-3800/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.ecolmodel.2008.02.043
Introduction
Ayu, Plecoglossus altivelis altivelis, is the migratory fish which inhabits in Japan, Korean peninsula and the east of the continent of China (see Seki et al., 1988). They spawn at the lower reach of the river in autumn and drift downstream to the sea
134
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Fig. 1 – Life cycle of ayu, Plecoglossus altivelis altivelis.
as soon as hatching (see Ishida, 1959, 1976; Tsukamoto et al., 1987, 1988; Tsukamoto and Uchida, 1992; Uchida et al., 1990). After about half year, they migrate to the upper and middle reach of the river and feed on attached algae on rocks. In autumn, mature ayu spawns and then dies (see Fig. 1). The ayu is one of the most important fish for fishery from a point of view of economy in Japan. Further the environmental assessment law was established in 1997. One of the most famous environmental assessment methods is PHABSIM. However, the environmental assessment with PHABSIM has not been conducted quantitatively. This is because what kind of parameters affect on the suitability of spawning has not been investigated. Nagaya et al. (2004) conducted this research with a statistical analysis and found that the flow depth, velocity and substrate are the most effective parameters which affect on the suitability of spawning. Onitsuka et al. (2005) investigated on the preference curves of flow depth and velocity on the basis of field surveys. As a result, the preference curves of flow depth and velocity were proposed. On the other hand, the preference curve of substrate has not been investigated. Therefore, the following investigations are necessary to establish the environmental assessment. At first, the preference curve of substrate must be investigated. Second, the accuracy of the PHABSIM must be verified on the basis of field data. At this phase, the spawning suitability of ayu can be predicted after calculation of hydraulics such as flow depth, velocity. Finally, an evaluation of environment when the discharge changes can be conducted to compare the hydraulics of each discharge case.
In this study, a preference curve of substrate is investigated on the basis of field survey. A numerical simulation by making use of a horizontal 2D numerical model was carried out with changing the discharge at the Gokasegawa River. Further, an evaluation of environment when the discharge changes was conducted to compare the hydraulics of each discharge case. Such an examination was firstly conducted.
2.
PHABSIM
One of the most famous environmental assessment methods is PHABSIM. The available physical habitat is evaluated from the composite suitability index CSI. CSI is calculated from the suitability index such as flow depth SI(h), velocity SI(Um ), substrate SI(s), cover SI(c) and so on. CSI = SI(h) × SI(Um ) × SI(s) × SI(c) × . . .
(1)
The parameters on the right hand side in Eq. (1) are chosen according to the fish species. The suitability of spawning for ayu can be also evaluated from Eq. (1). The cover may not affect on the suitability of spawning for ayu, due to the area of the spawning bed of ayu is located downstream of the river. Unfortunately, what kinds of parameters are effective on the suitability of spawning of ayu. The effective parameters for spawning of ayu were selected from the several values such as water temperature, velocity, flow depth, channel slope, BOD, COD, pH, SS and so on, by making use of a principal compo-
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
nent regression analysis on the basis of the field data obtained from the Gokasegawa River, Japan, by Nagaya et al. (2004). As a result, it was found that the most effective parameters are water temperature, velocity and flow depth. Therefore, the suitability of spawning for ayu is calculated by the following equation: CSI(suitability of spawning for ayu) = SI(h) × SI(Um ) × SI(s) (2) Suzuki (1958) found that the ayu can spawn even if the flow depth is the same as the body height of the ayu. Onitsuka et al. (2005) also pointed out that the effects of the flow depth SI(h) on the suitability of spawning for ayu are quite small on the basis of their field survey and the previous data which was obtained by Ishida (1976). The preference curve of the flow depth SI(h) is described as follows (Shiraishi et al., 2007): SI(h) = 33.3h (0 ≤ h(m) < 0.03) SI(h) = 1 (0.03 ≤ h(m))
(3)
The velocity strongly affects on the suitability of spawning for ayu. The preference curve of the velocity for spawning of ayu SI(Um ) is suggested by Onitsuka et al. (2005) as shown by the following equation: SI(Um ) = 0 (0 ≤ Um (m/s) < 0.3) SI(Um ) = 3.3Um − 1 (0.3 ≤ Um (m/s) < 0.6) SI(Um ) = 1 (0.6 ≤ Um (m/s) < 1.0)
(4)
SI(Um ) = −1.4Um + 2.4 (1.0 ≤ Um (m/s) < 1.7) SI(Um ) = 0 (1.7 ≤ Um (m/s)) The preference curve of the substrate SI(s) is not cleared at present. Ishida (1976) pointed out that the bed condition must be “soft bed” to spawn. Soft bed condition means that the water is running through the aperture between the bed materials. On the other hand, “firm bed” means that the fine sediments are caught between the bed materials, so that the
135
water cannot run through the aperture between the bed materials. When the bed condition is soft bed, the ayu only can spawn, because the spawns of the ayu needs a lot of oxygen in the water. Fishermen pointed out that if someone walks on the soft bed, the bed materials around foots start to move. In contrast, if someone walks on the firm bed, the bed materials around foots do not start to move. Unfortunately, the soft bed and also firm bed is not made clear quantitatively. Therefore, the prediction of spawning suitability for ayu with PHABSIM is not able at present.
3.
Field surveys
Field surveys were conducted to obtain the preference curve of the substrate SI(s). Fig. 2 shows the map of Nobeoka city where is located at the middle part of Kyushu Island, Japan. The Gokasegawa River, Hourigawa River and Kitagawa River are running through Nobeoka city. Those rivers belong to Class A river. The administrator of the Class A river is Ministry of Land, Infrastructure and Transport (MLIT). The Gokasegawa River has a branch, i.e., Osegawa River. Those rivers reach to the Pacific Ocean. The length and watershed catchment area of the Gokasegawa River are 106 and 1820 km2 , respectively. Nobeoka city is frequently damaged by the flood, i.e., August 1993, September 1997 and September 2005. Fig. 3 shows the riverside situation of the Gokasegawa River in September 1997 when the flood occurs. A lot of cars sink into the Gokasegawa River. Therefore, the river improvement must be conducted as soon as possible. The Gokasegawa River has three spawning beds of ayu, i.e., Hyakken, Misu and Agata Rapid (see Nagaya et al., 2004). Those rapids were also described as the black circles in Fig. 2. The ayu spawns at the lower reach of the river in autumn. The field surveys were conducted in the Agata Rapid which is located at 4 km from the river mouth 5th January in 2005. Fig. 4 shows the measurement points. Measurement points from R1 to R16 are located at the spawning bed so that the bed
Fig. 2 – There are three spawning beds of ayu, i.e., Hyakken, Misu and Agata Rapid, in the Gokasegawa River.
136
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Fig. 3 – A sample of Flood damage in the Gokasegawa River (September 1997).
condition of almost these areas is soft bed. However, the bed condition of some area is firm bed. On the other hand, measurement points from P1 to P2 are not located at the spawning bed, so that the bed condition is firm bed. The mark time experiments were conducted. A man marks time on the soft bed and also on the firm bed and if bed materials start to move, he must stop the marking time experiments and count his cumulative steps. The bed materials which move were caught by a net and the weight of those were measured. The velocity measurements of three components were conducted with an electromagnetic current meter. The probe of the current meter was traversed in the vertical direction about 20–25 points. The measurement frequency and measurements time at each point are 10 Hz and 51.2 s, respectively.
Fig. 5 – The number of cumulative steps of each bed conditions, i.e., firm bed, semi-soft bed and soft bed.
Further, bed materials at measurement points were sampled within 0.1 m depth. The same surveys were conducted in Kitagawa River in 27th December in 2005.
4.
Preference curve of substrate
Fig. 5 shows the number of cumulative steps and its bed condition. There is no relationship between the number of cumulative steps and the bed condition. Fig. 6 shows the relationship between total weight of moved bed material and its
Fig. 4 – Measurement points in Gokasegawa River. R means the rapid and P means the pool.
137
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
depends on the velocity near the bed. The velocity near the bed concerns with the friction velocity. The friction velocity is calculated from the log-law: 1 y + y U = ln + Ar U∗ ks
(5)
In which y is the reference bed level, is the Karman constant and Ar is the integration constant. In the case that the grain size is same and the shape of grain is sphere, the reference bed level y is equal to 0.25 ks , the Karman constant is 0.412 and the integration constant Ar is 8.5 (Nezu and Rodi, 1986; Nezu and Nakagawa, 1993). The equivalent sand roughness ks can be obtained from the grain size distribution as follows: ks = d65
Fig. 6 – Relationship between total weight of removed bed material and bed conditions, i.e., firm bed, semi-soft bed and soft bed.
(6)
As a result, it was found that the ranges of the friction velocity at the soft beds are 0.05–0.12 m/s. Therefore, the following preference curve is suggested: SI(s) = 0 (0.05 > U∗ (m/s), 0.12 ≤ U∗ (m/s)) SI(s) = 1 (0.05 ≤ U∗ (m/s) < 0.12)
bed condition. The total weights of almost data in soft bed conditions are large. However, the weights of some data in soft bed conditions are small, so that both conditions are cannot be distinguished by the total weight with high accuracy. Fig. 7 shows the weights of moved bed material by one step and its bed condition. It was found that the soft bed condition is satisfied if the weights of the removed bed materials by one step are larger than 0.1 kg when someone marks time on the river bed. In contrast, this result is not useful to make a river channel planning, because flow characteristics such as the velocity, flow depth and so on are necessary to make it. Therefore, the relationship between the bed condition and hydraulics must be investigated quantitatively. The fine sediments between each large bed materials are washed away by the flow. The strength of this function
5.
(7)
2D model and calculation condition
A horizontal 2D numerical model in Cartesian coordinates system was described as following equations: ∂h ∂M ∂N + + =0 ∂t ∂x ∂z
(8)
∂M ∂Um M ∂Wm M ∂H 0x ∂U U ∂U W + + + gh =− −h m m −h m m ∂t ∂x ∂z ∂x ∂x ∂z (9)
W ∂N ∂Um N ∂Wm N ∂H 0z ∂U W ∂Wm m + + + gh =− −h m m −h ∂t ∂x ∂z ∂z ∂x ∂z (10)
U = 2D −Um h m
W = D −Um h m
∂U
∂U
W = 2D −Wm h m
Dh = 0.3hU∗
Fig. 7 – Relationship between weight of removed bed material by one step and bed conditions, i.e., firm bed, semi-soft bed and soft bed.
m
∂x m
∂y
+
−
∂Wm ∂x
∂W m
∂y
2 k 3
−
(11.a)
2 k 3
(11.b)
(11.c) (12)
In which x and z are the streamwise and spanwise coordinate, respectively, t is the time, g is the gravitational acceleration, is the density of water, H is the water level, 0 is the bed shear stress, Dh is coefficient of eddy viscosity, k is the depth averaged turbulence energy, Wm are the depth
138
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Table 1 – Calculation condition Upstream
Inlet discharge was given as following discharge Case A. 15 m3 /s: corresponding to 335-day discharge Case B. 25 m3 /s: corresponding to 185-day discharge Case C. 35 m3 /s: corresponding to discharge between 185-and 75-day discharge
Downstream
Predicted flow depth at Mitsuse water level observation station, where is located in the calculation region, was given. Predicted flow depth is calculated from H–Q curve 0.033 s/m1/3
Manning’s coefficient of roughness Time interval Judge of convergence of calculation
0.1 s When flow becomes steady
and W are the turbuaveraged velocity in the z direction, Um m lence fluctuation in the x and z directions, respectively. M and N are the discharge flux in the x and z directions, respectively. The generalized coordinate system proposed by Nagata et al. (2000) was used in order to apply to the complicated landform of the actual river. Therefore, Eqs. (8)–(12) were changed to the generalized coordinate system (see Nagaya and Matsuo, 2002). It is better to use small mesh and also large area when the calculation was conducted. Unfortunately, computer power and calculation cost are limited, so that the mesh sizes of the streamwise and spanwise directions were set to be 100 and 15 m, respectively. These data is obtained by the interpolation of the cross-sectional bed level data measured each 200 m pitch by Ministry of Land, Infrastructure and Transport (MLIT). The calculation domain is shown in Fig. 8. Table 1 shows the calculation condition. Three sets of discharges at the inlet edge of calculation domain were given as 15.0 m3 /s (corresponding to 335-day discharge), 25.0 m3 /s (corresponding to 185-day discharge) and 35.0 m3 /s (corresponding to discharge
between 185- and 75-day discharge). The predicted flow depth at Mitsuse water level observation station, where is located in the calculation domain, was given. The predicted flow depth is calculated from H to Q (flow depth and discharge relationship) curve. Manning’s coefficient of roughness n and time interval t were set to be 0.033 s/m1/3 and 0.1 s, respectively. The judge of convergence of calculation was conducted whether the flow becomes steady or not (Fig. 8).
6.
Results and discussions
6.1.
Velocity
Fig. 9 shows the velocity vector for each case (case A, B and C). In the case B, high velocity area is seen near the sand bank which is located at the upstream area of the calculation domain. This area is corresponding to the Hyakken Rapid. High velocity is also seen at the upstream region of the corner where is located at the center of this figure. This area is corresponding to the Misu Rapid. These imply that the accuracy of this calculation is high. It was found that the water surface area increases with an increase of the discharge. A small branch is generated inside of the inner bank of the corner when the discharge is 25 and 35 m3 /s. In particular, the side cavity is generated at the left side of the sand bank when the discharge is 25 m3 /s. There is a possibility that the small fish and crab can hide in such a branch, due to the velocity is lower than that in the main channel.
6.2. ayu
Composite suitability index CSI of spawning for
Flow characteristics such as flow depth, velocity and also friction velocity are already obtained from the 2D numerical calculation. The suitability index of flow depth, velocity and friction velocity of spawning for ayu is calculated from Eqs. (3), (4) and (7). Therefore, the composite suitability index CSI of spawning is calculated from Eq. (2). Fig. 10 shows the contour of composite suitability index CSI of ayu spawning for each case (case A, B and C). CSI takes high
Fig. 8 – Calculation domain includes three spawning bed of ayu.
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
139
Fig. 9 – (a) Velocity vector when discharge is 15 m3 /s (case A), (b) velocity vector when discharge is 25 m3 /s (case B) and (c) velocity vector when discharge is 35 m3 /s (case C).
value near the Hyakken and Misu Rapid when the discharge is 15 and 25 m3 /s. In contrast, CSI takes high value in almost area when the discharge is 35 m3 /s. The former result is corresponding to the field data (see Nagaya et al., 2005). However, the latter result is not corresponding to the field data. There are two kind of swimming speed such as burst speed and cruising one. The burst speed depends on fish species,
surrounding velocity and body length BL of fish. Unfortunately, such a relationship has not been investigated up to now. At present, the burst speed of the ayu UBS is estimated roughly from the following formula (Nakamura, 1995):
UBS = 12.4BL (cm/s)
(13)
140
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
Fig. 10 – (a) Contour of composite suitability index CSI of spawning for ayu when discharge is 15 m3 /s (case A), (b) contour of composite suitability index CSI of spawning for ayu when discharge is 25 m3 /s (case B) and (c) contour of composite suitability index CSI of spawning for ayu when discharge is 35 m3 /s (case C).
Fish can maintain the burst speed shorter than a few second. On the other hand, fish can maintain the cruising speed more than 1 h. The cruising speed UCS is estimated from the following formula (Nakamura, 1995):
UCS = (2 − 4)BL (cm/s)
(14)
The body length of the ayu in the spawning season is almost 15 cm, so that the cruising speed UCS is 30 cm/s. The velocity in almost area when the discharge is 35 m3 /s is larger than the cruising speed UCS (30 cm/s). In such a situation, ayu cannot spawn. This implies that PHABSIM has fault. PHABSIM does not include the time dependent, because there is no time term in Eqs. (1) and (2). The time depen-
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
141
Fig. 12 – Variation of water surface area with changing discharge.
dent will be included in the PHABSIM near future by the authors.
6.3.
Biodiversity
There are various lives in the river such as fish, crab, shell, insect, plankton and so on. Suitable environment for each live is different from each other. Therefore, biodiversity is necessary in the river. The biodiversity is satisfied if the various velocity areas exist in the river. Fig. 11 shows the present occupying area for each velocity range. In the case of 15 m3 /s, almost velocities are in the range between 0.0 and 0.2. High velocity area is few. This implies that it is difficult to inhabit for lives that prefer high velocity area. In contrast, there are
Fig. 11 – (a) Present occupying area for each velocity range when discharge is 15 m3 /s (case A), (b) present occupying area for each velocity range when discharge is 25 m3 /s (case B) and (c) present occupying area for each velocity range when discharge is 35 m3 /s (case C). Fig. 13 – Variation of water edge length with changing discharge.
142
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
a lot of high velocity area in the cases of 25 and 35 m3 /s. Therefore, biodiversity may increases with an increase of the discharge. Biodiversity depends not only on the velocity but also on the water surface area and length of water edge. Fig. 12 shows the variation of the water surface area with changing discharge. Fig. 13 shows the variation of the length of water edge with changing discharge. The water surface area almost increases with an increase of the discharge. In contrast, the length of water edge increases with an increase of the discharge in the condition that the discharge is less than 25 m3 /s. The length of water edge decreases in the condition that the discharge is in the range between 25 and 35 m3 /s. The side cavity is seen when the discharge is 25 m3 /s (see Fig. 9(b)). On the other hand, such a side cavity disappears when the discharge is 35 m3 /s (see Fig. 9(c)). Nezu and Onitsuka (2001, 2002) pointed out that a lot of river lives inhabit in the side cavity and also near the vegetation. Therefore, it can be said that the biodiversity is high when the discharge is 25 m3 /s.
7.
Conclusion
The preference curve of substrate was investigated in the Gokasegawa Rivear and Kitagawa River and also a numerical simulation by making use of a horizontal 2D numerical model was carried out with changing the discharge at the Gokasegawa River. A new finding is described as follows: (1) The spawning suitability for ayu depends on flow depth, velocity and substrate. The preference curves of flow depth and velocity are already suggested by Onitsuka et al. (2005) and Shiraishi et al. (2005). In contrast, the preference curve of substrate has not been investigated at all. The preference curve of substrate was investigated on the basis of field surveys and that was suggested (Eq. (7)). (2) Evaluations of spawning suitability for ayu with a 2D numerical simulation and also PHABSIM may be conducted for the first time. (3) The results of composite suitability index (CSI) do not correspond to the field data. In particular, the accuracy becomes lower with an increase the flow discharge. This implies that PHABSIM is not always valid, due to the time dependent term is not included. PHABSIM must be included time dependent near future. (4) Biodiversity was also investigated from viewpoints of velocity range, length of water edge and water surface area. It was found that biodiversity is high when the discharge is 25 m3 /s in the case of the Gokasegawa River.
Acknowledgements The authors are thankful to M. Suda, H. Kudoh, M. Kai and S. Tsuchida who belong to Nobeoka Gokasegawa fisherman’s cooperative association and also K. Nagase who is a manager of Kitagawa fisherman’s cooperative association. The authors are also thankful to H. Takao, H. Kai and S. Aoki who belong
to Nobeoka Work Office, Ministry of Land, Infrastructure and Transport.
references
Ishida, R., 1959. On the spawning of the ayu, Plecoglossus altivelis T. & S.-I. Structure of the spawning shoal and spawning behaviour. Bulletin of the Japanese Society of Scientific Fisheries 25 (4), 259–268 (in Japanese). Ishida, R., 1976. An investigation of spawning behavior of ayu. Anima (43), 12–20 (in Japanese). Nagata, N., Hosoda, T., Muramoto, Y., 2000. Numerical analysis of river channel processes with bank erosion. Journal of Hydraulic Engineering, ASCE 126 (4), 243–252. Nagaya, T., Matsuo, K., 2002. Evaluation of fish habitat using two-dimensional numerical analysis. Advances in River Engineering, JSCE 8, 73–78 (in Japanese). Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Matsumoto, K., 2004. Evaluation of weight of parameter for spawning of ayu, Plecoglossus altivelis, and prediction of spawn density per unit area. In: Lee, J.H.W., Lam, K.M. (Eds.), Environmental Hydraulics and Sustainable Water Management, vol. 2. AA. Balkema Publishers, UK, pp. 947–954. Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Shiraishi, Y., 2005. Prediction of spawning suitability for Ayu, Plecoglossus altivelis, with horizontal 2-D numerical simulation. In: Proceedings of 31th Congress of IAHR, Korea, pp. 1920–1928. Nakamura, S., 1995. Topics of Fishways. Sankaido Publishing Co. Ltd (in Japanese). Nezu, I., Nakagawa, H., 1993. Turbulence in Open-Channel Flows, IAHR/AIRH Monograph. Balkema Publishers. Nezu, I., Onitsuka, K., 2001. Turbulent structures in partly vegetated open-channel flows with LDA and PIV measurements. Journal of Hydraulic Research, IAHR 39 (6), 629–642. Nezu, I., Onitsuka, K., 2002. PIV Measurements of side-cavity open-channel flows—Wando model in rivers. Journal of Visualization, The Visualization Society of Japan 5 (1), 77–84. Nezu, I., Rodi, W., 1986. Open-channel flow measurements with a laser doppler anemometer. Journal of Hydraulic Engineering, ASCE 112 (5), 335–355. Onitsuka, K., Nagaya, T., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Ozeki, H., Matsumoto, K., Shiraishi, Y., 2005. Suitable flow depth and velocity for spawning of Ayu, Plecoglossus altivelis. In: Proceedings of 31th Congress of IAHR, Korea, pp. 1850–1858. Seki, S., Taniguchi, N., Jeon, S., 1988. Genetic divergence among natural populations of ayu from Japan and Korea. Nippon Suisan Gakkaishi 54 (4), 559–568 (in Japanese). Shiraishi, Y., Onitsuka, K., Nagaya, T., Higashino, M., Takami, T., Otsuka, N., Akiyama, J., Ozeki, H., 2005. Field survey on suitable condition of habitat for ayu, Plecoglossus altivelis. In: Proceedings of 31th Congress of IAHR, Korea, pp. 1929–1938. Shiraishi, Y., Nagaya, T., Onitsuka, K., Higashino, M., Takami, T., Matoba, S., Akiyama, J., 2007. A proposal of river improvement project considering spawning of ayu, Plecoglossus altivelis altivelis, in Osegawa River. In: Proceedings of 32th Congress of IAHR, Italy, p. 10 (on CD-ROM). Suzuki, K., 1958. Suisanzosyoku 5 (3), 21–26. Tsukamoto, K., Aida, K., Otake, T., 1988. Plasma thyroxine concentration and upstream migratory behavior of juvenile ayu. Nippon Suisan Gakkaishi 54 (10), 1687–1693. Tsukamoto, K., Ishida, R., Naka, K., Kajihara, T., 1987. Switching of size and migratory pattern in successive generations of
e c o l o g i c a l m o d e l l i n g 2 1 5 ( 2 0 0 8 ) 133–143
landlocked ayu. American Fisheries Society Symposium 1, 492–506. Tsukamoto, K., Uchida, K., 1992. Migration mechanism of the ayu. In: Ilyichev, V.I., Anikiev, V.V. (Eds.), Oceanic and Anthropogenic Controls of Life in the Pacific Ocean. Kluwer Academic Publishers, pp. 145–172.
143
Uchida, K., Tsukamoto, K., Kajihara, T., 1990. Effects of environmental factors on jumping behaviour of the juvenile ayu Plecoglossus altivelis with special reference to their upstream migration. Nippon Suisan Gakkaishi 56 (9), 1393–1399.