Introduction

This study presents a quick scan of several sediment flushing scenarios for the Funagira dam in Japan on the habitat suitability of the fish species Ayu (Plecoglossus altivelis) both upstream of the dam in the reservoir and in the downstream Tenryuu river. Several scenarios have earlier been investigated in the study of Omer et al. (2017) to increase the sediment transport through the dam and to prevent erosion of the Tenryuu River downstream of the dam. The results of these scenarios were used to assess the relative effect on the habitat suitability of Ayu compared to the current release scheme.  The Ayu (Figure 1) is a commercially valuable fish that inhabits Japan, Korea and Eastern China. It is a migratory fish species that spawns in the lower river reaches in autumn, and drifts towards the sea after hatching. After about half a year they migrate up the river to spawn and then die (Nagaya et al., 2008).

Figure 1  Ayu (Plecoglossus altivelis). Source: Wikipedia.

Methods

Habitat suitability calculation

The spatial analysis tool HABITAT (Haasnoot & van de Wolfshaar, 2009; habitat.deltares.nl) was used (Early Preview version 3.0.1.30997) to establish the habitat suitability model for feeding and spawning of Ayu. HABITAT uses response curves based on a set of critical habitat requirements as input. These curves indicate the boundaries of environmental conditions for the expected presence of species. HABITAT translates each environmental parameter map to a habitat suitability index map on a scale from zero to one. The limiting environmental parameter determines the eventual suitability, which is calculated as the minimum suitability of all input parameters per grid-cell.  J-Power provided information on environmental thresholds for a range of parameters and several life stages of Ayu. Several filters were applied to select the final parameters and life history events that were used in the habitat assessment. Only water depth and flow velocity for adult fish could be selected as input variables, since water quality parameters were not taken into account in the hydro-morphodynamic modelling. For both parameters, the thresholds for the life history events of feeding and spawning were fully covered by the data. The response curves that were used for the habitat assessment are depicted in Figure 2.

Figure 2.1   Response curves for water depth and flow velocity used for the habitat assessment of Ayu fish.

The sediment flushing scenarios investigated in the study by Omer et al. (2017) were used to extract water depth and flow velocity for the habitat suitability assessment for each scenario. These scenarios comprise combinations of three different discharges and three different reservoir water levels (Figure 2).

Figure 2.  Overview of sediment flushing scenarios from Omer et al. (2017).

The parameters were extracted for all scenarios from the hydro-morphodynamic model output from the last time step of the simulation and converted into *.asc format. After the habitat calculations, the final habitat suitability index map for feeding and spawning was projected back onto the curvilinear grid.  

Data Analysis

The habitat suitability maps were classified into habitat suitability classes of low suitability (0 – 0.2), fair suitability (0.2-0.4), mediocre suitability (0.4-0.6), good suitability (0.6-0.8) and very good suitability (0.8-1.0). For each class the total area was calculated per scenario for feeding and spawning upstream and downstream of the Funagira dam. For all scenarios a relative comparison was made to the current dam operating scheme (Q1000, wl50) and expressed in positive or negative percentage of areal change. Difference maps of habitat gain and loss, water gain and exceedance of flow velocity thresholds were made to evaluate the cause of habitat loss and gain based on combinations of water depth and flow velocity.

Results

Current conditions

In the current dam operating regime, most of the potential suitable habitat for the Ayu is located upstream of the Funagira dam (Figure 3 and Table 1). The largest habitat patches are found on the lee-side of the point bars or mid-channel bars (Figure 3.1 b, boxed areas). These zones are relatively shallow with lower flow velocities (Figure 3 a, boxed areas). Downstream of the dam the potential habitat is restricted to a thin line, representing the shallower zones adjacent to the channel. The habitat suitability for feeding and spawning is very similar due to the limited difference in response curves for water depth and flow velocity.

Figure 3. Spatial results for the current dam operating regime (Q1000, wl50.6). a) Water depth and flow velocity serving as input parameters to calculate habitat suitability, b) calculated habitat suitability for Ayu for feeding and spawning.

Table 1. Statistics of Ayu habitat suitability (hectares) of the current operating regime (Q1000, wl 50.6) per habitat suitability class.           

 scenario results

For all flushing scenarios, except the Q1000wl47 regime, the total habitat suitability for the whole modelled area is larger than the current operating regime (Figure 4a). However, if the upstream and downstream sections are analysed separately, it is clear that this increase in habitat occurs mostly upstream of the dam. All scenarios, except Q1000wl48 and Q1000wl48, show a large increase in spawning habitat in the upstream area compared to the current operating regime (Figure 4b). The downstream section shows contrasting results. Here, all scenarios except Q1000wl47 and Q1000wl48 show a decrease in potential spawning habitat.

Figure 4   Potential spawning habitat (sum of mediocre, good and very good habitat) for all flushing scenarios upstream and downstream of the dam. Due to large similarities between feeding and spawning habitat, only results for spawning are shown here. a) Sum of suitability for spawning (in hectares). b) Relative difference of the flushing scenarios compared to the current operating regime for upstream and downstream spawning habitat.

The increase in suitable habitat upstream of the dam mainly occurs on the lee-side of the uppermost meander bend (Figure 5 and spatial habitat suitability maps in Appendix I). This is due to the increased amount of water that creates a net increase in suitability (Figure 6). Additionally, the largest habitat decline downstream occurs in the high discharge scenarios adjacent to the channel where flow velocities now exceed the critical habitat threshold (Figure 6). This shows there is a trade-off where higher discharges increase the amount of water which increases the wetted area and the shallow zones that are suitable for the Ayu, but also creates high flow velocities that limit Ayu suitability at other locations.

Figure 5. Habitat loss and habitat gain maps for two distinct scenarios compared to the current operating regime. On the left scenario Q2000wl47, that generates the largest habitat gain upstream and on the right scenario Q5000wl50, that generates the largest habitat loss downstream.

Figure 6  Relative water gain (left) and exceedance of the flow velocity thresholds for Ayu (right) of two distinct flushing scenarios compared to the current operating regime.

Discussion

Results show that most sediment flushing scenarios result in an increase of potential spawning and feeding habitat for Ayu upstream of the dam and a habitat decrease downstream of the dam compared to the current operating regime. This is caused by an increased flooded area due to higher discharges which leads to more shallow and sheltered zones upstream. However, when discharges increase, the flow velocity downstream exceeds the suitability threshold at many locations, which then become unsuitable. Ayu is a migratory fish that migrates from the sea up to the river to spawn. Due to the lack of a fish passage in the Funagira dam, the increased habitat suitability upstream can most likely not be utilized by the Ayu because it cannot pass the dam to reach these areas. If this is the case, then Ayu will be negatively affected by most sediment flushing scenarios due to the decrease in habitat downstream of the dam, except for scenarios Q1000wl47 and Q1000wl48, that show an increase in downstream habitat area. 

Due to the limited amount of parameters and the relatively short running time of the hydro-morphodynamic simulations, this habitat suitability assessment could only give a very general indication of the potential effects relative to the current operating regime. In reservoirs and dam impacted systems water quality parameters like temperature, suspended sediment and dissolved oxygen will also affect the habitat suitability of different Ayu life stages. Therefore it is important to include response curves of these parameters for several Ayu life stages (eggs, larvae and juveniles) in the assessment. Moreover, the flushing scenarios represent the situations at peak flows. The water depths and flow velocities during the low flows could generate very different habitat suitability results. Also, fish life history strategies are strongly linked to flow dynamics. Therefore, the hydro-morphodynamic model simulations should capture these long term flow dynamics. Finally, to build confidence in the habitat suitability models as predictive tools for Ayu habitat and to refine the response curves, model results need to be compared to field observations of Ayu occurrence and abundance under various flow and water quality circumstances. 

Ayu is an important commercial fish and was therefore chosen as a flagship species for the habitat assessment. However, different species can show distinct responses to changes in flow magnitude, flow timing and water quality. To evaluate the broad ecosystem response to reservoir flushing, the effect on keystone species inhabiting other types of habitats (e.g. riparian vegetation) should be included as well.

Conclusions and recommendations

• Most sediment flushing scenarios show an increase of potential spawning and feeding habitat for Ayu upstream and a decrease downstream of Funagira dam relative to the current dam operating regime.
• These results should only be considered as a broad indication of the effects of changing hydro-morphological conditions.
• The habitat assessment of Ayu can be improved by:

–           Adding water quality response curves (Temperature, DO, SS).

–           Considering more Ayu life stages (eggs, larvae and juveniles).

–           Increasing the hydro-morphological simulation time by running several years and capturing seasonal discharge dynamics.

–           Validating the habitat predictions with field observations.

• To evaluate the ecosystem wide response, keystone species with other habitat types should be included in the assessment.