Next Article in Journal
Nanoscale Zero-Valent Iron and Chitosan Functionalized Eichhornia crassipes Biochar for Efficient Hexavalent Chromium Removal
Next Article in Special Issue
Analysis of Potential Risks Associated with Urban Stormwater Quality for Managed Aquifer Recharge
Previous Article in Journal
Burden of Disease in Coastal Areas of South Korea: An Assessment Using Health Insurance Claim Data
Previous Article in Special Issue
Supporting Sustainable Development of Water Resources: A Social Welfare Maximization Game Model

Int. J. Environ. Res. Public Health 2019, 16(17), 3045; https://doi.org/10.3390/ijerph16173045

Article
The Ecological Water Demand of Schizothorax in Tibet Based on Habitat Area and Connectivity
State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu 610065, China
*
Author to whom correspondence should be addressed.
Received: 12 June 2019 / Accepted: 16 August 2019 / Published: 22 August 2019

Abstract

:
Water resource regulation is convenient for humans, but also changes river hydrology and affects aquatic ecosystems. This study combined a field investigation and two-dimensional hydrodynamic model (MIKE21) to simulate the hydrodynamic distribution from 1 March to 30 April of 2008–2013 and establish the HDI (habitat depth suitability index) and HVI (habitat velocity suitability index) based on static hydraulic conditions at typical points. Additionally, by using MIKE21 to simulate the hydraulic state in the study area under 20 flow conditions from 530–1060 m3/s, and combining these states with the HCI (habitat cover type suitability index), HDI, and HVI, we simulated the WUA (weighted usable area) and habitat connectivity under different runoff regulation scenarios to study the water requirements of Schizothorax during the spawning period in the Yanni wetland. The results showed the following: (1) the suitable cover type was cobble and rock substrate, with nearby sandy land; furthermore, the suitable water depth was 0.5–1.5 m, and the suitable velocity was 0.1–0.9 m/s. (2) Using the proximity index to analyse the connectivity of suitable habitats, the range of ecological discharge determined by the WUA and connectivity was 424–1060 m/s. (3) Habitat quality was divided into three levels to detail the flow demand further. When the flow was 424–530 m3/s or 848–1060 m3/s, the WUA and connectivity generally met the requirements under natural conditions. When the flow was 530–636 m3/s or 742–848 m3/s, the WUA and connectivity were in a good state. When the flow was 636–742 m3/s, the WUA and connectivity were in the best state. This study complements existing research on the suitability of Schizothorax habitat in Tibet, and introduces the connectivity index to enrich the method for calculating ecological water demand, providing a reference for resource regulation and the protection of aquatic organisms.
Keywords:
connectivity; ecological flow; habitat; Schizothorax; spawning period

1. Introduction

The intensity of the exploitation and utilisation of water resources increases with continuous economic development [1]. The runoff regulation alleviates the frequency of flood disasters and the uneven distribution of water resources but also has a strong negative impact on the ecological environment. Reservoirs change downstream runoff processes, which directly or indirectly affect habitat quality and change the structure, composition, and distribution of the biological community [2,3]. A lack of ecological discharge causes the lower reaches to dry up, reduce the number of species, and deteriorate the ecological environment. For example, construction of the Aswan Dam on the Nile River in Africa accelerated coastal erosion and salinisation [4], and different discharge rates of the Glen Canyon Dam in the United States affected the spawning and hatching rate of rainbow trout [5].
The Yarlung Zangbo River originates from the Jemayangzong Glacier in the Himalayas in the southwest of Tibet, with high altitude, cold temperatures, strong radiation, and other extremes, resulting in a unique and fragile ecosystem [6]. The annual average runoff of this river in China is 166 billion m3, with an uneven distribution throughout the year. As water resource regulation is imperative, it is urgent to study the ecological water demand in Tibet to ease the conflict between the development and ecological sustainability.
Fish represent the apex community in many aquatic ecosystems and are an important aquatic biological resource [7]. Dynamic changes in the number and population structure of fish can comprehensively and effectively reflect the overall changes in aquatic organisms and water quality status [8]. Thus, studies commonly use fish as a proxy to analyse aquatic ecological water demands. Tibetan Schizothorax is of great scientific value for studying topographic changes, evolution, and extreme environmental adaptation mechanisms on the Tibetan Plateau [9]. Schizothorax is also one of the main economic fishes in Tibet [10], with a slow growth rate, poor adaptability to environmental changes, and late sexual maturity (usually 3–6 years) [11], and is mainly distributed in the Yarlung Zangbo River. The spawning period of Schizothorax is 3–4 months, with a peak in March [10,12].
Currently, based on a global review, environmental flow methodologies could be classified as follows [13]: hydrological [14,15], hydraulic rating [16,17], habitat simulation [18,19], and holistic [20]. The habitat simulation method, including IFIM (Instream Flow Incremental Methodology) and CASIMIR (Computer-Aided Simulation Model for Instream Flow Requirements In Diverted Stream), associates the physical habitat (e.g., depth and velocity) with biological response indicators (e.g., oviposition and fertilization). The resultant outputs of the habitat simulation models, mainly including the PHABSIM (physical habit simulation model) [21], the RHYHABSIM (river hydraulics and habitat simulation system) [22], and River-2D [23], are often depicted as effective habitat time series and duration curves and used to quantify EFRs (environmental flow requirements) and evaluate alternative flow regulation scenarios [13].
Hydraulic characteristics are the main factors defining fish habitat and the biological composition [24]. Choosing suitable hydraulic parameters as indicators to characterise spawning habitat requirements is important for ensuring that the habitat quality simulated by environmental flow methodologies is accurate under different flow conditions. Scholars in China and elsewhere have used various hydraulic parameters to describe the characteristics of fish spawning grounds, including water depth [25], velocity [25,26], the Froude number [25,27], the kinetic energy gradient [27,28], and vortex intensity [29]. Appropriate water depth can create space for fish to move within and provide a suitable incubation environment for viscous fish eggs, and appropriate water velocity can stimulate fish to lay eggs and affects the gonadal development of fish, which requires sufficient dissolved oxygen [30]. Therefore, we used water depth and velocity to represent hydraulic requirements and to analyse habitat suitability in this study.
The existing habitat simulation method considers the suitable area demand without considering suitable habitat size or spatial connectivity. The movements of individuals, materials, nutrients, energy, or disturbances through a landscape involve more than the boundary configuration, permeability, context, and patches arrayed in the mosaic. The probability that an organism or an ion at one location in a landscape will move to some other location is a function of the combination of patch types and boundaries that separate those locations [31]. Composite metrics, such as the WUA (weighted usable area), may not sufficiently define the conditions of the river environment. The spatial distribution of aquatic habitat and size of habitat patch are also important [32]. The relationship among habitat patches could be quantified using connectivity metrics based on the size, arrangement and hydrographic distances of habitats to complement traditional habitat measures [33].
Because of the fragile habitat and urgent need for ecological water demand research in Tibet, in this study, we determined the HSI (habitat suitability index) of the substrate and hydraulic conditions during the spawning period based on a field investigation, literature investigation, and numerical simulation. We calculated the ecological water demand using the habitat simulation method. Additionally, the connectivity calculation method from landscape ecology was employed to analyse the quality of suitable habitat and enrich the modelling of ecological water demand.

2. Materials and Methods

2.1. Study Area

The Yanni wetland is located at the confluence of the Yarlung Zangbo River and the Niyang River (Figure 1), and contains diverse habitats and rich species; the mainstream part of the wetland is an important area for Schizothorax spawning and hatching [34]. The mainstream section of the Yanni wetland (from Channa to Nuxia) was used as the study area. In the section, the length of the reach was approximately 16 km, the width was 1450–3350 m, the perennial water surface width of the river was 427–1282 m, and the maximum water depth was 5.6–16.1 m. The river is of a wide valley type, with developed terraces on both sides and beaches in the middle.
The river channel is dominated by bifurcated-compound channels and is developed with terraces on both sides (Figure 2). During the period of normal water level, there are two main steps composed of cobble and rock substrates. The river channel develops into sandy land, grassland, shrubland, and farmland on both sides. In the non-flood season, the water is mainly distributed in the main channel, and in the flood season, the flow significantly increases, flooding the beaches.

2.2. Data

According to survey results, the main fish species in the reach are Schizothorax (94.2% of the catch), including Schizothorax oconnori Lloyd (EN (endangered, [35]), 65.8%), Schizothorax macropogon Regan (EN, 26.7%), and Schizothorax watoni Regan (EN, 2.1%). The length of the fish ranged from 123 to 504 mm, and the spawning period was from March to April, with a peak in March.
Based on the daily flow data from the Nuxia hydrological station from 2000–2013, Figure 3 shows the flow frequency curves in March and April. The average flow during the spawning period was 530 m3/s (expressed below as “Q”). The flow in March was stable (600–400 m3/s), and the 90% guaranteed frequency flow was used as the lower limit of flow in the spawning period (420 m3/s, 0.8 Q). In April, the flow and its amplitude increased due to the snowmelt and rainfall.

2.3. Methods

Figure 4 shows the flowchart of the research. Based on the topographic and hydrological data, a field investigation, literature research, and satellite processing, we determined the habitat suitability curve, WUA, and habitat connectivity during the spawning period of Schizothorax; and finally determined the ecological water demand during this period.

2.3.1. Hydrodynamic Model

For rivers, where the horizontal scale is much larger than the vertical scale, changes in hydraulic parameters such as water depth and velocity in the vertical direction were much smaller than those in the horizontal direction. Two-dimensional plane models could be used to simulate the hydraulic conditions of rivers by homogenizing the three-dimensional governing equations along with the water depth. Among these models, MIKE21, developed by the DHI (Danish Hydraulic Institute), is widely used worldwide [36,37,38,39].
We selected the MIKE21 hydrodynamic model to simulate the distribution of water depth and velocity in the region based on wide cross-sections of the study area. The results were used to determine the suitable curve of hydraulic habitat during the spawning period of Schizothorax and to provide hydrodynamic data for the calculation of the WUA.
The MIKE21 model is based on three Navier–Stokes equations uniformly distributed by incompressible and Reynold’s values, and the model is subject to the Boussinesq assumption and hydrostatic pressure assumption:
(a)
Continuity equation:
h t + h u x + h v y = h S
(b)
Momentum equation:
h u ¯ t + h u ¯ 2 x + h u v ¯ y = f v ¯ h g h η x h ρ 0 p a x g h 2 2 ρ 0 ρ x + τ s x ρ 0 τ b x ρ 0   1 ρ 0 ( S x x x + S x y y ) + x ( h T x x ) + y ( h T x y ) + h u s S ,
h v ¯ t + h u v ¯ x + h v ¯ 2 y = f u ¯ h g h η x h ρ 0 p a y g h 2 2 ρ 0 ρ y + τ s y ρ 0 τ b y ρ 0   1 ρ 0 ( S y x x + S y y y ) + x ( h T x y ) + y ( h T y y ) + h v s S ,
where η is the water level; d is the static water depth; h = ηd is the total water depth; u and v are the velocity components (m/s); f = 2 ω sin φ is the Coriolis force, where ω is the rotational angular velocity of the earth and φ is the local latitude; g is the gravity (m2/s); ρ is the water density; Sxx, Sxy, and Syy are the respective radiation stress components; S is the source; us and vs are the velocity of the source, respectively; and Tij is the horizontal viscous stress term.

2.3.2. Habitat Model

(1) Habitat Suitability Index:
(a) Channel suitability index
Only by adhering to cobbles and other substrates can Schizothorax fish eggs hatch successfully and undergo short-distance reproductive migration [40]. According to previous studies, the spawning area of Schizothorax in Tibet is generally located in places where the riverbed gravel is relatively thick and is covered by pebbles or stones with small amounts of silt [10,12]. Young fish forage in slow flows on sediments, usually in areas with higher water temperatures, such as a migratory bays, wide valleys, and sinks [34].
Mark Hampton divided the cover type criteria of channels into cobble, boulder, and wood [41]. In this study, considering that increased flow may cover the floodplain, and according to the field investigation and satellite image interpretation technology, we confirmed the cover type throughout the entire riverway within the research area; the cover types included rock, cobble, silt, grassland, shrub, agricultural land, and forest.
(b) The depth and velocity suitability indices
Existing studies suggest that spawning areas contain interconnected areas of rapid and slow flows [42,43], with water depth of approximately 0.3–1.5 m [10,12]. These results do not meet the accuracy requirements of habitat simulation but provide a reference for the establishment of a hydraulic suitability index.
With a developed floodplain, rich hydraulic conditions, and scattered spawning areas, in the present study area, we used typical points to construct the HSI of hydraulic conditions, which is similar to the real requirements but also requires higher accuracy in the selection of monitoring points. Considering the results of existing hydraulic habitat research, the investigation of spawning areas in the study area, and the topographic data, the Schizothorax spawning areas are scattered throughout the study area, and most of the sites are distributed near the edge of the beach and in the middle of the beach at the junction of fast and slow currents (i.e., at the second step area in Figure 2). In the selection of typical points, we considered the following: (1) the fast flow rate in the main channel provides simulation signals for spawning and hatching. (2) The flat terrain and slow flow rate on the beach facilitate predation of young fish. (3) The water mixing at the junction of the main channel and beach is strong, which is conducive to the fertilisation of fish eggs. Therefore, we chose this junction as the typical point (shown in Figure 5). On this basis, we selected points that met the requirements for determining typical points among all cross-sectional topographic data in the study area and formed a set of typical points, the spatial distribution of which is shown in Figure 5. Subsequently, MIKE21 was used to simulate the water depth and flow velocity of the typical points from 1 March to 30 April from 2009–2013. After eliminating the hydraulic calculation results (from existing research on hydraulic habitat demand) whose water depth was not between 0.3 and 1.5 m, the number of samples at the different water depths and velocity ranges was counted, and the statistical results were normalised to establish the hydraulic suitability index during the spawning period of Schizothorax.
(2) Suitable Area
By combining the distribution of water depth and velocity simulated by MIKE21 with confirmed cover types within the study area, and according to the HSI of the cover type, depth, and velocity, we determined the CSI (comprehensive habitat suitability index) and simulated the UA (usable area) and the WUA under different runoff regulation scenarios:
C S I i = H C I i × H D I i × H V I i 3
UA = j = 1 k A j
WUA = i = 1 n ( A i × C S I i )
where HCIi, HDIi, and HVIi are the cover type suitability, depth suitability, and velocity suitability of the ith grid, respectively; k is the number of grids where CSIj > 0; Aj is the area of the jth grid where CSIj > 0; k is the number of grids; and Ai is the area of the ith grid.
Table 1 shows five habitat suitability levels, which are divided based on CSI values. As suitability increases, ecological significance increases. The total habitat was used to analyse the variation in the total WUA and connectivity and to determine the ecological flow; however, different suitable habitats were used to enrich and refine the ecological flow demand.

2.3.3. Connectivity Model

In this study, we used the PROX (proximity index) to analyse habitat connectivity. PROX describes the spatial relationships between different habitat patches [33] and distinguishes between sparse distributions of small patches and clustered distributions of large patches [44]. In a landscape, patches refer to the spatial units that are different from the surrounding environment in appearance or nature and have certain internal homogeneity, which affects biomass, species composition, and diversity. Patches of the same nature are combined into a class. In this study, patches refer to the regions with a CSI >0 and are divided into five classes according to the range of CSI value (Table 1). The patch connectivity of each spawning habitat was determined by PROXij, and each PROXij value was weighted by the area to calculate the PROXi of class i:
PROX i j = s = 1 m a i j s h i j s 2
PROX i = j = 1 n a i j A i × PROX i j
where PROXij is the proximity index of the jth patch in class i; m is the number of patches within the specified neighbourhood (r, m) of patch ij; aijs is the area (m2) of patches ijs within the specified neighbourhood (r) of patch ij; hijs is the distance (m) between the patch ij and patch ijs; PROXi is the proximity index of class i; n is the number of patches in class i; aij is the area (m2) of patches ij; and Ai is the area (m2) of class i.
The specified neighbourhood (r, m) must be identified before the PROX calculation. As the spatial distribution of aquatic habitat quality and size is an important factor affecting the quality of aquatic habitat in a 0.01 m resolution macrohabitat [32] and the channel in the study area is wide, we set the specified neighbourhood to 100 m (r = 100).
PROX ≥0 is dimensionless and is used as a comparison index. PROX = 0 indicates that there are no other patches of the same type around the patch in the specified neighbourhood. PROX increases when there are more homogeneous patches in the neighbourhood and when there is more compact distribution.

3. Results

3.1. Hydrodynamic Model

We used the data of the water boundary line measured in the study area in May 2013 to carry out parameter determination for the MIKE21 hydrodynamic model. Figure 6 shows the verification results of the model, where points 1 to 11 are water edge extraction points determined by field investigation, which are different from the typical points in the habitat model. As the water level error is 0–0.1 m, the model can well represent the water dynamics in the study area.
Figure 7 shows the distribution of water depth in the study area under the natural flow rate (1 Q): the water depth is distributed in the main channel, sub-channel, etc., while the velocity is distributed only in the main channel, as the sub-channel is mostly in a static state under the influence of the floodplain. Compared with the natural river network structure (Figure 5), the simulation results in Figure 7 were similar.

3.2. Habitat Model

3.2.1. Habitat Suitability Index

(1) Channel suitability index:
Controlled by bifurcated compound channels and terraces on both sides, the cross-section along the elevation in the channel could be divided into the first step, the second step, and land (Figure 2). The first step in the channel mainly consists of rocks, and the cover type does not satisfy the conditions for hatching; thus, the HCI is 0. The second step mainly consists of cobble and develops into sand towards the banks, the cover type is suitable for spawning and hatching; thus, the HCI is 1.
According to satellite data, the cover types of land above the water surface are sandy land, grassland, shrubland, and farmland, none of which meet the spawning requirement; thus, the HCI is 0.
(2) Hydraulic suitability index:
Based on the statistics of the typical points’ hydraulic, simulated by MIKE21, from March to April of 2009–2013, and normalisation processing, we obtained the HDI and HVI (Figure 8). The suitable depth range was 0.5–1.5 m, in which at a depth of 0.6–1.0 m, the HDI >0.6, indicating good suitability, and at 0.7–0.8 m, the HDI >0.8, indicating the best suitability for spawning. The suitable velocity range was 0.1–0.9 m/s, in which at a velocity of 0.3–0.7 m, the HVI >0.6, indicating good suitability, and at a velocity of 0.5–0.6 m, the HVI >0.8, indicating the best suitability for spawning.

3.2.2. Suitable Area

MIKE21 was used to calculate the distribution of the water depth and velocity under different runoff scenarios in the study region. Figure 9 shows the cover type, depth, velocity suitability, and CSI distributions during the spawning period under natural conditions. (a) The suitable cover type is mainly distributed in the second step along the river, with good continuity. (b) Under natural conditions, the distribution of the suitable depth areas is relatively dispersed because of the widely distributed floodplain and bifurcated river channels. (c) Under the influence of the bank-up, the water flow in shoals and grooves was mostly in a static state, and suitable velocities were obviously concentrated in the main channel. (d) Considering the suitability of the cover type, water depth, and flow velocity, the distribution of spawning areas suitable for Schizothorax decreased significantly, which was mainly affected by the cover type and water depth requirements.
The total suitability of cover type (HCItotal), depth (HDItotal), and velocity (HVItotal) and comprehensive suitability (CSItotal) in the statistical region were used to analyse the causes of suitability change in the spawning area of Schizothorax under different runoff conditions, as shown in Figure 10.
HCItotal: when the flow is 0.1–0.5 Q, the water is mainly distributed in the first step of the main channel, where the cover type suitability is 0. When the value increases from 0.6–0.8 Q as the water level rises, the submerged area expands to the second step and HCItotal increases. When the value is 0.9 Q, the water level submerges the channel, reaching the maximum HCItotal. When the value continues rising, and the water level covers the banks, of which the main substrata are sandy land and grassland, which are not suitable for Schizothorax fish reproduction, HCItotal remains unchanged.
HDItotal: when the flow is 0.1–0.5 Q, the water depth is suitable within the first step of the main channel. When the value increases from 0.6–0.8 Q, as the water level rises, the depth of the first step is too high, the depth of the second step is shallower than that of the main channel, and the HDItotal decreases slightly. When the value increases from 0.9–1.3 Q, the water level rises slowly because of the wide beach, while HDItotal is almost unchanged. When the value is higher than 1.4 Q, the water in the channel is too deep, and the suitable area is transferred from the channel to the beaches and grooves, which have better depth suitability.
HVItotal: when the flow increases to 0.9 Q, the HVItotal increases as the velocity and the cross-section area increase. Then, with the increase in flow, the water spreads to the beach, and the velocity suitability of the main channel increases slowly, while the beach flow velocity is slow and the suitability is poor; thus, the HVItotal increases slowly.
CSItotal: when the flow is 0.1–0.5 Q, CSItotal is 0 and is mainly affected by HCItotal. When the value is 0.6–0.8 Q, HCItotal and HVItotal increase, while HDItotal decreases slightly and CSItotal rises slowly. When the value is 0.9–1.3 Q, HCItotal maintains the maximum value, and HVItotal increases slowly, while HDItotal is almost unchanged. The overlapping between the categories is intensified, and CSItotal increases significantly and reaches a maximum at 1.3 Q. When the value is 1.4–2.0 Q, the suitable depth area develops towards the beach, and the overlaps between the three suitability categories decrease, causing a reduction in CSItotal.
WUAi is controlled by the ith grid area and CSI. In this model, the total WUA is significantly correlated with CSItotal, as the area of each grid is relatively uniform (Figure 11, R2 = 0.9999).
Figure 12 shows the variation in WUA, WUA(0, 0.3), WUA(0.3, 0.6), WUA(0.6, 0.8), and WUA(0.8, 1.0) with the flow change. At 0.1–0.5 Q, CSItotal = 0, and WUA = 0; thus, there is no suitable spawning area for Schizothorax. At 0.6–0.9 Q, the WUA increases as CSItotal increases slowly. At 0.9–1.1 Q, WUA increases significantly. At 1.2–1.3 Q, WUA maintains a high value and WUAmax = 258 × 104 m2 at 1.3 Q. At 1.4–2.0 Q, the WUA decreases as the CSItotal is reduced.
The WUA values corresponding to different suitability ranges were calculated, and WUA(0.8, 1.0) changed dramatically with the flow. WUA(0.8, 1.0) significantly increased from 0.2–14.5 × 104 m2 as the flow increased from 0.6–1.1 Q. Then, the value remained high from 14–19.2 × 104 m2 from 1.1–1.4 Q and was reduced to 0.1 × 104 m2 when the flow reached 2.0 Q. WUA(0.6, 0.8) significantly increased from 0.4–9.5 × 104 m2 as the flow increased from 0.6–1.1 Q; it maintained a value from 9.2–9.9 × 104 m2 from 1.1–1.4 Q and then fell to 0.3 × 104 m2 when the flow reached 2.0 Q. WUA(0.3, 0.6) increased from 0.4–14.3 × 104 m2 as the flow increased from 0.6–1.1 Q, and it slightly changed (+0.8, 1) from 1.1–1.7 Q and then decreased to 7.4 × 104 m2 at 2.0 Q. WUA(0, 0.3) maintained a lower value or even a value of 0 under all flow conditions.
According to the change in WUA, the ecological water demand of Schizothorax during the spawning period was determined: (1) the spawning peak is in March, and the 90% confidence rate is 420 m3/s (approximately 0.8Q). With WUA0.8Q = 49 × 104 m2 as the minimum habitat area required by Schizothorax during the spawning period, the WUA under different runoff regulations could meet the demand and the spawning habitat requirements are generally met; the corresponding flow range is from 0.8–2.0 Q. (2) With 60% × WUAmax as the minimum habitat area, the WUA spawning habitat area meets 60% × WUAmax and above, and the corresponding flow range is from 1.0–1.6 Q. Within this flow range, WUA(0.3, 0.6) and WUA(0.6, 0.8) are in a good state. Based on the WUA and CSI distributions, the spawning habitat is in a good state. (3) With 80% × WUAmax as the minimum habitat area, the WUA spawning habitat area meets 80% × WUAmax and above, and the corresponding flow range is from 1.1–1.4 Q. Within this flow range, WUA(0.3, 0.6) and WUA(0.6, 0.8) are in the best state. Based on the WUA and CSI distributions, the spawning habitat is in the best state.

3.3. Connectivity Model

We identified the MPS (mean patch size) and NP (number of patches) of the a, b1, b2, b3, and b4 classes and calculated the PROX(proximity) of each to analyse the habitat connectivity under different flow conditions.
The connectivity of class a reflects the overall suitable habitat distribution in the region. Figure 13 shows the following: at 0.6–0.8 Q, NPa grows quickly and MPSa is constant, but PROXa maintains a lower value (near 0) because of the scattered patches, resulting in poor connectivity. At 0.8–1.1 Q, NPa and MPSa increase, and the distance between each patch gets closer, PROXa increases slowly, the connectivity enhanced but still poor. At 1.2–1.5 Q, NPa is stable, MPSa experiences a small change, the distance decreases, and the patch distribution is dominated by a large patch cluster, thus PROXa increases significantly and remains at a high state. At this time, the habitat connectivity is great, and concentrated spawning sites might be formed. At 1.6–1.9 Q, NPa remains high but MPSa decreases and the degree of the cluster weakens; thus, PROXa shows an obvious reduction and the connectivity remains in a good state. At 2.0 Q, NPa decreases with a sparse distribution, MPSa is thin, and the degree of patch aggregation significantly declines.
Figure 14 shows the PROX and connectivity of the b1, b2, b3, and b4 classes.
B4 class: At 0.1–2.0 Q, MPSb4 barely changes, and the change in PROXb4 is mainly affected by NPb4 and the distribution. At 1.3–1.4 Q, with a large NPb4 and a small patch distance, PROXb4 and connectivity are high. At 1.0 Q, with a lower NPb4 but a more concentrated distribution or a large patch cluster, PROXb4 is also high.
B3 class: At 0.1–2.0 Q, MPSb3 barely changes. At 1.3–1.4 Q and 1.6–1.7 Q, NPb3 is small with a concentrated distribution, and at 1.2–1.5 Q, NPb3 is large with a close distance, all resulting in good PROXb3 and connectivity.
B2 class: At 0.1–2.0 Q, MPSb2 barely changes, and the change in PROXb2 is mainly affected by NPb2 and the distribution. At 1.2 Q and 1.6 Q, the two peaks of PROXb2 appear, at which the patches are clustered and exhibited good connectivity.
B1 class: Compared with the former three habitat types, NPb1 significantly decreases and is sparsely distributed, resulting in a low PROXb1 and poor connectivity.
Considering the connectivity of the five classes comprehensively, the ecological water requirements of Schizothorax during spawning were determined to be as follows: (1) At 1.2–1.5 Q, with a high PROXa and a good PROXb3 and PROXb2, the habitat connectivity is good and the state of the habitat is best. (2) At 1.6–1.9 Q, with a moderate PROXa, PROXb3, PROXb2, and PROXb1, the habitat connectivity is moderate and the state of the habitat is good. (1) Using 0.8 Q corresponding to the lower limit of PROX, at 0.8–1.1 Q and 2.0 Q, the habitat connectivity is poor and the state of the habitat is generally satisfactory.

4. Discussion

In this study, the habitat simulation method was used to calculate the ecological water requirements in the main stream section of the Yanni wetland in Tibet. The rationality of the habitat suitability curve directly affects the simulation results of the habitat suitability and affects the scientific determination of the ecological discharge. The existing research results for the suitable habitat of Schizothorax in Tibet include the following: the spawning area is located in the boundary zone of severe slow current where the eggs are to be laid, the fish eggs need to attach to pebbles, and the water depth range is 0.3–1.5 m. The results of the research confirmed that the spawning ground of the fish should be composed of cobble and rock with sandy land nearby and provided the screening conditions allowing us to establish the hydraulic habitat suitability curve by using the numerical simulation method to ensure the accuracy of the habitat simulation.
In contrast to mountainous, straight river reaches, as the floodplain develops through the research area, scattered spawning grounds and a great variation of the hydraulic conditions along cross-sections arose. Therefore, we selected typical points in the study area to construct the hydraulic suability index. Rapid flow stimulates spawning in Schizothorax, slow flow helps young fish forage, and the violent water mixing in the boundary zone between rapid and slow flow is conducive to the fertilisation of fish eggs; therefore, we chose the junction of the main channel and the beach trough as the typical area and established a total of eight typical points along the river.
Using the measured hydrological data, topographic data and water boundary data, the MIKE21 hydraulic dynamic model was established and verified, and used to simulate the dynamic water distribution of typical points from March 1 to April 30 of 2008–2013 in order to calculate the HSI. Based on the hydrodynamic model results, the suitable depth range for Schizothorax in Tibet is 0.5–1.5 m (depth = 0.6–1.0 m, HDI ≥0.6), while the suitable velocity range is 0.1–0.9 m/s (velocity = 0.3–0.7 m/s, HVI ≥0.6, which differs from previous estimates of the spawning hydraulic habitat requirements of Schizothorax in a southwestern mountainous area (depth: 0.5–1.5 m; velocity: 0.5–2.0 m/s [45]). The main reasons for the difference are as follows: due to the influence of topographic factors, the study area has a wide river channel, flat terrain, developed terraces, and floodplains, and therefore, a slow flow rate. The southwestern mountainous areas are mostly located in canyon areas, with steep slopes, and the cross-section is of the “V” or “U” type. The floodplain does not develop; thus, the flow rate is fast. The establishment of the HDI and HVI enriched the research on the HSI of Schizothorax in Tibet, providing a reference for simulating ecological water requirements.
Based on the satellite data, topographic data, field sampling and HCI, the distribution and suitability of cover types in the study area were determined. Additionally, MIKE21 was used to simulate the hydraulic distribution under different runoff conditions to simulate the CSI according to the HDI, HVI, and HCI. Under all regulation scenarios, the water depth is distributed in the main channel, sub-channel, etc., while velocity is distributed in only the main channel, as the sub-channel is mostly in a static state under the influence of the floodplain. Thus, CSI primarily distributes in the main channel. At 0.1–0.5 Q, the water level is within the first step, and the cover type is unsuitable, CSI mainly affected by HCI and is 0; at 0.6–0.8 Q, HCI and HVI increase, while HDI decreases slightly and CSI rises slowly; at 0.9–1.3 Q, HCI maintains the maximum value, HVI increases slowly, HDI is almost unchanged, but the overlapping between the categories is intensified; thus, CSI increases significantly and reaches a maximum at 1.3 Q; at 1.4–2.0 Q, the suitable depth area develops towards the beach, and the overlaps between three suitability categories decrease, causing a reduction in CSI.
As the area of each grid is relatively uniform, WUA is significantly correlated with CSI. Taking the WUA corresponding to the 90% confidence flow rate in the peak spawning period (March) as the lower limit, the appropriate flow range for the runoff regulation is 80–200% × Q (average annual spawning period flow). Due to the wide cross-section, developed floodplain, and changeable terrain, the flow range suitable for spawning is wide when the hydrodynamic conditions are close to the natural state at low flow (0.8–1.0 Q), while at high flow (1.0–2.0 Q), the water level rises slowly as the flow increases and floods the bank on both sides. In contrast to the application of a habitat simulation method to simulate the minimum ecological discharge of rivers [46,47,48], as the runoff regulation is affected by dispatching, meteorology, and other factors, and could be a maximum or minimum flow in extreme cases, we proposed the upper and lower limits of ecological discharge in this study.
Under the influence of topography, the location of scattered spawning grounds changes under different discharge conditions, and their spatial relationship may affect the spawning of Schizothorax. Therefore, the connectivity index in the landscape ecology was introduced in this study, and the suitable habitat connectivity under different discharge scenarios was analysed to determine the range of ecological discharge: 80–200% × Q (average annual spawning period flow). The suitable flow range determined by the connectivity and the WUA are the same, but the habitat qualities of the two methods are different at different flow intervals (shown in Table 2).
The WUA and UA connectivity are considered to determine the ecological water demand of Schizothorax during the spawning period. According to the differences in the CSI, the suitability of spawning area was divided into four grades, and the habitat status was further divided into three types according to the corresponding WUA and connectivity of each grade—best state, good state, and generally satisfactory state (Table 2)—providing a reference for reasonably determining the dispatching flow.
For the first time, this study proposes to analyse the ecological water requirement of Schizothorax from the perspective of the WUA and habitat connectivity, which enriches the habitat simulation factors. Although the results revealed that the WUA and connectivity were consistent in the range of ecological flow, they were inconsistent in the range of flow corresponding to different habitat states. As the spatial distribution of the habitat is very important in the determination of the ecological flow, this method could be extended to calculate the water demand of various fishes. However, when calculating the connectivity, the value of the neighbourhood is key, as it is related to the habitats of fish. It is necessary to further study these conditions for different fish species.

5. Conclusions

In this study, we used a hydrodynamic model, habitat model, and connectivity model to calculate the HSI during the spawning of Schizothorax in the Yanni wetland and to predict the WUA and connectivity under different runoff regulation scenarios:
(1)
The model results indicated that the suitable spawning habitat of Schizothorax is cobble with nearby sandy land. Additionally, the suitable water depth is 0.5–1.5 m, and the suitable velocity is 0.1–0.9 m/s. The results enrich the research on the HSI of Schizothorax in the Brahmaputra River.
(2)
When the runoff regulation flow was from 424–1060 m3/s, the WUA and connectivity satisfied the requirement for spawning under natural conditions with different habitat states.
(3)
When the runoff was from 424–530 m3/s or 848–1060 m3/s, the habitat quality generally satisfied the requirements for spawning. When the runoff was from 530–636 m3/s or 742–848 m3/s, the habitat was in a good state. When the runoff was from 636–742 m3/s, the habitat was in a best state.

Author Contributions

Conceptualization—Z.Z., Y.L.; Methodology—Z.Z.; software, Z.Z.; Formal analysis—Z.Z.; Investigation—Y.L.; Data curation—Y.D.; Writing—original draft preparation—Z.Z.; Writing—review and editing—R.A.

Funding

This research was funded by the National Key Project for Research and Development Plan (2016YFC0401709).

Acknowledgments

Thanks to the Nuxia hydrologic station for providing valuable monitoring data.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Brown, C.M.; Lund, J.R.; Cai, X.; Reed, P.M. The future of water resources systems analysis: Toward a scientific framework for sustainable water management. Water Resour. Res. 2015, 51, 6110–6124. [Google Scholar] [CrossRef]
  2. Gerten, D.; Hoff, H.; Rockström, J.; Jägermeyr, J. Towards a revised planetary boundary for consumptive freshwater use: Role of environmental flow requirements. Curr. Opin. Environ. Sustain. 2013, 5, 551–558. [Google Scholar] [CrossRef]
  3. Wang, B.; Shao, D.G.; Mu, G.L.; Wang, Z.M. An eco-functional classification for environmental flow assessment in the Pearl River Basin in Guangdong, China. Sci. China Technol. Sci. 2016, 59, 265–275. [Google Scholar] [CrossRef]
  4. Abd-El, M.H.; Smith, S.E.; Darwish, K. Impacts of the Aswan High Dam After 50 Years. Water Resour. Manag. 2015, 29, 1873–1885. [Google Scholar] [CrossRef]
  5. Avery, L.A.; Korman, J.; Persons, W.R. Effects of Increased Discharge on Spawning and Age-0 Recruitment of Rainbow Trout in the Colorado River at Lees Ferry, Arizona. N. Am. J. Fish. Manag. 2015, 35, 671–680. [Google Scholar] [CrossRef]
  6. Liu, Z.; Yao, Z.; Huang, H.; Wu, S. Land use and climate changes and their impacts on runoff in the Yarlung Zangbo river basin, China. Land Degrad. Dev. 2014, 25, 203–215. [Google Scholar] [CrossRef]
  7. Mcowen, C.J.; Cheung, W.; Rykaczewski, R.R.; Watson, R. Is fisheries production within Large Marine Ecosystems determined by bottom-up or top-down forcing? Fish Fish. 2015, 16, 623–632. [Google Scholar] [CrossRef]
  8. Wang, B. The Research on the Method of Optimal Operation of Cascade Reservoirs Based on Suitable Habitats for River Fishes; Wuhan University: Wuhan, China, 2016. [Google Scholar]
  9. Wang, J.L.; Mou, Z.B.; Wang, Q.L. Research Progress on Schizothoracinae Fishes in Tibet. Anhui Agric. Sci. 2018, 46, 16–19. [Google Scholar]
  10. Zhou, X.J. Study on the Biology and Population Dynamics of Schizothorax waltoni; Huazhong Agricultural University: Wuhan, China, 2014. [Google Scholar]
  11. Ye, F.L.; Zhang, J.D. Fish Ecology; Guangdong Higher Education Press: Guangzhou, China, 2002. [Google Scholar]
  12. Ma, B.S. Study on the Biology and Population Dynamics of Schizothorax o’connori; Huazhong Agricultural University: Wuhan, China, 2011. [Google Scholar]
  13. Tharme, R.E. A global perspective on environmental flow assessment: Emerging trends in the development and application of environmental flow methodologies for rivers. River Res. Appl. 2003, 19, 397–441. [Google Scholar]
  14. Reiser, D.W.; Wesche, T.A.; Estes, C. Status of Instream Flow Legislation and Practices in North America. Fisheries 1989, 14, 22–29. [Google Scholar] [CrossRef]
  15. Zabet, S. A Comparison of 7Q10 Low Flow between Rural and Urban Watersheds in Eastern United States. Master’s Thesis, University of Tennessee, Knoxville, TN, USA, 2012. [Google Scholar]
  16. Jha, R.; Sharma, K.D.; Singh, V.P. Critical appraisal of methods for the assessment of environmental flows and their application in two river systems of India. Ksce J. Civ. Eng. 2008, 12, 213–219. [Google Scholar] [CrossRef]
  17. Liu, C.; Zhao, C.; Xia, J.; Sun, C.; Wang, R.; Liu, T. An instream ecological flow method for data-scarce regulated rivers. J. Hydrol. (Amst.) 2011, 398, 17–25. [Google Scholar] [CrossRef]
  18. Barrett, M.P.J. An Evaluation of the Instream Flow Incremental Methodology (IFIM). J. Ariz.-Nev. Acad. Sci. 1992, 24–25, 75–77. [Google Scholar]
  19. Leclerc, M.; Boudreault, A.; Bechara, T.A.; Corfa, G. Two-Dimensional Hydrodynamic Modeling: A Neglected Tool in the Instream Flow Incremental Methodology. Trans. Am. Fish. Soc. 1995, 124, 645–662. [Google Scholar] [CrossRef]
  20. Petts, G.E. Water allocation to protect river ecosystems. Regul. Rivers Res. Manag. 1996, 12, 353–365. [Google Scholar] [CrossRef]
  21. Gore, J.A.; Crawford, D.J.; Addison, D.S. An analysis of artificial riffles and enhancement of benthic community diversity by physical habitat simulation (PHABSIM) and direct observation. River Res. Appl. 2015, 14, 69–77. [Google Scholar] [CrossRef]
  22. Nicolas, L.; Jowett, I.G. Generalized instream habitat models. Can. J. Fish. Aquat. Sci. 2005, 62, 7–14. [Google Scholar]
  23. Gallagher, S.P. Use of Two Dimensional Hydrodynamic Modelling to Evaluate Channel Rehabilitation in the Trinity River, California; USAUS Fish and Wildlife Service, Arcata Fish and Wildlife Office: Arcata, CA, USA, 1999. [Google Scholar]
  24. Lamouroux, N.; Oliver, J.M.; Persat, H.; Pouilly, M. Predicting Community Characteristics from Habitat Conditions: Fluvial Fish and Hydraulics. Freshw. Biol. 1999, 42, 275–299. [Google Scholar] [CrossRef]
  25. Moir, H.J.; Soulsby, C.; Youngson, A. Hydraulic and sedimentary characteristics of habitat utilized by Atlantic salmon for spawning in the Girnock Burn, Scotland. Fish. Manag. Ecol. 2010, 5, 241–254. [Google Scholar] [CrossRef]
  26. Wang, Y.K.; Xia, Z.Q. Three-dimensional Hydraulics Characteristics of Chinese Sturgeon Spawning Site in the Yangtze River. J. Sichuan Univ. (Eng. Sci. Ed.) 2010, 42, 14–19. [Google Scholar]
  27. Li, J.; Xia, Z.Q.; Wang, Y.K.; Zheng, Q. Study on River Morphology and Flow Characteristics of Four Major ChineseCarps Spawning Grounds in the Middle Reach of the Yangtze River. J. Sichuan Univ. (Eng. Sci. Ed.) 2010, 42, 63–70. [Google Scholar]
  28. Crowder, D.W.; Diplas, P. Evaluating spatially explicit metrics of stream energy gradients using hydrodynamic model simulations. Can. J. Fish. Aquat. Sci. 2000, 57, 1497–1507. [Google Scholar] [CrossRef]
  29. Crowder, D.W.; Diplas, P. Vorticity and circulation: Spatial metrics for evaluating flow complexity in stream habitats. Can. J. Fish. Aquat. Sci. 2002, 59, 633–645. [Google Scholar] [CrossRef]
  30. Sang, L.H.; Chen, X.Q.; Huang, W. Evolution of environmental flow methodologies for rivers. Adv. Water Sci. 2006, 17, 754–760. [Google Scholar]
  31. Wiens, J.A. Riverine landscapes: Taking landscape ecology into the water. Freshw. Biol. 2010, 47, 501–515. [Google Scholar] [CrossRef]
  32. Carnie, R.; Tonina, D.; Mckean, J.A.; Isaak, D.J. Habitat connectivity as a metric for aquatic microhabitat quality: Application to Chinook salmon spawning habitat. Ecohydrology 2015, 9, 982–994. [Google Scholar] [CrossRef]
  33. Le Pichon, C.; Gorges, G.; Baudry, J.; Goreaud, F.; Boët, P. Spatial metrics and methods for riverscapes: Quantifying variability in riverine fish habitat patterns. Environmetrics 2009, 20, 512–526. [Google Scholar] [CrossRef]
  34. Zhu, T.B.; Chen, L.; Yang, D.G.; Ma, B.; Li, L. Distribution and habitat character of Schizothoracine fishes in the middle Yarlung Zangbo river. Chin. J. Ecol. 2017, 36, 2817–2823. [Google Scholar]
  35. Wang, S.; Jie, Y. China Species Red List; Higher Education Press: Beijing, China, 2004. [Google Scholar]
  36. Zhu, C.J.; Liang, Q.; Yan, F.; Hao, W.L. Reduction of Waste Water in Erhai Lake Based on MIKE21 Hydrodynamic and Water Quality Model. Sci. World J. 2013, 2013. [Google Scholar] [CrossRef]
  37. Chubarenko, I.; Tchepikova, I. Modelling of man-made contribution to salinity increase into the Vistula Lagoon (Baltic Sea). Ecol. Model. 2001, 138, 87–100. [Google Scholar] [CrossRef]
  38. Xu, M.J.; Yu, L.; Zhao, Y.W.; Li, M. The Simulation of Shallow Reservoir Eutrophication Based on MIKE21: A Case Study of Douhe Reservoir in North China. Procedia Environ. Sci. 2012, 13, 1975–1988. [Google Scholar] [CrossRef]
  39. Duong, T.A.; Long, P.H.; Minh, D.B.; Peter, R. Simulating Future Flows and Salinity Intrusion Using Combined One- and Two-Dimensional Hydrodynamic Modelling—The Case of Hau River, Vietnamese Mekong Delta. Water 2018, 10, 897. [Google Scholar]
  40. Lin, J.Q.; Peng, Q.D.; Huang, Z.L. Review on hydraulics research of fish eggs’ movement in rivers. J. Hydraul. Eng. 2015, 46, 869–876. [Google Scholar]
  41. Hampton, H. Development of Habitat Preference Criteria for Anadromous Salmonids of the Trinity River; US Fish & Wildlife Service, Division of Ecological Services: Sacramento, CA, USA, 1988; 93p.
  42. Ding, R.H. The Fishes of Sichuan, China; Sichuan Publishing House of Science and Technology: Chengdu, China, 1994. [Google Scholar]
  43. Cao, W.X.; Chang, J.B.; Qian, Y.; Duan, Z.H. Fish Resource of Early Life History Stages in Yangtze River; China Waterpower Press: Beijing, China, 2007. [Google Scholar]
  44. Gustafson, E.J.; Parker, G.R. Using an index of habitat patch proximity for landscape design. Landsc. Urban Plan. 1994, 29, 117–130. [Google Scholar] [CrossRef]
  45. Zhang, Z.G.; Tan, Q.L.; Zhong, Z.G.; Jin, Y.; Du, J. study on ecological flow regime based on habitat requirement of fish. Water Power 2016, 42, 13–17. [Google Scholar]
  46. Gard, M. Comparison of spawning habitat predictions of PHABSIM and River2D models. Int. J. River Basin Manag. 2009, 7, 55–71. [Google Scholar] [CrossRef]
  47. Fu, J.J.; Huang, B.; Rui, J.L.; Tan, S.K.; Zhao, S. Application of Habitat Simulation to Fishery Habitat Protection in Heishui River. J. Hydroecol. 2016, 37, 70–75. [Google Scholar]
  48. Yang, Z.F.; Yu, S.W.; Chen, H.; She, D.X. Model for defining environmental flow thresholds of spring flood period using abrupt habitat change analysis. Adv. Water Sci. 2010, 21, 567–574. [Google Scholar]
Figure 1. (a) Location of the Yanni wetland in the Yarlung Zangbo River. (b) The Yanni wetland river network (data from January 2018 Landsat eight satellite images).
Figure 1. (a) Location of the Yanni wetland in the Yarlung Zangbo River. (b) The Yanni wetland river network (data from January 2018 Landsat eight satellite images).
Ijerph 16 03045 g001
Figure 2. Channel morphology of Section A, which is typical of the study area.
Figure 2. Channel morphology of Section A, which is typical of the study area.
Ijerph 16 03045 g002
Figure 3. Measured daily discharge during the spawning period from 2000–2013, and the frequency curves simulated by a Pearson type-III curve.
Figure 3. Measured daily discharge during the spawning period from 2000–2013, and the frequency curves simulated by a Pearson type-III curve.
Ijerph 16 03045 g003
Figure 4. Flowchart for calculating the ecological water demand in the study area.
Figure 4. Flowchart for calculating the ecological water demand in the study area.
Ijerph 16 03045 g004
Figure 5. (a) shows the extraction position of a typical point in the cross section; (b) shows the distribution of typical points along channels, including eight points; (c) shows the cover type near the bank.
Figure 5. (a) shows the extraction position of a typical point in the cross section; (b) shows the distribution of typical points along channels, including eight points; (c) shows the cover type near the bank.
Ijerph 16 03045 g005
Figure 6. Contrasting the measured water level and simulated value to confirm the model parameter, in which 1–11 are the points shown in Figure 7.
Figure 6. Contrasting the measured water level and simulated value to confirm the model parameter, in which 1–11 are the points shown in Figure 7.
Ijerph 16 03045 g006
Figure 7. Under natural conditions (1 Q), (a) shows the distribution diagram of the water level, and (b) shows the flow rate.
Figure 7. Under natural conditions (1 Q), (a) shows the distribution diagram of the water level, and (b) shows the flow rate.
Ijerph 16 03045 g007
Figure 8. (a) shows the sample numbers used for the water depth statistics and the HDI (habitat depth suitability index); (b) shows the sample numbers used for the velocity statistics and the HVI (habitat velocity suitability index).
Figure 8. (a) shows the sample numbers used for the water depth statistics and the HDI (habitat depth suitability index); (b) shows the sample numbers used for the velocity statistics and the HVI (habitat velocity suitability index).
Ijerph 16 03045 g008
Figure 9. Picture of habitat suitability index distribution under natural conditions including habitat cover type suitability index (a); the habitat water depth suitability index (b); the habitat velocity suitability index (c); and the comprehensive habitat suitability index (d).
Figure 9. Picture of habitat suitability index distribution under natural conditions including habitat cover type suitability index (a); the habitat water depth suitability index (b); the habitat velocity suitability index (c); and the comprehensive habitat suitability index (d).
Ijerph 16 03045 g009
Figure 10. The Variation in HCItotal (sum of the habitat cover type suitability index of the study area), HDItotal (sum of the depth suitability index of the study area), HVItotal (sum of the habitat velocity suitability index of the study area), and CSItotal (sum of the habitat comprehensive habitat suitability index of the study area) under different runoff regulation scenarios.
Figure 10. The Variation in HCItotal (sum of the habitat cover type suitability index of the study area), HDItotal (sum of the depth suitability index of the study area), HVItotal (sum of the habitat velocity suitability index of the study area), and CSItotal (sum of the habitat comprehensive habitat suitability index of the study area) under different runoff regulation scenarios.
Ijerph 16 03045 g010
Figure 11. WUA relevance to CSItotal.
Figure 11. WUA relevance to CSItotal.
Ijerph 16 03045 g011
Figure 12. The Variation in the WUA (weighted usable area) under different runoff regulation scenarios.
Figure 12. The Variation in the WUA (weighted usable area) under different runoff regulation scenarios.
Ijerph 16 03045 g012
Figure 13. Variation in MPS (mean patch size), NP (number of patches), and PROX(proximity) of class a with runoff regulation.
Figure 13. Variation in MPS (mean patch size), NP (number of patches), and PROX(proximity) of class a with runoff regulation.
Ijerph 16 03045 g013
Figure 14. Variation in MPS, NP, and PROX of classes b4 (a), b3 (b), b2 (c), and b1 (d) with runoff regulation.
Figure 14. Variation in MPS, NP, and PROX of classes b4 (a), b3 (b), b2 (c), and b1 (d) with runoff regulation.
Ijerph 16 03045 g014
Table 1. Categories of suitable area.
Table 1. Categories of suitable area.
ClassCSI 1 RangeSuitability Level
a1.0 > CSI > 0Total habitat, including all suitable areas
b10.3 > CSI > 0 Poor habitat
b20.6 > CSI > 0.3Intermediate habitat
b30.8 > CSI > 0.6Good habitat
b41.0 > CSI > 0.8Best habitat
1 CSI—comprehensive habitat suitability index.
Table 2. Ecological water demand of Schizothorax during the spawning period.
Table 2. Ecological water demand of Schizothorax during the spawning period.
×Q 1Habitat StateCorresponding Flow (m3/s)
WUA 20.8–1Generally satisfactory530≥; ≥424
1–1.1Good state583>; ≥530
1.1–1.4Best state742>; ≥583
1.4–1.6Good state848>; ≥742
1.6–2.0Generally satisfactory1060>; ≥848
Connectivity0.8–1.1Generally satisfactory583≥; ≥424
1.1–1.2Good state636>; ≥583
1.2–1.5Best state795≥; ≥636
1.5–1.9Good state1007≥; >795
1.9–2.0Generally satisfactory1060≥; >1007
Overall
consideration
0.8–1.0Generally satisfactory530≥; ≥424
1.0–1.2Good state636>; ≥530
1.2–1.4Best state742>; ≥636
1.4–1.6Good state848>; ≥742
1.6–2.0Generally satisfactory1060≥; ≥848
1 The average flow during the spawning period, Q = 530 m3/s; 2 WUA—weighted usable area.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Back to TopTop