Development and application of a hybrid model to analyze spatial distribution of macroinvertebrates under flow regulation in the Lijiang River
Highlights
► A hybrid model for macroinvertebrates presence is developed. ► Quantitatively assess impact of reservoir operation on macroinvertebrates dynamic. ► Case study showed increase of discharge in dry season is negative to S. amurensis. ► The hybrid approach is applicable for other aquatic organisms.
Introduction
Modification of natural flow regime for human consumption, agricultural production and power generation has dramatic effects on aquatic biota distribution due to altering water quality and habitat connectivity, disrupting sediment transport and nutrient cycling, and blocking fish and invertebrate movements (Dynesius and Nilsson, 1994, Santucci et al., 2005). In the last decades, great efforts have been made to better understand the mechanisms between the spatial structure of aquatic species and the related hydro-environmental conditions. The effects of small experimental floods on the differential mortality of native and exotic riparian trees have been studied on the Bill Williams River (Mouton et al., 2007, Shafroth et al., 2010), and an integrated ecohydraulics model was developed to simulate and assess ecological effects of physical habitat changes at four different flow rates in the Zwalm River in northern Belgium (Mouton et al., 2007). Relationships between colonially-nesting water birds reproductive performance and flow variability were quantified in the Murray River in south-eastern Australia (Leslie, 2001).
Macroinvertebrates play a central role in river ecosystems. They uptake detritus and algae and provide food to water-bird and fish. Great attentions had been paid to the dynamics of macroinvertebrate. One important aspect is to understand the relationships between their distribution and flow regimes. Zigler et al. (2008) used physical and hydraulic factors to predict the distribution of freshwater mussels in a reach of the upper Mississippi River and suggested that seasonal droughts and floods were important in structuring mussel distributions. Stubbington et al. (2009) presented a conceptual model of the processes influencing benthic and hyporheic invertebrates under low-flow conditions and highlighted the potential importance of surface water and hyporheic zone linkages to riverine invertebrate communities under a range of flow conditions. James and Suren (2009) assessed the impacts of low flows on invertebrates through six experimental channels located in a lowland stream in the Kaiapoi River in New Zealand and concluded that invertebrates were resistant to the experimental flow reduction. Plenty of studies on macroinvertebrate community under different rivers showed that flow regulation had great effects on macroinvertebrate structure and distribution (Cortes et al., 2002, Dewson et al., 2007, Dunbar et al., 2010, Grzybkowska et al., 1996, Lind et al., 2007, McKay and King, 2006, Sheldon and Thoms, 2006).
However, these studies are mainly qualitative rather than quantitative. In recent years, with a growing awareness of the value of natural ecosystems, there are strong demands to quantitatively investigate other than qualitatively assess the impacts of flow regulation, so as to seek for practical remediation measures (Nagaya et al., 2008, Tollefson, 2008). Merigoux et al. (2009) combined the hydraulic preferences for 66 invertebrate taxa with a statistical habitat model (FSTress) to predict the effects of minimum flow restoration on invertebrate abundances in two bypassed sections of the Rhone River and expected that the overall invertebrate diversity was to increase.
This research took a compound channel of the Lijiang River as the study case, and aimed at developing a hybrid macroinvertebrate habitat model to quantitatively assess the impact of upstream reservoir operation on the distribution of macroinvertebrates. The model integrated a two-dimensional water quality module with a species presence module. The integrated model simulated water temperature, dissolved oxygen, water depth, flow velocity, and then predicted consequentially the presence of S. amurensis. Field survey records collected during 2009–2010 were used for model calibration and verification. Both realistic and hypothesized scenarios were analyzed and compared to provide suggestions to adjust the flow regulation for ecological concerns.
Section snippets
Study area
The study focused on a compound channel in the middle reach of Lijiang River, which is located in the southwest China (Fig. 1). Due to the special Karst landscape and the strong seasonality of rainfall, the discharges in the channel (gauged at the Yangshuo hydrological station) vary from 12 m3/s to 8000 m3/s, and the annual average is about 120 m3/s. During dry season from October to next March, there is a serious problem of water-deficiency, which imposes great threats to the local water supply
Model developments
The developed model consisted of a two-dimensional water quality module and a macroinvertebrate presence module based on artificial neural networks (ANN) approach. The water quality module provides the external driving forces to the macroinvertebrates module.
Results of scenario analyses
The validated hybrid model was used to analyze the distribution of S. amurensis under the designed discharge of 60 m3/s in dry season. Fig. 4 presented the simulated flow velocity, water depth, WT and DO. Compared to the current condition of 28 m3/s, the averaged water depth and flow velocity have an obvious increase when discharge is 60 m3/s. Due to the increase of water level, the shallow area drops from 5.12 × 105 m2 to 4.69 × 105 m2. Regarding water temperature and dissolved oxygen, the difference
Discussions
This paper developed a hybrid model that integrates a two-dimensional water quality module and an ANN based macroinvertebrate presence module. The model aimed at quantitatively investigating the impacts of reservoir operations on the downstream macroinvertebrate dynamics.
The developed model was applied to a compound reach in the middle of Lijiang River, where the flow is seriously altered by the upstream Qingshitan Reservoir. Five factors were used to construct the S. amurensis presence model.
Acknowledgments
The authors were grateful to the funding from National Basic Research Program 973 (No. 2010CB429004), National Natural Science Foundation of China (No. 50879086) and the Chutian Scholarship (KJ2010B002).
References (52)
- et al.
Comparison of modelling techniques to predict macroinvertebrate community composition in rivers of Ethiopia
Ecological Informatics
(2010) - et al.
An evaluation of three statistical methods used to model resource selection
Ecological Modelling
(2010) - et al.
Effects of cell size and neighborhood type in a cellular automata based prey predator model
Simulation Modelling Practice and Theory
(2003) - et al.
Integration of data mining techniques and heuristic knowledge in fuzzy logic modelling of eutrophication in Taihu Lake
Ecological Modelling
(2003) - et al.
Optimization of Artificial Neural Network (ANN) model design for prediction of macroinvertebrates in the Zwalm river basin (Flanders, Belgium)
Ecological Modelling
(2004) - et al.
The potential for computer-aided identification in biodiversity research
Trends in Ecology & Evolution
(1995) - et al.
A heuristic approach to predicting water beetle diversity in temporary and fluctuating waters
Ecological Modelling
(2010) - et al.
Multilayer feedforward networks are universal approximators
Neural Networks
(1989) - et al.
Predicting fish yield of African lakes using neural networks
Ecological Modelling
(1999) - et al.
Individual-based modelling of fish population dynamics in the river downstream under flow regulation
Ecological Informatics
(2010)
Effects of local and regional landscape characteristics on wildlife distribution across managed forests
Forest Ecology and Management
Fish habitat modelling as a tool for river management
Ecological Engineering
Evaluation of suitable hydraulic conditions for spawning of ayu with horizontal 2D numerical simulation and PHABSIM
Ecological Modelling
Illuminating the “black box”: a randomization approach for understanding variable contributions in artificial neural networks
Ecological Modelling
Generalizability of artificial neural network models in ecological applications: Predicting nest occurrence and breeding success of the red-winged blackbird Agelaius phoeniceus
Ecological Modelling
Developing an empirical model of phytoplankton primary production: a neural network case study
Ecological Modelling
Modelling the riparian vegetation evolution due to flow regulation of Lijiang River by unstructured cellular automata
Ecological Informatics
Seasonal dynamics of zooplankton in a shallow eutrophic, man-made hyposaline lake in Delhi (India): role of environmental factors
Hydrobiologia
Effects of water-depth and water-level fluctuations on the macroinvertebrate community structure in the littoral zone of Lake Constance
Hydrobiologia
The distribution of macroinvertebrate assemblages in a reach of the River Allier (France), in relation to riverbed characteristics
Hydrobiologia
Hydraulic microhabitats and the distribution of macroinvertebrate assemblages in riffles
Freshwater Biology
A robust fuzzy logic approach to modelling algae biomass
Journal of Hydraulic Research
Hydroinformatics techniques in eco-environmental modelling and management
Journal of Hydroinformatics
Cellular-automata-based ecological and ecohydraulics modelling
Journal of Hydroinformatics
Non-linear approach to grouping, dynamics and organizational informatics of benthic macroinvertebrate communities in stream by artificial neural networks
Macroinvertebrate community structure in a regulated river segment with different flow conditions
River Research and Applications
Cited by (14)
Multi-objective ecological reservoir operation based on water quality response models and improved genetic algorithm: A case study in Three Gorges Reservoir, China
2014, Engineering Applications of Artificial IntelligenceCitation Excerpt :For example, water recreational activities cut down the space for aquatic organisms (Jiang et al., 2006a, 2006b; Rohdea et al., 2006); water transfer projects change the spatial and temporal pattern of regional water resources (Zeilhofer and Rubem, 2009;); and reservoirs aggravate the degradation of water quality and the spread of water-borne diseases (Mukand et al., 2011; Lu et al., 2012). Considering these changes, modern reservoir management demands a new paradigm, which integrates ecological components into its management decisions (Camdevyren et al., 2005; Suen and Eheart, 2006; Cabecinha et al., 2009; Chen et al., 2011). Ecological reservoir operation involves complex decision making processes with multiple decision variables and objectives as well as considerable risk (Mehrdad et al., 2010).
Comparison of habitat suitability models using different habitat suitability evaluation methods
2014, Ecological EngineeringCitation Excerpt :In contrast to the preference curve method, the fuzzy logic method uses statements such as “low”, “medium”, and “high” to verbally describe input variables and mathematical methods for data processing. This method can solve two problems effectively: nonlinearity and correlation between variables, which cannot be solved using a preference function (Marsili-Libelli et al., 2013; Chen et al., 2011). In the 1990s, the Computer-Aided Simulation Model for Instream Flow Requirements (CASiMiR), which is based on fuzzy logic, was developed.
Protocol for developing ANN models and its application to the assessment of the quality of the ANN model development process in drinking water quality modelling
2014, Environmental Modelling and SoftwareSpring micro-distribution of macroinvertebrate in relation to hydro-environmental factors in the lijiang river, china
2013, Journal of Hydro-Environment ResearchCitation Excerpt :The relative importance of these hydro-environmental factors varied not only with the region, but also with the spatial scale (Park et al., 2003, 2007; Song et al., 2007). Meanwhile, models to patterning and predicting macroinvertebrates dynamics were gradually established by using artificial neural networks (Park et al., 2001, 2003), data-driven method (Boets et al., 2010; Bayih et al., 2010), fuzzy logic (Adriaenssens et al., 2004), flexible asymmetrical approach (Dufrene and Legendre, 1997) and hybrid approach (D'Heygere et al., 2003; Chen et al., 2011b). However, knowledge to macroinvertebrate distribution and the governing factors at an entire river is scarce due to the scale of studied area and the complexity of spatial variation (Arab et al., 2004; Principe and Corigliano, 2006; Park et al., 2007; Xiao et al., 2010).
Optimizing the operation of the Qingshitan Reservoir in the Lijiang River for multiple human interests and quasi-natural flow maintenance
2012, Journal of Environmental Sciences (China)