National Conference on Ecosystem Restoration Los Angeles, CA July 20-24, 2009 Models Quantify the Relationship Between Water Flows/Levels and Ecological Endpoints Joseph V. DePinto, Todd M. Redder, Scott Bell, Laura Weintraub LimnoTech (www.limno.com) Ann Arbor, MI
Presentation Outline Background: two initiatives that recognize importance of hydrology/hydraulics to ecosystem structure and function ESWM framework of The Nature Conservancy Great Lakes Water Compact Previous Flow/Level Ecological Response Modeling Lake Ontario St. Lawrence River Regulation Evaluation (IERM) Muskegon River Watershed linked flow ecological response modeling (GLECO) Conceptual Approach for Sacramento San Joaquin Bay Delta System
Ecological Sustainable Water Management - The Nature Conservancy (from Richter, et al., 2003) Goal: Meet human needs for water by storing and diverting water in a manner that sustains ecological integrity of aquatic ecosystems
Great Lakes Water Compact Signed by all Great Lakes States Separate Agreement with Canada (Ontario, Quebec) Goal: Protect, conserve, restore, improve and effectively manage the Waters and Water Dependent Natural Resources of the Basin Specifically: prevent significant adverse impacts of water Withdrawals and Losses on the Basin's ecosystems and watersheds
Lake Ontario St. Lawrence River Water Level/Flow Regulation Study Goal: Evaluate existing regulation plan and recommend alternative plan that best satisfies the needs of multiple interests: environment, riparian landowners, hydropower, commercial navigation, recreational boating, water supplies) Lower St. Lawrence River Moses Saunders Dam (Plan 1958DD) Trois-Rivieres Quebec Montreal Canada U.S. Toronto Lake Ontario Rochester Upper St. Lawrence River
Integrated Ecological Response Model (IERM) Designed to compare response of Ecological Performance Indicators (PI s) to alternative Hydrologic/Hydraulic (H&H) conditions Compare alternative Regulation Plans under a given Basin Supply Scenario Other stressors assumed constant for comparison Composed of sub-models for each PI group PI response algorithms only as complex as data will allow Range of complexity from simple empirical relationships (PI vs water level function) to more complex processoriented population sub-models Work with researchers to integrate the science Build Conceptual model: understand data availability & connections between various studies (2002-04) Identify specific performance indicators and associated metrics Evaluate and interpret the model
Reduce Analysis to 32 Key PIs SAR (4) Vegetation (1) Mammal (1) Herptiles (0) Lake Ontario / Upper St. Lawrence Key PIs (19) Birds (2) Fish (11) Vegetation (0) Fish (3) Lower St. Lawrence Key PIs (13) SAR (4) Key PIs based on: Representativeness/significance Certainty Sensitivity to regulation Geographic coverage Mammal (1) Herptiles (1) Birds (4)
IERM Conceptual Model (Lake Ontario) Lake Ontario Water Level Weekly WL time series Nest access Water Temperature Wetland access Stranding potential Frequency of lowwater years Frequency of highwater years Nest flooding Nearshore temperature Wetland temperature Timing of spawning events Weighted usable area for various life stages Wetland Habitat Weighted usable area Meadow marsh area Cattail area Floating leaf area Cattail usage Wetland Birds Suitable habitat area (acres) Nesting success Fish (multiple species) Year-class strength (no./yr) Biomass (kg/ha) Production (no./ha/yr) Muskrats Muskrat houses per acre Habitat loss due to flooding/stranding Weighted usable area Weighted usable area Endangered Species Suitable habitat area Amphibians/Reptiles Suitable habitat area WL fluctuations Flood magnitude/duration
Wetland Plant Sub-Model (LO/USL) Sub-Mode el Inputs LO/USL Water Level Time Series Flooding elevations inundated for 4 consecutive QM during growing season Dewatering elevations dry during entire growing season Typical Wetland Topography Barrier Beach Drowned River Mouth Protected Embayment Unprotected Embayment LO/USL Wetland Area Barrier Beach Wetland Plant PI Measures % Species Composition @ Specific Elevations Barrier Beach Drowned River Mouth Protected Embayment Unprotected Embayment Total Estimated Area of Plant Species (ha) Barrier Beach Drowned River Mouth Protected Embayment Unprotected Embayment Feed to Faunal Sub-Models Feed to Faunal Sub-Models Sub-M Model Outputs Drowned River Mouth Protected Embayment Unprotected Embayment Wetland Plant Effects
IERM PI Time Series Diagram
IERM Target Diagram
Plan A Net # of PIs w/ Sign nificant Gains 20 15 10 5 0-5 IERM Plan Evaluation Results (Lake Ontario / Upper River 19 PIs) Historical (1900-2000) Stochastic #1 - Wettest Century Stochastic #2 - Driest Century Stochastic #3 - Like Historical Stochastic #4 - Longest Drought -10 Plan A Plan D Plan B PreProject
Plan A Net # of PIs w/ Sign nificant Gains 20 15 10 5 0-5 IERM Plan Evaluation Results (Entire LOSL System 32 PIs) Historical (1900-2000) Stochastic #1 - Wettest Century Stochastic #2 - Driest Century Stochastic #3 - Like Historical Stochastic #4 - Longest Drought -10 Plan A Plan D Plan B PreProject
Addressing Great Lakes Withdrawal Issues Great Lakes Watershed Ecosystem Model (GLECO) Stressors Channel Modifications Other Management Actions & System Stressors Natural Hydrological & Climatological Forcings Water Use Drainage Basin Properties Flow Response Groundwater Flow Regime River Channel Flow Regime Watershed Runoff Quantity Watershed Runoff Quality Nutrient loads Solids loads Assessment Indicators Fish Population Riparian Wetland Vegetation Plant diversity Fish Habitat Weighted usable habitat area Temperature River Hydraulics Flow/Volume Depth Velocity Water Quality Nutrients Solids
GLECO development and application to Muskegon Watershed (MI) Configure Hydrologic Simulation Program FORTRAN (HSPF) to watershed based on geomorphic, hydrologic, land use/cover data Link HSPF to required ecological sub-models Calibrate flow and water quality parameters at various sampling locations Establish baseline condition for relevant ecological endpoints. Run model forecasts of withdrawal scenarios & compare to baseline
Model Scenario Application: Little Muskegon River Scenario A Single withdrawal Withdrawal of 5 MGD from the Little Muskegon River catchment #1 Scenario B : Cumulative withdrawal Withdrawal of 5 MGD from the Little Muskegon River catchment #1 & catchment #2 (10 mgd total) Withdrawals modify channel hydraulics and water temperature Withdrawals reduce suitable spawning habitat area for brown trout by 20-50%
Brown Trout: Spawning Habitat Suitability Functions Brown trout spawning habitat is impacted by water temperature, depth, and stream velocity. Based on U.S. Fish & 0.2 Wildlife Service brown trout 0.0 habitat report (1986). Habitat Suitability Index (HSI) 1.0 0.8 0.6 0.4 0.0 0 10 20 30 40 50 60 70 80 Water Temperature (deg. F) Habitat Suitability Index (HSI) 1.0 0.8 0.6 0.4 0.2 0.0 0.0 0.5 1.0 1.5 2.0 Stream Depth (feet) Habitat Suitability Index (HSI) 1.0 0.8 0.6 0.4 0.2 0.0 0.0 1.0 2.0 3.0 4.0 Stream Velocity (ft/sec)
Withdrawal Scenarios for Little Muskegon River sub-basin Model generates geographic comparisons of withdrawal scenarios (red, green) relative to baseline (blue) for spawning habitat (average annual weighted suitable area): Muskegon River GW: 5 mgd GW: 5 mgd GW: 5 mgd Little Muskegon River
Comparison of Habitat Results Model generates temporal comparisons of scenario simulation results relative to baseline:
Sacramento San Joaquin Bay Delta Integrated Ecological Response Model: Concepts The Bay Delta system is managed both in terms of water quantity and water quality. There are multiple stakeholders with varying priorities Reclamation and other management and project decisions must consider ecological impacts Reclamation Project Dam Canal New flow mgmt. plan Modification of System Conditions Flow volume Velocity Temperature Salinity Ecosystem Receptor Fish Waterfowl Wetland habitat 20
Bay Delta Integrated Ecological Response Model: Challenges Must provide for protection and recovery of endangered and sensitive species, as well as protection and restoration of water supplies Multiple projects to consider, each may affect multiple environmental drivers Multiple ecological receptors: delta smelt, Chinook salmon, green sturgeon, riparian wetlands, migratory waterfowl, etc. Cumulative impacts may exist Must overlay management on natural hydrometeorological conditions (climate change?) 21
System Stressors Hydrology SJR System Operations Tributary boundary inflows, diversions, return flows DSM2-SJR Model Mathematical Mo odels Hydraulic Sub- Model (HYDRO) Predicted flow @ Vernalis Predicted flow @ Vernalis Chinook Salmon Smolt out-migration (mean flow in spring) Adult escapement (mean flow in fall) Water Quality Sub- Model (QUAL) Predicted salinity in delta area Delta Smelt Population indices: 1) SJR-Sacramento confluence 2) Suisun Bay Predicted salinity in delta area Ecological Response Sub-Models 22
Model Scenario Application: Little Muskegon River Little Muskegon River Watershed