Assessing Fish Passage Success in Culvert Structures with the Development of a Two-Dimensional Algorithm Considering Physical Capabilities of Juvenile Salmonids ALEXANDRA WEST JEFFERIES, P.E. A THESIS M.S. CIVIL ENGINEERING UNIVERSITY OF ALASKA ANCHORAGE MARCH 2017
Special Thanks Gillian O Doherty (Alaska Department of Fish & Game) William Rice, PE (U.S. Fish & Wildlife Service) My thesis committee Tom Ravens, PhD Rob Lang, PhD William Rice, PE
Purpose & Background Fish Passage Barriers Habitat fragmentation Law [33 CFR 323.4 (a)(6)(vii)] [Alaska Statute 16.05.841] Stream Simulation Design Preferred, but not always feasible due to size or budget limitations Modeling FISHPASS (Power & Energy) FishXing (Velocity)
Purpose & Background - FishXing 1-dimensional Compares against average cross-sectional velocity Tests passage of specific species and life stage through culvert Models and calculates water profiles for different types of culverts and uses gradually varied flow equations. Hydraulic output (average velocity and normal depth) is then compared against the swimming availabilities of the species of interest Velocity reduction factors allow input of decreased velocity to better simulate occupied velocity
Occupied Velocity Furniss et al., 2006
Occupied Velocity Q
Objectives 1. Develop and test a two-dimensional algorithm to assess fish passage movement and success through culverts based on velocity. 2. Compare FishXing and the two-dimensional algorithm to actual passage results.
River2D 2-dimensional Models transient conditions Developed for the assessment of fish habitat based on the USGS Physical Habitat Simulation (PHABSIM) Assesses habitat suitability (not passage) Input mesh Topography Ice Output water depths & velocities across a grid or at nodes Can dump grid or nodes into.csv format
River2D River2D Velocity Output for 14.1 cfs (legend in m/s)
River2D River2D Depth Output for 14.1 cfs (legend in m)
Site Characteristics ADF&G study site Buddy Creek Largest tributary to Montana Creek Near Talkeetna Studied species coho, chinook, & rainbows Had two separate conditions Culvert array (monitored 2013-2014) Arch pipe (replaced in 2015 and monitored through end of 2016)
Site Characteristics continued ADF&G-provided data included: Passive Integrated Transponder (PIT) tag data on juvenile coho, chinook, and rainbow trout between 65 mm and 120 mm from 2013-2016 Buddy Creek summer gage data (water surface elevations, temperatures, and discharges) downstream of the crossing Likely time of vertical stream adjustments
PIT Tagging Culvert Array, 2014 After Arch Construction, 2015
Methodology Overview 1. Collected survey data in 2014 (before arch installation), 2015 (after arch installation), and 2016 (after stream adjustments) 2. Modeled culvert array and arch hydraulics FishXing (both conditions) River2D (arch only) 3. Validated hydraulic models with actual data Discharge, point velocities, depths in structure 4. Used output from River2D and physical capabilities of fish to create a 2D fish passage algorithm 5. Tested fish passage algorithm against actual fish passage data at known discharges Fish passage data provided by ADF&G from PIT tag data 6. Tested FishXing for culvert array and arch against actual fish passage data
Data Collection ADF&G fish passage study 2013-2016 Stream gage Point velocities & depths PIT tagging 65-120 mm juvenile salmonids Species: Chinook, coho, and rainbow trout Date of tagging Size at tagging Date/time of passage Qmin, Qmax, Qaverage Time between d/s and u/s passage
Data Collection Multiple stream surveys were required due to vertical stream adjustment after arch installation. A) Summer 2015 B) Summer 2016
Fish Swimming Abilities Cruising Speed the speed a fish can maintain for an extended period of travel without fatigue. Metabolic activity in this mode is strictly aerobic and utilizes only red muscle tissues Prolonged (Sustained) Speed the speed that a fish can maintain for a prolonged period, but which ultimately results in fatigue. Metabolic activity in this mode is mixed anaerobic and aerobic and utilizes some white muscle tissue and possibly red muscle tissues Burst (Darting) Speed the speed a fish can maintain for a very short period, generally 5-7 seconds, without gross variation in performance. Burst speed is employed for feeding escape, ad negotiating difficult hydraulic situations, and represents maximum swimming speed. Metabolic activity in this mode is strictly anaerobic and utilizes all of the white muscle tissues
Fish Swimming Abilities Species Modeled Swim Type V (m/s) t (s) Length of Fish (mm) Water Temp ( C) Reference Coho & Chinook Prolonged 5.87L.7 t -.1 1800 40-133 18-20 Hunter & Mayor, 1986 Burst 0.8 7.5 55 No data Gubernick Rainbow Trout Prolonged 0.8 1200-22.6 Burst 12.03L.62 t -.51 5-7-19 Jones et al., 1974 Hunter & Mayor, 1986
2D Fish Passage Algorithm No Keep swimming Burst Swim swim t swim> t burst Yes Yes No BARRIER V>V p V>V burst No No Keep swimming Yes BARRIER (velocity) Keep swimming t swim> t p Yes Swim in burst? If-then statements Input: Values along a 3 grid: Depth from River2D V from River2D Select coho/chinook or rainbow trout & size Output: Pass or Fail
2D Fish Passage Algorithm Input: sample from River2D grid dump (5 cfs) & results: Assesses specific fish abilities and determines if each grid is passable If at least one path exists - Successful FAIL
Results Passage events used in models: Structure Culvert Array Arch Species Coho & Chinook Rainbow Trout Coho & Chinook Rainbow Trout No. of Size Range Flow Range (cfs) Passage (mm) Events Min Max Min Max 121 65 110 6 72 17 94 134 11 56 291 65 124 3 91 58 70 95 3 38
Results - Primary Objective Comparing Algorithm Results to PIT Tag Data 55% congruency for coho/chinook 100% congruency for rainbow trout Combined velocity/depth barrier at higher flows except at higher flows when fish were able to swim in or directly over riprap at edges (higher roughness).
Results-Secondary Objective FishXing Culvert Array Results vs. PIT Tag Data Favored middle culvert 2% congruency for coho/chinook 94% congruency for rainbow trout Difficult to model accurately above results may not mean much FishXing Arch Results vs. PIT Tag Data 55% congruency for coho/chinook 100% congruency for rainbow trout Approximately the same results as the 2D algorithm
Results Comparison 100% Congruency to Passage Data 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% <70 70 80 90 100 110 120 Rounded Fish Size (mm) FishXing Algorithm 200 Successful Passage Events 180 160 140 120 100 80 60 40 20 FishXing Algorithm PIT Tag 0 <70 70 80 90 100 110 120 Rounded Fish Size (mm)
Conclusions The 2d algorithm and FishXing (1D) passed the same fish at approximately the same flows Velocity variations in 2D more accurately represents the inside of a passage structure and the occupied velocity FishXing velocity reduction factors fairly accurately predict occupied velocity Current modeling techniques result in conservative design Hydraulics (with use of velocity reduction factors) not the issue Clear need for better understanding of juvenile salmon swimming behavior
Limitations Algorithm grid spacing Algorithm only developed for one structure Limited juvenile salmon swimming ability data Lack of overall swimming data Specific lack of burst speed data Swimming abilities depend on water temperatures Exact size of fish
Recommendations Use Stream Simulation design whenever possible When use FishXing, use good engineering judgement Areas of further study: FishXing velocity reduction factors Fish swimming abilities More data Burst data for juvenile salmon (none found in published literature) Field data (most data is from laboratory studies forced velocity)
Questions?