Influence of conductivity and depth on lowfrequency underwater radio telemetry signal attenuation

1
Influence of conductivity and depth on
low-frequency underwater radio telemetry signal
attenuation
Jason G. Freund Kyle J. Hartman
West Virginia University Division of
Forestry Wildlife and Fisheries
2
Radio Telemetry Overview

• Unique frequency for each tag (individual)
• Low frequency (30-70 MHz) or
• high frequency (gt100 MHz)
• Detection of signal through an antenna and
• a receiver
• Strength of signal to determine direction
• of signal origin
• Signal transmitted better through air than
• water

3
Signal Attenuation
Definition To reduce in force, value,
or amount weaken (The American Heritage
Dictionary)
Implications The Ohio River has highly
conductive water and the main channel generally
exceeds 6m deep.
This impacts radio telemetry signal detection!
4
Necessity for Research

• Many fish found to be unaccounted for
• during a search are located in later
• search efforts.
• Previous research mentions attenuation but
• fails to address it

• Lucas and Batley (1996), Barbel in the River Ouse
• Otis and Weber (1982), Carp in Lake Winnebago
  system
• Distance of signal detection greater at 2 feet
• than at 5 feet

5
Consequences of High Signal Attenuation

• Not the entire volume of the water column
• is effectively sampled
• May account for lost and found fish

6
Signal Attenuation and Depth
Signal strength is inversely proportional to
transmitter depth
7
Signal Attenuation and Conductivity
Signal strength is inversely proportional to
water conductivity
8
Ohio River Conductivity

• Manufacturer and literature suggests using
• low frequency equipment when water
• conductivity is gt400-500 ?S.
• Main channel conductivity averages 518.3
• ?S with a range of 146.3 ?S to 886.0 ?S.
• Conductivity varies greatly with seasons, but
• overall, conductivity exceeded 500 ?S 66.0
• of all observations within the main channel.

9
Ohio River Macro-Scale Habitat Types

• Main Channel
• Typically deep and wide
• Islands and back channels

• Embayment
• shallower and not as wide
• May be bay-like

• Tributary
• Not as wide, nor typically
• as deep as the main channel

10
Experimental Design
1 m
3 m
6 m
9 m
1 distance measure
1 distance measure
1 distance measure
1 distance measure
Block of 4 transmitters
Block of 4 transmitters
1 distance measure
1 distance measure
1 distance measure
1 distance measure

• Replicated 4 times per trial, 2 trials
• Repeated measures AVOVA (8 transmitters)

11
Experimental Field Design
RV Bucketmouth
Block of 4 transmitters
12
(No Transcript)
13
(No Transcript)
14
Measuring Distances
a2 b2 c2 where a ? Longitude b ? Latitude
Boat
c
b
a
Transmitter

• Using GPS, initial and final locations are
• recorded and distance (converted to km) is
• found using simple trigonometry.

15
Statistical Method

• Repeated measures ANOVA with 2 blocks
• of 4 transmitters (total of 8 transmitters
• used over all treatments)
• Depths were assigned randomly to blocks
• Post-Hoc Test if significant ANOVA
• Tukeys Honestly Significant Difference
• ? 0.05 for all tests

16
Results of Statistical Tests

• The distance at which transmitters can be
• detected varies significantly with depth.
• Use DateDepth as the error term as it
• is out best estimate of the variance
• P 0.0020 (df 3, n 256)
• Differences between transmitters was
• significant, differences between dates
• were not.

17
Tukey HSD Multiple Comparison Results

• Mean distance of detection varies
• significantly at all depths (p lt 0.05)

18
Distances from Transmitters
Black 1 m Red 3 m Green 6 m Yellow 9 m
Location of Transmitters
19
Detection Distance as a Linear Function of Depth
20
Plot of Residuals from Linear Regression
21
Detection Distance as a Non-Linear Function of
Depth
22
Plot of Residuals from Non-linear Regression
23
Inherent Study Design Limitations
Many factors besides depth and conductivity
affect signal detection

• Location of transmitter relative to antenna peak

24
Inherent Study Design Limitations

• Maximum distance seldom applies in the field

Distance A
Distance A
25
Inherent Study Design Limitations
Many factors besides depth and conductivity
affect signal detection

• Location of transmitter relative to antenna peak
• Habitat complexity (meanders, aquatic vegetation,
  
• structure, etc.)
• Area searched per scan (scan time, travel speed,
• number of transmitters)
• transmitter size and battery power
• Surface area of the water body (lake, large
  river, small
• river, etc)

26
Implications for Current Research

• Detection of main channel fish more
• difficult, particularly in the lower pool
• Embayment, tributary, and main channel
• border fish may be biased for
• May explain lost and found fish
• Difficulty in determining an exact cause
• of signal loss (tag failure, mortality,
• dispersal, deep water, etc.)

27
Acknowledgements

• Dr. Bill Thayne
• Dr. Dean Coble
• R. C. Tipton
• West Virginia Division of Natural Resources

28
(No Transcript)