Estimating Abundance
Reading: Chapter 10
– Survey design
– Visual censuses
– Acoustic methods
– Trawl surveys
– Depletion estimates
– Mark-recapture estimates
– Egg Production Methods
– Fishery-dependent CPUE
Estimating Abundance
Why do we need to estimate abundance?
To estimate:
1. Stock size
2. Recruitment
3. Mortality
4. Spatial distribution
Estimating Abundance
Survey design
– A central problem is obtaining an abundance
index that is proportional to stock size
– Well-designed survey should provide estimates of:
•
•
average fish abundance or density and
Spatial distribution (survey boundaries?)
– Accuracy vs. Precision
1
Accuracy
Precision
Estimating Abundance
Survey design
– A central problem is obtaining an abundance
index that is proportional to stock size
– Well-designed survey should provide estimates of:
•
•
average fish abundance or density and
Spatial distribution (survey boundaries?)
– Accuracy vs. Precision
– Bias vs. Variance
– ↑ precision (↓ error) = ↑ $
Sample error vs. sample size
110
Sample Error (%)
100
90
80
70
60
50
40
5
10
15
20
25
30
35
Sample size (n)
2
Estimating Abundance
Survey design
–
–
–
–
–
Stratification by habitat type or depth
Combine abundance estimates across strata
Increases precision
Systematic vs. Random sampling
Systematic can be more precise and generally
reduces costs
Estimating Abundance
Visual censuses
¾
¾
¾
¾
¾
¾
Require clear, shallow waters
Best with non-cryptic fish that don’t avoid divers
Can see fish and habitat
Transects most common
Point counts (timed or instantaneous)
Behavior
3
4
Estimating Abundance
Acoustics
¾ Use of sound waves to detect fish (swim bladder)
¾ Best for pelagic fishes
¾ Target strength is species-specific and must be
determined experimentally
¾ Simultaneous trawling to ‘ground-truth’ catch
¾ Problems with acoustic shadows and avoidance
¾ Very promising for well understood pelagic stocks
Estimating Abundance
Depletion (or Removal) estimates
¾ Relation between abundance and catch rate
¾ Requires:
¾ Closed population
¾ Short fishing period (no recruitment)
¾ Catchability proportional to abundance
CPUE (C/f) = qNt
N t = N0 – K t
CPUE (C/f) = qN0 – qKt
Plot CPUE vs. cumulative catch (K) (known as Leslie method)
5
CPUE
“Leslie Method”
Slope = -q
Estimate of N0
Kt
Consecutive sweeps with 100ft. seine (Fall 2003)
45
40
30
25
20
15
10
5
0
1st haul
2nd haul
3rd haul
Depletion estimates of abundance (Fall 2003)
50
25
Mullet
40
20
30
15
CPUE
CPUE
Pinfish
20
10
10
5
0
0
40
50
60
70
80
90
20
100
25
30
35
40
Cumulative Catch
Cumulative Catch
10
14
Spot
12
Shrimp
8
10
8
CPUE
CPUE
Number captured
35
6
6
4
4
2
2
0
0
5
10
15
20
Cumulative Catch
25
5
10
15
20
25
Cumulative Catch
6
60ft seine pulled inside 100ft seine (Fall 2003)
25
Mullet
Spot
Pinfish
Blue crab
Number captured
20
15
10
5
0
1st haul
2nd haul
3rd haul
4th haul
Fort Fisher Field trip 2004
Depletion estimation using 100ft. seine
140
Pinfish
Numbers captured
120
Mojarra
Atl silverside
100
Ladyfish
Total fish
80
60
40
20
0
1st haul
2nd haul
3rd haul
Seine haul
Depletion estimates of abundance
(Fall 2004)
40
Pinfish
100
Mojarra
30
CPUE
CPUE
80
60
20
40
10
20
0
0
70
80
90
100
110
120
130
0
50
100
K
Atl. silverside
200
All species
180
160
15
140
CPUE
CPUE
150
K
200
20
10
120
100
80
60
5
40
20
0
0
0
10
20
30
K
40
50
75
100
125
150
175
200
225
250
275
300
K
7
Depletion estimates of abundance (Fall 2005)
600
80
Pinfish
Atl. silverside
70
60
400
CPUE (# per haul)
CPUE (# per haul)
500
300
200
100
50
40
30
20
0
10
-100
0
500
520
540
560
580
70
80
Cumulative catch (K)
90
1000
All species
Atl. croaker
800
CPUE (# per haul)
8
CPUE (# per haul)
100
Cumulative catch (K)
10
6
4
2
600
400
200
0
0
-200
7.8
8.0
8.2
8.4
8.6
8.8
9.0
9.2
9.4
850
900
Cumulative catch (K)
950
1000
1050
Cumulative catch (K)
Size-selectivity of beach seines (Fall 2005)
60
Relative frequency (%)
20 ft seine
50
n = 71
40
30
20
10
0
0
20
40
60
80
100
120
140
Total Length (mm)
16
Relative frequency (%)
14
60 ft seine
12
n = 210
10
8
6
4
2
0
0
20
40
60
80
100
120
140
Total Length (mm)
Estimating Abundance
Depletion (or Removal) estimates
¾ DeLury Method
CPUE (C/f) = qNt
CPUE (C/f) = qN0(Nt/N0)
ln CPUE = ln qN0 + ln (Nt/N0)
Substitute Nt/N0 = e-qE
ln CPUE = ln qN0 –qE
Plot ln CPUE vs. cumulative effort (E)
8
DeLury Method
ln CPUE
y-int. = ln qN0
slope = -q
Cumulative effort (E)
Estimating Abundance
Trawl surveys
¾ Very widely used, most common
¾ Mesh size regulates fish size
¾ Constant catchability (q) essential; lack of
standardization is major problem
¾ Consistent gear design, tow speed, duration help
to maintain q
C = qfN
CPUE = qD
Stock biomass = D x area
Estimating Abundance
Trawl surveys
¾ Many factors affect catchability (q)
¾
¾
¾
¾
Tow speed
Depth
Time of day
Vessel noise
¾ Mostly, q is unknown, but…..
¾ If q is constant, then estimated stock biomass will
be proportional to actual stock size
9
Estimating Abundance
Mark-recapture methods
¾ Successful in terrestrial and freshwater systems
¾ Can also provide growth and movement data
¾ Assume:
¾
¾
¾
¾
Tagged fish mix randomly with untagged fish
Catchability equal
No tag loss or mortality due to tagging
Relatively closed population
T/N = R/C
so, N = TC/R
Estimating Abundance
Egg production estimates
¾
¾
¾
¾
Provide estimate of size of spawning stock
Used for large pelagic fish stocks
Annual method for determinate spawners
Daily method for indeterminate spawners
Prod = BiomassRatioFecundity
so, B = P/RF
¾ Need to account for atresia, mortality, age
10
Annual method
Surveys
Daily method
Estimating Abundance
What’s wrong with using CPUE from fishery?
¾ It provides catch and effort data from large areas
over long time scales, so why not use it?
¾ Often times it is used, only data available
¾ Landings data omits discards (bycatch, undersize)
¾ Catch/effort data hard to get for every boat
¾ CPUE (LPUE) rarely proportional to abundance
¾ No gear standardization
¾ Capture efficiency increases with time
¾ Fishers don’t fish randomly
11
Fig. 10.15. Spatial
distribution
of commercial
trawling effort
(hours per year)
in the North Sea
Fig. 4.16. Distribution
of Atlantic cod in the
Gulf of St. Lawrence,
showing range expansion
and contraction over
twenty years
Fig. 4.17. Occurrence of
low, medium, and high
catches of Atlantic cod
in research vessel
surveys as the
fishery collapsed
12
Fig. 4.18. How the
calculation of mean
catch rate can affect
the interpretation of
fishery trends, example
from northern cod
9
Catchability Coefficient
8
7
6
5
4
3
2
1
0
0
1000
2000
3000
4000
Biomass (Tons)
Abundance
13
Density
Hyperaggregation
Hypoaggregation
Abundance
Catch per Effort (CPUE)
CPUE remains high due to aggregation of fish
Abundance
14