Transcript Lecture 1 topics

Lecture 2 review
• Maximizing long term harvest can generally be achieved
by following a “fixed escapement” harvest rule WHICH
BRINGS STOCK TO ITS MOST PRODUCTIVE SIZE AS
QUICKLY AS POSSIBLE THEN HOLDS STOCK AT
THAT SIZE
• Nearly the same long term harvest can be achieved by
following a “fixed exploitation rate” rule, much less
damaging to fishers
• Tactics for regulating harvest rates involve either input
(effort) or output (catch) controls
• Output controls are dangerous and require accurate
assessments of stock size
• Complex management objectives and performance
measures are an invitation to gridlock in decision making
Limits to compensatory responses
• Most populations exhibit high juvenile
survival at very low densities
SJ
(Invasive species have to
exhibit this ability)
N
• But occasionally (5-10%?) compensation
fails at low densities, leading to low
equilibrium or extinction
-Allee effect (eggs don’t get fertilized, eg scallops); rare
SJ
-Cultivation/depensation (competitors/predators of
juveniles increase when N is low, eg bass-bluegill)
N
-Trophic cascades (green water/clear water states)
-Botsford’s effect (size dependent cannibalism)
Life history trajectories
• Whenever you handle a fish, ALWAYS ask
yourself these questions:
– How old is it?
– Where was it spawned?
– Where will it spawn?
Life history stanzas (partitions of the
life
history
trajectory)
The eggie
Juvenile migration
Larval drift,
densityindependent
mortality
First juvenile nursery
area: small, strong
density-dependence
in mortality
Spread into larger
juvenile nursery
area(s), mortality
much lower
Spawning
migration
Adult foraging areas,
most often with complex
seasonal migration
patterns
Fractal, complex
diurnal
movement
Characteristics of LHT
• There is typically very strong selection for
behaviors that take fish back to spawn in
the places where they were successfully
produced (this is not just a salmon thing)
• Seasonal migrations become more
pronounced as fish grow
Distance
from
tagging
site
Random model
Time
Distance
from
tagging
site
Migration model
Time
Characteristics of LHT
• Natural mortality rates vary as M=k/(body
length), starting at a few percent per day and
often falling to a few percent per year
• Body growth typically follows a vonBertalanffy
length curve of the form
length=L[1-e-K(a-ao)]
• Sometimes there is a “kink” in the growth curve,
with small juveniles either showing extra fast
growth (if they seek warm microhabitats) or extra
slow growth (if they face very high predation
risk).
Is the Beverton-Holt invariant M/K=1.6
a valid generalization for stock
assessment?
100
y = 2.0041x
y = 2.1161x
2
2
R = 0.8567
Natural mortality rate M
R = 0.5872
10
Age structure data
y = 1.6372x
R2 = 0.3182
Length converted catch
curve
1
Tagging
Z vs E plot
y = 1.1363x
2
R = 0.2195
0.1
0.01
0.01
0.1
1
vonBertalanffy K
10
Characteristics of LHT
• Maturation typically occurs at 50%-70% of maximum
body length, with fecundity then being proportional to
body weight
Length (cm)
Hayes model Length
wt (kg)
Hayes Model weight
70
4000
3500
60
Length (cm)
2500
40
2000
30
1500
20
Weight (g)
3000
50
1000
10
500
0
0
0
2
4
6
8
10
12
Age (years)
But some fish like these New Zealand brown trout
practically stop growing at maturity, and make massive
(45%) investments in eggs (Hayes et al TAFS 2000)
Representing LHT in models
• Age structure accounting (block trajectory by
even age intervals)
[N1 N2 N3 …]t  [N1 N2 N3…]t+1 (easy in spreadsheets)
• Stanza structure accounting (Ecosim)
Weight
at age
Log
Numbers
at age
Age (months)
• Individual-based models (track movement)
X,Y positions and
fates of large
sample of
individuals
Ways to represent space in models
• Total areas by habitat class, without regard
to spatial arrangement (A1,A2,…)
• Irregular spatial areas (“polygons”)
A1
A2
• Regular spatial cells (“rasters”)
Spatial Management
Dealing with complex dynamics
Spatial management is not just
about MPAs
• Dynamic organization of shrimp fisheries:
lessons for assessment, cooperative
management
• Fishing for information: using logbook data
to understand spatial stock structure and
opportunities for more selective fishing
practices
• Methods for modeling spatial stock
dynamics
A tale of two gulfs
Spatial stock structure (Western king prawn)
Spatial life history
Nursery
Juveniles
Fishery
(Adults)
St Vincent Gulf fishery
Spencer Gulf fishery
Contrasting management regimes
• St. Vincent Gulf (collapsed)
– Ethnic fishery
– Combative participants, severe misreporting
– Assessments based on simple catch-effort
relationships
• Spencer Gulf (sustained)
– Cohesive fishing communities
– Neil Carrick: dogged persistence, many bar fights to
develop cooperative approach
– Regulatory structure based on adaptive fishing policy
(time-area closures) based on repeated surveys and
openings each year
Cooperative spatial surveys in
Spencer Gulf
Assessment modeling options
• Empirical approach: fishing for information to
map stock distribution and abundance several
times during each season, cpue-based rule for
ending annual fishery before depletion of
spawning stock
• “Mechanistic” approach: develop detailed
spatial model of physical drivers (currents,
temperature, salinity), predict prawn recruitment,
survival, movement
• (The mechanistic approach led to very costly
research and modeling, never worked)
Multispecies fisheries: tradeoffs
caused by technical interactions so
some stocks overfished at MSY
SALIX CREEK
700000
MOTASE LAKE
AZUKLOTZ CREEK
PINKUT ABOVE WEIR
PINKUT CHANNEL # 1
500000
FULTON BELOW WEIR
FULTON ABOVE WEIR
Averge Yield
Skeena River
sockeye salmon
example: many
stocks overfished
at Fmsy
PINKUT BELOW WEIR
600000
400000
FULTON CHANNEL # 2
FULTON CHANNEL # 1
300000
TWAIN CREEK
TSEZAKAWA CREEK
TAHLO CREEK
200000
TACHEK CREEK
SOCKEYE CREEK
100000
SIX-MILE CREEK
SHASS CREEK
0
PIERRE CREEK
0
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6
Overall exploitation rate
PENDELTON CREEK
NINE-MILE CREEK
MORRISON CREEK
FOUR-MILE CREEK
BABINE - UNACCOUNTED
BABINE RIVER (SECTION
4)
BABINE RIVER
Selective fishing practices to achieve
variable F targets over species
• Most common: forced discarding of
sensitive species (e.g. escape ramps for
dolphins)
• Modification of gear deployment (e.g. bait
types, set depths, mesh sizes, escape
gaps and grids)
• Selective space-time openings (e.g.
salmon)
Temporal selectivity: Skeena River
gillnet fishery example
Spatial selectivity
• Use detailed logbook and survey data to
map species distributions, identify areas of
high overlap and/or density of sensitive
species (e.g. Fishmap)
• Adaptive spatial closures based on the
mapping (e.g. Carrick’s shrimp fishery)
• Also use the mapping to develop “follyfantasy” spatial cpue indices for long-term
stock assessment
Partial separations in spatial
distributions, BC trawl fishery
Sensitive species
(longspine rockfish,
Fmsy=0.05)
Productive
species (English
sole, Fmsy=0.2)