Transcript 3 major dilemmas on small-scale fisheries management:

Perspectives and dilemmas for smallscale fisheries management in African
freshwater fisheries
Jeppe Kolding (University of Bergen)
Small-scale fisheries - A challenge for fisheries management Experiences and lessons from developing countries and Norway.
Fisheries Forum (Fiskerifaglig Forum)
4 – 5 October 2007
Background
• Inland fisheries (and small-scale fisheries
in general) are the ‘social security
system’ in Africa – A common good!
• Serves as the ‘last resort’ when
everything else fail.
 How should they be managed?
 How can they be managed?
3 major dilemmas on small-scale
fisheries management:
• Why does the common property theory
•
•
(CPT) only apply to man, and not to other
predators? – they are also harvesting a
common.
Why are we controlling effort increase (f),
while we at the same time refine and
develop the catchability (q)?
Why is man the only predator that has a
harvesting pattern completely opposite all
other predators?
Small-scale fisheries comprise:
•
•
•
•
> 30 % of total world captures
> 50 % of total landings for human consumption
> 90 % of all fishermen
≈ 80 % live in Asia
Many ecosystems only exploitable on a small-scale
• Coastal lagoons
• Tidal flats, shallow shores
• Estuaries
• Coral reefs
• Most freshwaters (contribute 25% of global production)
Marine and inland capture fisheries –
top 10 producers 2002
China, India and Indonesia
have populations of nearly 1
billion people living below
the UNDP poverty line of
US$ 1 per day (Staples et al. 2004)
(SOFIA 2004)
Importance of fish for people
The richer the less
dependent on fish
Relationship between the proportion of fish protein in human diets and the
relative wealth (measured as GPD) of the nations they live in. From Kent (1998)
Small-scale fisheries
•
•
•
•
Research generally very, very small
(mostly socio-economic)
Most fisheries biologist are dealing with
large scale industrialized fisheries
Quantitative SSF data limited or nil
Problems and management?
 copied from industrial fisheries
(the ruling paradigm)
•
•
Small-scale fisheries
Mostly associated with developing countries





traditional - antiquated - primitive
poor - needs development
unmanaged - resource depleting - challenge
overfished
“Tragedy of the commons”
Having a negative image:






Illegal and destructive gears
Ignore regulations and legislation
Unruly members of society
Subject to “Maltusian” overfishing
“Poverty trap”
Unselective, indiscriminate fishing methods
Small-scale fisheries
Thus, plenty of arguments for:
They need to be managed!
90 % of projects use one or several of the above reasons for justifications
But how to manage them? when:
• Little research = little knowledge
•
•
•
•
Multi-species and multi-gear situations
Negligible monitoring
Unwillingness to abide
Costly to enforce
But do we
understand
their best ?
Traditional fisheries management
The present answer (panacea) seems to be:
Co-management and MPAs
“They must learn to understand their own good”
Management paradigm
The present mainstream research is focused on:
• Industrial (valuable) fisheries
• Single-species considerations
•
 F, TAC, Quotas, size-limits
Enhanced selective harvesting strategies
Purpose: A selective kill on targeted species and sizes
Result: Dominate our thinking (paradigm) and forms
our perception on small-scale fisheries
•
EAF (ecosystem approach to fisheries) is only
recent on the agenda (Johannesburg 1992) and
only conceptually debated = we don’t know how!
Patterns of exploitation
Selectivity = rooted in all fisheries theory:
•
•
•
•
Mesh size regulations
Gear restrictions
By-catch
Destructive methods
seining
beat fishing
barriers, weirs
small mesh sizes
Almost universally
banned in Africa
In industrial fisheries non-selectivity = BAD
Result: Universally applied - also in co-management!
The selectivity paradigm
•
FAO 2003 (Ecosystem Approach to Fisheries):
"Selectivity, or lack of it, is central to many biological issues
affecting fisheries. Bycatch or incidental capture is
responsible for endangering and contributing to extinction of
a number of non-target species…. In addition, the discarding
of unwanted catch, which is particularly important in
unselective fisheries, is being considered by society not only
as wasteful but as unethical.
The Code of Conduct dedicates a whole section to the issue
(8.5). It promotes the use of more selective gear (7.6.9;
8.4.5) and calls for more international collaboration in better
gear development (8.5.1; 8.5.4), as well as for the agreement
on gear research standards.”
SSF management = Copy+paste
• Industrial fisheries are single species
fisheries with single species management
They are so large and valuable that research
and CMS is invested for management decisions
• Small-scale fisheries are multi-species,
multi-gear, too small to warrant research.
Our ‘understanding’ and assumptions on which
we base our management is directly inherited
from large-scale fisheries.
Q1: Why does the common property
theory (CPT) only apply to man, and
not to other predators?
In the ‘balance of nature’ it is generally assumed that:
•
•
•
•
predation is the most important factor in natural mortality
of fish (Sissenwine 1984; Vetter 1988; ICES 1988),
adaptations tend to maximize fitness through optimal
utilization of resources (Slobodkin 1974; Stearns 1976;
Maynard-Smith 1978),
predators and prey are co-evolved (Slobodkin 1974; Krebs
1985) and,
there is an uni-modal response of prey productivity to
predator densities:
(sigmoid curve theory = logistic Gordon Schaefer model)
Logistic growth –Surplus production
Max = MSY
No growth
No surplus
No stock
The rate of change in
biomass production as a
function of the biomass is
uni-modal
No growth
No surplus
Stock = Max
Logistic growth - predator-prey
•
•
From above principles it is reasonable to assume
that predation would 'maintain' prey populations
close to their highest average production rate
(Slobodkin 1961, 1968; Mertz & Wade 1976; Pauly 1979; Caddy & Csirke
1983; Carpenter et al. 1985).
The argument follows simply from the sigmoid curve
where the highest surplus production of the prey
population (dN/dt = max) is the 'carrying capacity'
(K) of the predator population.
Predator-prey
Thus predators can in theory grow to reach K (= MSYprey), but if
they overshoot they will reduce prey production and
consequently decline themselves.
This is the background for density dependent cascade theory,
and the coupled time-lagged oscillations observed between
predator and prey
Predator-prey: Cascading effects
Inverse biomass trends illustrating trophic cascades
in the Black Sea (from Daskalov 2002)
What has this to do with CPT?
• The big question is if effort is controlling the
productivity or if the productivity is controlling
the effort? Are small-scale fishermen different
from other predators?
•
•
•
The answer to this dilemma is fundamental for
applying CPT and co-management!
If we close for open access it will have severe
consequences for the ‘last resort’ option
By closing open access we are in fact, closing the
social security system of Africa!
Productivity in African lakes
•
Morphology


•
•
…of a lake, particularly area, volume, depth, and shoreline development
or gradient, is of major importance to the productivity (Ryder 1978).
The mean depth encapsulates several of these attributes and is
considered as the most important (Rawson 1952, Ryder et al. 1974, Mehner et al.
2005).
Nutrients

Lakes do not maintain fertility unless continual external loading of
nutrients is applied (Schindler 1978, Moss 1988, Karenge and Kolding 1995). Water
inflow is a major contributor and serves as a proxy for nutrient load.
Hydrology

The ‘flood pulse advantage’ is the amount by which fish yield per unit
mean water area is increased by a natural, predictable flood pulse (Bayley
1991). The ‘flood pulse’ keeps the environment in a stage of early
succession, which means that it is dominated by biota with r-selected
traits (Junk et al. 1989).
The physical basis for lake productivity
Climatic
Latitude
Altitude
Morphological
Edaphic
Hydrologic
Nutrient
loading
Size, duration
and variability
of flood pulse
Not
considered
here
Area
Depth
Volume
Generalised effects of climatic, morphological, edaphic and
hydrological factors (X-axis) on productivity (Y-axis)
Relative Lake Level Fluctuation Index (RLLF) …
mean lake level amplitude
RLLF 
100
mean depth
•
•
•
…encapsulates the morphological, edaphic and
hydrological driving forces for productivity into a
single quantity.
… is a dynamic extension of the MEI index that
only incorporated morphological and edaphic
factors
… builds on the ‘flood pulse’ concept (Junk et al. 1989)
and the ‘flood pulse advantage’ (Bayley 1991)
Hydrology and
fish yields
• Variability around
the trend of total
inland catches of the
SADC countries
show decadal
fluctuations possibly
influenced by long
term climate
variations (water
levels)
Lake levels as drivers of fish productivity
•
Lake Turkana 1972-1989
Lake Kariba 1982-1992
Kolding (1992)
Karenge and Kolding (1995)
Mean annual catch rates varies with water levels in most
African fisheries. This has long been known by local
fishermen, but not much investigated.
Data on catch, effort and water levels
•
•
17 major lakes and
reservoirs in Africa.
Monthly time series (Min #
years = 9) of lake levels from
gauge readings (N = 13) or
satellites (N = 4)


•
ESA
http://earth.esa.int/riverandlake/
TOPEX-POSEIDON
http://www.pecad.fas.usda.gov/crop
explorer/global_reservoir/
Yield and effort estimates
from 1990’s (Updated from
Jul-Larsen et al. (2003) and
various projects we have
been involved with).
Kolding and van Zwieten (2007)
Data..
Lake
Area
km2
Tanganyika
32600
580
240
73000
40000
0.04
0.14
2699
240
40
315
2868
0.06
0.14
Malawi
30800
290
545
28000
27296
0.10
0.30
Victoria
68800
40
288
571000
105000
0.60
1.10
Edward
2325
17
16031
5443
1.43
5.60
Turkana
7570
31
47
1500
1500
2.12
3.73
Kariba
5364
30
48
30311
7060
4.32
9.65
390
5.5
90
7500
2371
6.00
20.40
Volta
8500
18.8
121
250000
71861
7.02
19.49
Nasser
5248
Fluctuating
25.2
50
30000
6000
7.14
23.63
Mweru
2700
8
94
42000
15791
7.20
25.70
Bangweulu
5170
3.5
68
10900
10240
7.39
34.34
Kainji
1270
11
40
38246
17998
8.78
69.41
Rukwa
2300
3
17
9879
13.79
31.97
Chilwa
750
3
13
Chiuta
113
2.5
Itezhi-tezhi
370
15
Kivu
Malombe
Mean depth
m
Stable
#Species
Yield
ton/yr
#Fishermen
RLLF
annual
RLLF
seasonal
Highly fluctuating
15000
3485
17.80
39.70
40
1400
350
19.53
59.30
24
1200
1250
21.16
54.47
Africa – Yield (production) is highly
correlated with RLLF
Productivity (annual yield/km2) vs Seasonal RLLF
35
y = 0.42x + 3.97
R2 = 0.64
30
Volta
Kainji
25
20
Chilwa
t/km2
Malombe
15
Mweru
1
Chiuta
10
1)
2)
3)
Victoria
Edward
Nasser
Kariba
5
1
Rukwa
Tanganyika
Malawi
Kivu Turkana
0
0
5
10
Data too old for comparison (1970s)
Oligotrophic – large areas inaccessible
Unreliable records – Kapenta not incl.
3
2
Itezhi-tezhi
Bangweulu
15
20
25
30
35
40
Seasonal RLLF
45
50
55
60
65
70
Similar results from Asian
reservoirs..
8
2
Mean yield (t/km /yr)
7
y = 0.038x - 1.9
r2 = 0.65
6
5
4
3
2
1
0
0
20
40
60
80
100
120
140
160
180
200
Relative lake level fluctuations (RLLF-s)
From Kolding and van Zwieten (2006)
Relationship between mean annual yield (t/km2) and relative seasonal lake
level fluctuations (RLLF-s = %(annual draw downs/mean depth)) in 15
reservoirs of the lower Mekong countries. Data from Bernascek (1995)
Africa – Fishing effort is highly correlated
with RLLF…
# fishers/km2 vs seasonal RLLF
16
y = 0.16x + 0.83
R2 = 0.64
14
Kainji
12
fishers/km2
10
Volta
8
Victoria
Malawi
Tanganyika
Kivu
6
Malombe
Mweru
Chilwa
4
2
Edward
Itezhi-tezhi
Bangweulu
2
Kariba
1)
2)
Nasser
Turkana
1
Chiuta
Data too old for comparison (1970s)
Unreliable records
0
0
10
20
30
40
seasonal RLLF
50
60
70
80
… but catch rates are not correlated with RLLF
Catch rates vs seasonal RLLF
6
y = 0.0046x + 2.61
R2 = 0
Victoria
5
Nasser
Chilwa
Kariba
Chiuta
ton/fisher/year
4
Edward
3
Volta
Malombe
Mweru
Kainji
2
Tanganyika
Malawi
Bangweulu
1
Itezhi-tezhi
Turkana
Kivu
0
0
10
20
30
40
seasonal RLLF
Indicating…….
50
60
70
80
…effort seems self-regulating (from CPUE)
50
45
y = 3.0417x
R2 = 0.8228
Average yield per fisher is 3 ton per year
irrespective of system
Catch (t/Km^2/year)
40
35
Volta
30
Kainji
25
Chilwa
20
15
10
Malombe
Nasser & Kariba
Mweru
Chiuta
Is yield driven by effort or
is effort driven by yield?
Victoria 1990 Victoria 2002/4
Edward
5
Albert
Adapted from Jul-Larsen et al. 2003
Itezhi-tezhi
Victoria 1970/2
Bangweulu
0
Turkana
0
Kivu
Malawi
•
2 Tanganyika 4
6
8
10
12
Density (#fishers/km^2)
‘No management’ = Natures management
14
16
Does CPT apply?
• Effort in African lake fisheries seems selfregulated by system productivity
• Effort grows until the average catch rate per
fisher reaches around 3 ton per year
• Highest effort in most productive and resilient
systems. Less effort in low productive
vulnerable systems.
is there need for co-management?
Q2: Why are we afraid of effort
increase (f), while we at the same time
refine and develop the catchability (q)?
• What are the options of management regulations?
They can all be traced back to the simplest version
of the so-called catch equation:
•
•
We can regulate directly or indirectly on: Yield (Y),
Fishing mortality (F) or Biomass (B).
That is all. Any available or conceivable regulation
can be reduced to one of the three terms.
Management regulations
what are the options?
BMSY, Minimum SSB, MBAL, Bpa
B
Y  FB
Y
MSY, TAC, ITQ
F
Size of capture: tc
Mortality index: Z=F+M
Exploitation rate: E = F/Z
Effort control: f = F/q
F control: F0.1, Fmed etc.
Closed area
Closed season
Management regulations
• The choice of management regulations
depends on:
 Knowledge of the stock (research, monitoring)
 Control of the fishery (compliance, statistics)
 Management level (distribution, quotas…)
• In terms of required knowledge
(= management costs) then:
B > Y > F, where for the latter f > q
expensive
cheap
•
SSF = ‘q’ - management
For fisheries where little or nothing is known,
management regulations are always based on
regulating catchability q (in particular selectivity):
 Mesh size
 Size of capture
 Gear regulations (e.g beach seines…)
 Closed area or season (e.g. MPAs)
•
•
•
Find one example
where one or
several of these
do not apply
When nothing is known these regulations are based
on assumptions (often based on model results).
Next step is effort (f) control, then TAC etc.
Each new step requires exponential increase in
research and monitoring.
Co-management
•
•
•
•
Introduced because of the failures of enforcing
existing management regulations
Based on the same assumptions as conventional
management (CPT, i.e. avoiding the ‘tragedy’)
‘Tragedy’ can be avoided if the ‘common’ (read open
access) is removed → fishers become responsible
for the resources
Regulations are the same (always q-based) but
 Who are the ‘fishers’?
 Who will control access? Who will benefit?
Fishing mortality (F)
Better methods
Increasing these is
development
efficiency
catchability (q)
So while we
‘manage’ and
‘develop’ the fishing
mortality stays the
same.
Fishing mortality (F)
Effort (f)
Number of units
Who are we helping?
More of the same
Decreasing these is management
Catchability vs. effort
• Increased efficiency (q) requires increased
•
•
investments
Decreased effort (f) requires increased control
But the exploitation pressure on the fish stocks
will often be the same
- or even higher with investment (q) driven
development (exit is no longer easy)
• Only difference is a few rich vs. many ‘poor’
fishermen – but that is not a biological issue!!
• The conclusion of
Jul-Larsen et al.
(2003) was that
investment driven
growth (q), was
much more
dangerous than
population driven
growth (f)
• But
this is exactly
what we promote!!
Q3: Why is man the only predator that has
a harvesting pattern completely opposite
all other predators?
• Related to previous question
• Per definition then:
F = q = s when effort = 1
 Fishing mortality = catchability = selectivity for
one effort unit
• Harvesting pattern is how the fishing
mortality, catchability, or selectivity is aimed
at the target species (prey) over its lifetime
Predation vs fishing mortality..
Instantaneous rate of mortality
.. is almost exactly opposite
Predation
mortality
Fishing mortality
Age (years)
From ICES (1997).
..and this is what happens:
Median age-at-maturation (sexes combined) of Northeast Arctic cod based
on spawning zones in otoliths (from Jørgensen, 1990).
But we know that – we even use it
as a sign of fishing
Age and size structure changes
under selective fishing to younger
and smaller individuals.
effort
Age and size structure
As age and size structure changes
under selective fishing to younger
and smaller individuals, there will be
a decrease in:
• size (age) of maturity
• fecundity,
• egg quality
• egg volume,
• larval size at hatch,
• larval viability,
• food consumption rate,
• conversion efficiency,
• growth rate.
So, is this inevitable?
Life history and natural selection
Dying is more certain than giving birth!
Most ecological processes and life history traits can
be related to the prevailing mortality pattern:
• The unstable environment: characterised by
discrete, density independent, non-predictive, nonselective mortality induced by physical changes
• The stable environment: characterised by
continuous, density-dependent, predictive, and sizeselective mortality induced by the biotic community.
Mean size of organisms
Cope’s rule
Stable period
Stable period
Stable period
Stable period
Geological time
Cope's rule states that evolution tends to increase body size
over geological time in a lineage of populations.
But the precondition is geological stability. During unstable
periods with mass extinctions the large lineages are more
susceptible. Investment in age (size) is investment in future.
Life history: r-K selection
•
r-selected species:
 Small
 Rapid growth
 Early maturation
 No parental care
 Opportunistic
 Colonisers
 Unstable environment
 Resilient
•
K-selected species
 Large
 Slow growth
 Late maturation
 Parental care
 Specialised
 Competitors
 Stable environment
 Vulnerable
Logistic growth: r-K selection
Carrying capacity = B∞ = K
• r-selected species:
 Small
 Rapid growth
 Early maturation
 No parental care
 Opportunistic
 Colonisers
 Unstable

environment
Resilient
• K-selected species
 Large
 Slow growth
 Late maturation
 Parental care
 Specialised
 Competitors
 Stable

environment
Vulnerable
Abundance (Log N)
r-K selection as a function of
mortality pattern
Increased juvenile mortality
= K-selection = Z ↓
Slope = total mortality rate Z = rmax
Increased adult mortality
= r-selection = Z ↑
Age (size)
Kolding (1993)
K-selection: Stable environment, biotic mortality (predation) – predictive
r-selection: Unstable environment, abiotic mortality – non-predictive
Evidence: Size selection
= genetic changes
Increased mortality on:
Small
Random
Large
After Conover and Munch 2002
Mean individual weight at age for
six harvested populations after 4
generations. Circles, squares, and
triangles represent the small-,
random-, and large-harvested
populations, respectively.
Effect of size-selective fishing
Mortality on:
Small
Random
Large
Trends in average total weight harvested (A) and mean weight of harvested
Size selective fishing with large mesh sizes on adults
individuals (B) across multiple generations of size-selective exploitation.
Cope’slines,
rulesquares
in reverse.
Circles represent small=harvested
are the random-harvested
lines, and triangles are the large-harvested lines. Conover and Munch 2002
We are deliberately inducing r-selection on the stocks.
Instantaneous rate of mortality
Man has a harvesting pattern that is opposite to
what most fish stock are naturally adapted to
High juvenile mortality
= K-selection  large fish
High adult mortality
= r-selection  small fish
Predation
mortality
Fishing mortality
Age (years)
When yields are declining our prescription is even larger mesh sizes, which
will only make matters worse, and which we already know is wrong…
North Sea multispecies system
Percent changes in the long term
fishery yields for North Sea
stocks resulting from an
increase in trawl mesh size from
85 to 120 mm for the directed
fishery for cod.
Results are presented for
1) MSVPA including interspecies
predation and
2) single species (but multi-fleet)
assessment.
Lower yields in the MSVPA results
are due to greater predation
rates from large predatory fish
(cod, whiting, haddock, saithe)
released by the larger mesh
sizes.
Source: Anonymous 1989. Report of the
multispecies assessment working group.
Int. Counc. Explor. Sea., C.M.
1989/Asess: 20, Copenhagen.
Effect of mesh change from 85 to 120 mm
Multispecies
Single species
Cod
Whiting
Saithe
Mackerel
Haddock
Herring
Sprat
Norway pout
Sand eel
Total
-30
-20
-10
0
10
20
30
Percent change in long-term yield
40
Is non-selective fishing bad?
•
•
•
•
There is no empirical evidence
The notion comes from a theoretical model (Y/R)
which is 100% synthetic and biologically wrong
(constant parameters + no density dependence)
On the contrary we know that selective fishing is bad,
but we still advocate it!
But how do we impose gear-, mesh-, and size
restrictions in a multi-species fishery?
Multi-species community
How should it
be harvested?
What should
be minimum
mesh-size?
Biomass-size distributions
Selective fishing
will change the
slope
Size spectra
Abundance
The distribution of biomass by body
mass follows regular patterns
phytoplankton
zooplankton
small fish
big fish
Body mass
slope steepens when large fish removed
Jennings & Blanchard, 2004
Fishing effects on community size-structure
Trends in size-spectrum slopes of the North Sea
-5
Slope
-6
-7
-8
-9
-10
1976 1978 1980 1982 1984 1986 1988 1990 1992 1994
Year
Rice & Gislason (1996)
Rotate the size spectra and..
Quaternary consumers
Tertiary consumers
Secondary consumers
Primary consumers
Primary producers
… we get a Lindeman trophic pyramid
EAF: What is the ‘right’ fishing pattern?
How do we
A non-selective
manage
a multiharvesting
pattern
is
species
fishery?
what they
are criticised
for
What is the right
But a gears
non-selective
and mesh
harvesting
pattern is
sizes?
ecosystem conserving.
The system remains
unchanged,How
except
?
everything is less.
How much ?
Small-scale fisheries are often ‘non-selective’ !
Can we fish everything proportionally?
Example from lake Kariba, Zambia
where fishers are using illegal small-meshed nets
Parallel slopes, only
intercept lower
Kolding et al. 2003
•
The system remains unchanged, except everything is less
Conclusions
•
•
•
The ‘Tragedy of the commons’ is the tragedy of our
current management thinking.
Can we universally apply notions that are:
 Developed for single species fisheries
 Mostly theoretical
 Often dubious
On SSF - of which we know so little?
We are mammals and we apply all our models and
concepts based on “mammal biology”.
But fish – in their breeding strategy – are closer to
insects or trees!!
Final questions
• We try our best – but are we doing it right?
• Is our theory and paradigms appropriate?
• How much do we know?
• Why are we in the ‘management mode’ when
we have hardly started our ‘research mode’?
• Most SSF in Africa are still largely unmanaged,
but that is not a challenge - it is an opportunity!
For studying the impact of fishing and learn!
Because…
Thank you for your attention
• For SSF the real
challenge for
management is:
How can we
evolve our
theories?
Evolution of fisheries management
Slightly modified from Non Sequitur: Herald Tribune 11-12/8- 2007