Transcript E Chapter 13 Impacts of Aquaculture

Kaiser part three; Impacts
Chapter 13: Aquaculture
IMPACTS
FROM AQUACULTURE
Simultaneously with a stagnation or reduction in the output from fisheries on
wild fish stocks in recent decades (from reasons treated in chapter 12), the
production of marine species in aquaculture (mariculture) has increased.
While this may be an answer to the increasing demand for protein worldwide,
it has also created new problems for marine ecosystemes and natural stocks
of resource species. In Europe and North America the problems have first and
foremost been connected with the production of anadromous salmonids (Atlantic
salmon and trout, and several species of Pacific salmonids). Shortly, the problems
may be grouped in:
1. Problems connected with escapees and their genetic effects on wild stocks
2. Problems connected with escape and competition with wild stocks
3. Problems connected with transfer of disease and parasites to wild stocks
4. Problems connected with the use of marine fish as feed for farmed fish
5. Problems connected with local environmental effects of the farming industry
These aspects are also enlightened by other lecturers in the BI2060 course.
BI 2060 V09
1
Kaiser part three; Impacts
AQUACULTURE
Note almost linear growth 1950-2000
BI 2060 V09
2
Kaiser part three; Impacts
AQUACULTURE
I the year 2000 the
world production in
aquaculture (including
plants) was ca one
third of the outtake
from natural stocks.
This proportion has
increased in the last
decades.
Aquatic plants included
BI 2060 V09
The clearly biggest
producer is China.
Other asian countries
are also well represented on the list.
3
Kaiser part three; Impacts
AQUACULTURE
Numbers of persons employed doubled since 1970
The number of persons employed in fisheries and aquaculture have been more
than doubled on a world basis since 1970. Both the total increase and the rate of
increase have been similar in the two industries.
BI 2060 V09
4
Kaiser part three; Impacts
AQUACULTURE
The proportion
from finfish of the
total aquaculture
production has
increased much
faster than other
groups since
1970.
BI 2060 V09
5
Kaiser part three; Impacts
AQUACULTURE
NB! Greenland has a small
human population!
BI 2060 V09
6
Kaiser part three; Impacts
AQUACULTURE
HOW IS SEAFOOD FROM AQUACULTURE
PRODUCED?
BI 2060 V09
7
Kaiser part three; Impacts
AQUACULTURE
Below, a typical seawater pen for
fish production. Excess feed sinks to
the bottom under the pen and create
a milieu of decomposing materials
if not brought away by water currents.
Seawater pens are exposed
to strong natural forces and
have therefore been placed
in relatively sheltered areas.
However, the locations must
also allow a sufficient water
exchange in the pens. Pen
wreckage and fish escapes
are not uncommon problems.
BI 2060 V09
8
Kaiser part three; Impacts
AQUACULTURE
Ocean pens
Pen wreckage and excess feed problems in coastal waters have resulted in research
and development of ocean-based fish farming plants. Pilot plants have been dispatched
both in the North Atlantic and in the Gulf of Mexico. The idea is to avoid local pollution
by excess feed, and to reduce the wave strain by lowering the pens below the surface.
BI 2060 V09
9
Kaiser part three; Impacts
AQUACULTURE
Landbased fishfarms often use recirculated water. Among the advantages
are better insurance against poisonous algae blooms, bad weather and
predators (like seabirds and marine mammals).
BI 2060 V09
10
Kaiser part three; Impacts
AQUACULTURE
Fish produced in tanks and pens often
show physical signs of their captive life.
A very common sign is fin erosion due to
small living space and constant wear by
contact with the tank or pen walls.
The high density of individuals makes way
for infections and contagious diseases.
Oxygen deficiency during critical
embryonic development stages have been
shown to cause physical deformations
(skull- and spine deformations).
In Atlantic salmon, weared and rounded
fins is one of the criteria for identification
of escaped farmed fish.
BI 2060 V09
Solea solea (sole)
with damaged tail fins
caused by attempts to
bury themselves on the
tank bottom.
11
Kaiser part three; Impacts
AQUACULTURE
Air photo of Javanese coastline
Java has an extensive production of shrimps. The cultivation takes place in
small ponds in glennes cleared by deforestation of the coastline. Similar
conditions are found other places in the world. Mangrove forests have also
been removed to secure space for shrimp production (with bad results in
all aspects).
BI 2060 V09
12
Kaiser part three; Impacts
AQUACULTURE
Trading dried sea horse in China
Clown fish in
aquarium
(a) Dried sea horse (a small fish) is a highly appreciated medicine in China. Under
the danger of overexploiting the natural stocks, aquaculture production of
seahorses has become a lucrative business e.g. in New Zealand.
(b) Similar conditions are valid for popular aquarium fishes (here: Clown fish,
which naturally lives on coral reefs and was threatened by over-exploitation).
BI 2060 V09
13
Kaiser part three; Impacts
AQUACULTURE
PRODUCTION OF MOLLUSCS AND SHELLFISH
In Norway, blue mussle production based on natural settling on hanging ropes
has been an interesting enterprise for many local grunders. The professional
skills did not always match the enthusiasm, and a lot of bankruptsies were seen.
Poisoning by algae, and seabird predation have caused substantial problems in
many areas along the Norwegian coast.
Production
plants found
along most of
the coast.
Blue mussel is easy to cultivate.
Production is based on natural settling
and natural feed in form of plankton.
BI 2060 V09
14
Kaiser part three; Impacts
AQUACULTURE
PRODUCTION OF MOLLUSCS AND SHELLFISH
Aquaculture production of crustaceans has no long tradition in Norway. The type
of shrimp production which are so successful in asia is for natural reasons not so
viable and competitive i Norway. Nevertheless; in the last decades there has been
some activity in production of species with a particularly good market price (lobster).
As usual, when low production costs and high market value are the incentives for
an industry, experiments with imported species have also been tried for lobster
(i.e. american lobster in Norway). The american lobster show better growth than the
European, but is also more aggressive and will have some advantages if allowed
to compete for habitats.
American lobster imported for intensive production in containment has escaped
from captivity, and has been observed at large on several locations along the coast.
The danger for the European lobster is a reality.
BI 2060 V09
15
Kaiser part three; Impacts
AQUACULTURE
WHAT EFFECTS DO INTENSIVE PRODUCTION
SYSTEMS FOR SEAFOOD HAVE ON NATURAL
ECOSYSTEMS?
BI 2060 V09
16
Kaiser part three; Impacts
AQUACULTURE
Bunnforurensing under skjelloppdrett-redskap
Intensive production of mussles on hanging ropes often leads to organic stress at the
sea bottom below the plant. As usual with this type of pollution, this may often lead to
a dominance of a few opportunistic species (here the polychaete Schistomeringos loveni),
and a corresponding reduction of the species diversity. Even if the absolute effect may
vary with the general richness of nutrition, the difference between the bottom right below
the ropes and the adjacent areas is very clear (cf graphs above).
BI 2060 V09
17
Kaiser part three; Impacts
AQUACULTURE
Problems connected with escapes and genetic
introgression into natural populations
Species under domestication will inevitably undergo genetic changes, both intended
and unintended, while in captivity.
Intended genetic changes occur when stocks are bred to enhance certain genetic
traits which are advantageous for e.g. high production and captive life.
Unintended genetic changes are due to characteristics of the captive environment
itself. Also, the usually small captive populations are much more unstable with
respect to gene frequencies and will rapidly loose genetic variability.
Both types of genetic change will probability mean disadvantages to wild populations
if the escaped specimens are allowed to interbreed with wild relatives. It has been
documented that such introgression is taking place in Atlantic salmon in Norway.
The unwanted effects will increase with the number of generations the fish has
been under domestication, and with the magnitude of the introgression (i.e. the
number of escapees per generation or in total).
BI 2060 V09
18
Kaiser part three; Impacts
AQUACULTURE
Basic population genetics
Hardy-Weinbergs law says that in a statistical ideal population of diploid individuals,
the proportion of the genotypes for single-locus traits are determined by the allele
frequencies at the locus according to the binomial formula. Such a population is said
to be in Hardy-Weinberg equilibrium. Both allele frequencies and genotype proportions
are then stable over generations. If the population for some reason has been brought
out of H-W equilibrium, one generation of panmixia (random mating) is enough to
reconstruct the H-W equilibrium.
The H-W law rests on 5 specific assumptions:
1. Panmixia (random mating)
2. No mutation (can be relaxed in short term)
3. No random genetic drift (i.e. very large population)
4. No gene flow from other populations (with different allele frequencies)
5. No selection (neither natural nor artificial)
BI 2060 V09
19
Kaiser part three; Impacts
AQUACULTURE
HW genotype proportions acc. to
the binomial formula:
(f+s)2 = f2+ 2fs + s2
Suppose a population of diploid organisms
and a locus with two alleles F and S, which can
combine into genotypes FF, SF and SS.
We draw a sample of N=100, and count the
numbers of the different genotypes (table below).
where f and s are frequencies of
allele F and S, respectively, and
the genotypes are FF, SF og SS.
NB! These are frequencies; to get
numbers, multiply by N.
This is how to test if a population is in Hardy-Weinberg equilibrium
( kji-kvadrat Goodness-of-fit test )
FF
SF
SS
N
qF
qS
2
Obs.
35
50
15
100
.60
.40
0.677
Exp.
(36.0)
(48.0)
(16.0)
100.0
Expected numbers of the three genotypes under H-W equilibrium are found by putting the estimated allele frequencies in the
sample into the binomial formula. The number of degrees of freedom (DF) in this test is the number of different genotypes minus
the number of different alleles (i.e. DF = 3 – 2 = 1). The calculated chi-squared with DF=1 corresponds to P = 0.414 (not
significant) as looked up in a chi-squared table.
BI 2060 V09
20
Kaiser part three; Impacts
AQUACULTURE
Looking closer on the underlying assumptions for the H-W law:
1. Panmixia (random mating)
2. No mutation (can be relaxed in short term)
3. No random genetic drift (i.e. infinitely large population)
4. No gene flow from other populations (with different allele frequencies)
5. No selection (neither natural nor artificial)
No natural population fullfills all these assumptions, but some may come so close
that the error sources are small in practical use of the theorem.
Populations in captivity, on the other hand, often deviate rather strongly from
these assumptions, particularly numbers 1, 3, and 5 in the box above. This means
that allele- and genotype frequencies can change over few generations. This has
been documented e.g. for farmed salmon in Norway. Even if the original brood stock
was taken from wild stocks (9-10 generations ago), there are today clear changes in
allele frequencies and reduced genetic variability compared to wild salmon. At the
same time, selection programs for certain traits such as growth, sexual maturation
and behaviour have been undertaken and changed the gene pool accordingly.
BI 2060 V09
21
Kaiser part three; Impacts
AQUACULTURE
The introgression situation in Atlantic salmon in Norway
A rule of thumb in quantitative genetics says that the offspring will perform approx. at the average
of the parents phenotypic value for traits with a reasonable degree of heritability.
Hence, if escaped farmed salmon interbreed with wild relatives, the offspring from each crossing
will peform intermediate between farmed fish and wild fish. This will apply to traits like growth, age
at smoltification and -maturation, aggressivity and other traits with known (and quite high; 0.3-0.5)
heritability. The inheritance of genetically based behaviour traits do not differ from that of other
quantitative traits.
Because wild salmon stocks probably, during many generations of natural selection, have been
genetically adapted to their home rivers, hybrid offspring will probably be inferiour to natives with
respect to fitness in a specific habitat. Natural selection will act to "clean up" the stock, but as long
as the stock is not on or near its K (carrying capacity), the introgression will tend to reduce the total
fitness and productivity of the native salmon stock.
If the introgression is acute and massive, and/or is repeated over many generations, it will affect
the evolutionary potensial of the Atlantic salmon as a species. Norway has a particular international
responsibility for management of salmon because Norwegian rivers hold, by far, the largest part of
th total gene reservoir (gene pool) of the Atlantic salmon.
BI 2060 V09
22
Kaiser part three; Impacts
AQUACULTURE
Question: How large genetic effect can be expected in a situation where escaped
farmed fish interbreed with wild populations?
Answer: It will depend on the actual situation (case-by-case). The determining
factors will be:
1.
2.
3.
4.
5.
6.
Proportion of farmed fish immigrants breeding locally each generation
Number of generations with such immigrations impact
The genetic basis for the trait under change (number of genes affecting the trait)
If and how strongly the trait is selected for or against in the farmed population
The initial genetic difference for the trait between farmed and wild fish
The strength of local natural selection affecting the trait under study
To enlighten the effect of the various factors, one can perform ”what-if” analyses
by means of computer simulations. Starting with the simplest situation; a singlelocus polymorphism with two alleles A and B in a diploid organism, one can
follow the change in allele frequencies over generations using the software program
PopG.exe by Joe Felsenstein, or P14G.exe by J. Mork (see next slide).
BI 2060 V09
23
Kaiser part three; Impacts
AQUACULTURE
Simulation of evolution (change of allele frequencies) in a population by means
of the interactive Windows program PopG.exe by Joe Felsenstein (sample screen
dump).
Example of PopG simulation
BI 2060 V09
24
Kaiser part three; Impacts
AQUACULTURE
Screen dump of simulated evolution (change of allele frequencies) in a population by means
of the interactive DOS program P14g.exe by J. Mork, NTNU. (Sample screen dump).
Example of P14g.exe simulation
("Screen-dump" fra dataprogrammet P14.exe (J. Mork, TBS). X-akse: antall generasjoner, Y-akse: frekvens allel F). Kurven viser
10 uavhengige simuleringer av alleltap som følge av disruptiv seleksjon (heterozygotens er lavere enn homozygotenes).
Simuleringen er gjort med genetisk drift 'innkoblet'.
BI 2060 V09
25
Kaiser part three; Impacts
AQUACULTURE
Simulation of an introgression of escaped
farmed fish into a wild stock
First: The effect of random genetic drift in captivity and in the wild
Assume that we are studying a local stock of Atlantic salmon with an effective population size
(Ne) of 1000 individuals, which receives escaped farmed salmon each generation. The farmed
salmon origins from the wild stock, but genetic drift has lead to allele frequency differences
between the farmed and the wild fish over time, because the Ne of the farmed brood stock has
been only 10 individuals in 10 succeeding generations.
First, we will look at what is expected to happen over time in the two groups; immigrant
(donor) and resident (recipient), respectively, as an effect of random genetic drift. Thereafter,
given the new genetic characteristics of the two groups, we will look at the effect over time of
an immigration of escaped farmed fish each generation on the wild population’s allele
frequencies.
BI 2060 V09
26
Kaiser part three; Impacts
AQUACULTURE
Genetic drift in the captive stock (the immigrant). Assume that the culture has lasted for 10 generations, and
that the Ne has been 10 individuals each generation. The stock origined with a wild stock with the same genetic
characteristics as the one playing the role as resident (recipient) in the following simulations. The change in farmed
fish allele frequencies can be simulated (below are shown the outcome from ten independent simulations).
Ten independent
simulations with the
given parametres
resulted in the loss of
one of the two alleles in
three out ten cases
(33%)
BI 2060 V09
27
Kaiser part three; Impacts
AQUACULTURE
Genetic drift in the wild salmon population (recipient). Assume an effective
size (Ne) of 1000 individuals. During 10 generations the allele frequencies will
be affected by genetic drift, but to a much lesser degree than in the donor
population because of the much larger Ne. Below are shown outcomes from ten
independent simulations.
10 independent simulations
show that the changes in
allele frequencies due to
random genetic drift over 10
generations are considerably more moderate in the
large wild stock than in the
small captive stock. In the
follwowing simulations the
allele frequencies in the wild
stock will, for sake of
simplicity, be regarded as
constant.
BI 2060 V09
28
Kaiser part three; Impacts
GENETIC IMPACT FROM AQUACULTURE
Genetic effects of immigration
(without local selection)
From the 10 outcomes on previous slide we use,
again for sake of simplicity, one of the three
outcomes where one of the alleles was lost, i.e. qF=0.
After 10 generations with genetic drift
qF
qS
N (number)
Wild
0.500
0.500
900
Farmed
0.000
1.000
100
If we put these parametres into the
program P14g.exe, and simulates a
continued evolution in many generations,
we see that the escaped farmed fish
eventually will change the wild population
in its own direction. During only 10
generations the frequency of the F allele
will have changed from 0.50 to 0.25.
If this regime continues, also the wild
stock will loose its F allele (i.e. half of its
total genetic variability) in about 30
generations .
The graph to the right shows the outcome
of 5 simulations. They gave more or less
the same end result.
Effect of immigrations
(no local selection)
BI 2060 V09
29
Kaiser part three; Impacts
GENETIC IMPACT FROM AQUACULTURE
Genetic effects of immigration
(with local selection)
We insert the farmed population’s Ne and
allele frequencies, and simulates a sitiation
with 10% farmed fish immigration into the
wild population each generation (see table
at left for model input).
Start situation
qF
qS
N (number)
Wild
0.500
0.500
900
Farmed
0.000
1.000
100
A simulation of this scenario with the
P14g.exe program showed that after 40
generations, the continuous
immigration of farmed fish (which
lacked the F allele) had resulted in the
loss of this allele also in the wild
population (cf graph to the right).
Effect of immigration
(with local selection)
In this simulation, a ”self-cleaning”
local selection force which favoured
the F allel in the wild stock was
included, with fitness coefficients
shown in the graph heading (graph).
BI 2060 V09
30
Kaiser part three; Impacts
GENETIC IMPACT FROM AQUACULTURE
Genetic effects of introgression (with or without local selection)
Simulations like those shown above leave little doubt that an introgression of
farmed fish into wild populations may result in clear genetic changes. Such a oneway geneflow from a donor to a recipient will have as result that:
• The recipient will be more and more like the donor genetically
• If the donor is genetically altered and has lost genetic variability, this will
eventually also apply to the recipient population
• If such regimes goes on for extended time, the genetic diversity and
evolutionary potential of the wild stocks will be impaired.
• Selection will have a certain self-cleaning effect on the wild stock, but selection
can only work through increased mortality and therefore lead to reduced natural
productivity. If the farmed fish are genetically modified organisms, they will
transfer their genetic material to the wild stocks according to the same genetic
principles.
The consequences of this for natural populations, species and ecosystems can be
very unfortunate, and international nature management authorities generally
agree that such situations must be avoided.
BI 2060 V09
31
Kaiser part three; Impacts
NON-GENETIC IMPACTS FROM AQUACULTURE
Escaped farmed fish compete with wild fish on spawning grounds
• It is known from Atlantic salmon in Norway that farmed, fish because of
growth advantages, may outcompete their smaller wild relatives in the fight
for spawning redds in the rivers.
The hybrid offspring may be physically less fit for life in nature
• The offspring from introgression into wild stocks can, because of larger
body size, be unable to ascend small rivers. In this way they may represent
”useless” production from the human point of view.
In Norway, fish farms are "hatching sites" for salmon lice
• When the wild salmon return from the sea phase, they are exposed to an
unnatural high consentration of sea lice in coastal waters. Not being treated
for the problem, they will struggle with infections which can be decimating
for the wild stocks.
BI 2060 V09
32
Kaiser part three; Impacts
NON-GENETIC IMPACT FROMAQUACULTURE
"CONVERSION EFFICIENCY"
Salmon farming has been a huge economic success in Norway. Introductory
problems, mostly i form of disease, have found their solutions, and some believe
that this industry has an almost unlimited growth potential.
However; ther is one very apparent limitation to growth which cannot be
overlooked:
In today’s situation the feed used to produce salmon is taken from other marine
resources (in form of fish meal from tobis, herring, capelin and other industry
species). The problem is that these resource species themselves are limited in
size. Actually, some of them show clear signs of overexploitation, and may
therefore give limited output in the relatively near future.
It has often been held that todays’ salmon industry is not sustainable: It is
energetically inefficient ethically bad to use species high up in the food chain as
feed for salmon; if salmon farming shall be continuously growing, the industry must
change to use feed species from lower trophic levels in the food chain, that being
plants or animals.
BI 2060 V09
33
Kaiser part three; Impacts
AQUACULTURE
Simulation of the introgression of a genetically modified
organism (GMO) into a natural population
In the development of genetically modified organisms, genes are spliced into
the individuals’ genome in order to enhance specific traits (e.g. growth rate).
During the process, marker genes are used to track the incorporation. For this, it
has been common to use a gene that gives resistence towards antibiotics or
pesticides, and hence the incorporation of the new gene can easily be traced by
common bacteriological techniques (inoculation on treated agar gels).
If such a GMO is allowed to introgress into natural populations, it can lead to an
uncontrolled spreading of antibiotic resistance in nature by horizontal gene
transformation.
The entire process; escape rates, gene flow, local selection and introgression
rates can be simulated with software like that demonstrated on this course
(PopG.exe and P14g.exe).
NEXT SLIDE>
BI 2060 V09
34
Kaiser part three; Impacts
AQUACULTURE
Simulation of introgression of a GMO into natural population
Assume a GMO which
”leaks” one individual from
its containment each
generation, and a natural
population of Ne=1000
which is the recipient of
such leakages.
The simulation on the graph
to the right shows the
introgression when the
GMO is given a 20% better
fitness than wild relatives
relative to planticides, due
to its resistence against
drugs. Typically, the
frequency of the GMO gene
increased from zero to
fixation in the wild
population in only ~50
generations.
BI 2060 V09
35
Kaiser part three; Impacts
AQUACULTURE
BI 2060 V09
36