Transcript A new look at the interactions between marine mammals - FTP-UNU

A new look at the interactions between
marine mammals and fisheries
(or did we we mess up the North Atlantic, and now
blame them for what we did?)
Daniel Pauly
Fisheries Centre
University of British Columbia
Vancouver, Canada
United Nations University Fisheries Training Programme
Institute of Marine Research, Reykjavik, Iceland, December 18, 2002
For example, virtually all species at or neat the top of North
Atlantic food webs have been depleted by excess fishing, as
illustrated here by trends in biomass (blue) and fishing
mortality (red) from single-species assessments.
Trends from
R.M. Myers’
online database
http://www.mscs
.dal.ca/~myers/w
elcome.html
Iceland is no exception, as illustrated here by
the trend in cod biomass…
Figure courtesy of Hreidar Thor Valtysson, University of Akureyri
To generalize these results, we defined small spatial
cells, which can be mixed and matched as needed.
We chose ½ lat./long cells (of which there are about
260,000 in the world ocean, and 20,000 in the North
Atlantic).
Then we used a ruled-based algorithm…
Taxon (what)
Taxon distribution
database
NO
Country (who)
FAO Area (where)
Fishing access
database
Spatial reference
database
Common spatial
cells?
Improve
databases
YES
Assign catch rate to common cells
This routine now
assigns over 99 % of
FAO global marine
catches to ½ degree
spatial cells, and we
are still improving
the underlying
databases …
Here is an example of a map thus generated using
catch data from 1900 and about…
(t/km2year-1)
Catch of table fish in 1900
(fishes with trophic level > 3.75)
(map by R. Watson, SAUP)
Which can be compared with data from 100
years later…
Catch of table fish in 1999
(R. Watson, SAUP)
That we can see little difference is not surprising,
given the evolution of catches in the North
Atlantic…
Now to the ecological processes underlying all of
this. As we know, fisheries exploit resources
embedded within food webs…
wherein each organism has its own trophic level …
Thus, we can estimate (from catch data and the trophic
level of all species caught) the mean trophic level of
global fisheries landings. This is declining…
TL of landings
3.4
3.3
3.2
Marine
3.1
3.0
2.9
2.8
Freshwater
2.7
1970
1975
1980
1985
1990
Pauly et al. Science March 1998
This process, wherein fisheries catches are increasingly based on
organisms low in the food web, now known as ‘fishing down
marine food webs,’ is particularly strong in the North Atlantic…
Northwest Atlantic
3.4
3.3
3.2
3.1
3.0
2.9
3.7
Northeast Atlantic
2.8
1970
1975
1980
1985
1990
Pauly et al. Science (1998);
CJFAS (2000); etc…
TL of landings
TL of landings
3.5
3.6
3.5
3.4
3.3
1950 1955 1960 1965 1970 1975 1980 1985 1990
This can be verified for Iceland, thanks to an
excellent catch database assembled by Hreidar
Valtysson, of Akureyri University…
Available from the author, or online at www.saup.fisheries.ubc.ca)
This process is very strong in Icelandic waters...
Mean trophic level of landings
4.0
3.9
3.8
3.7
3.6
3.5
3.4
3.3
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
Figure courtesy of Hreidar Thor Valtysson, University of Akureyri
The ‘fishing down’
effect is strong and
everywhere.
The impact of ‘fishing down’ on marine ecosystems in
huge, as illustrated here by food web models from two datarich food areas. In these pyramids, the volume at each
trophic level indicates the amount of biological
production….
Note that the
corresponding
pyramids for
2002 would be
about half the
size of those for
the 1980s…)
Primary and secondary ecological data
We used 23 Ecopath
models to generalize
these values to the
entire North Atlantic:
We then:
• Predict biomasses from
multiple linear regression
stratified over the North
Atlantic area.
Data
Data
AVAS transformations
Temperature
GIS data
The approach is:
• based on ecosystem
models (23 models
representing 15 areas) to
supply information over
time and space.
• based on a ½ degree
latitude by ½ degree
longitude grid system;
Ecosystem
model (23)
Ecosystem
Ecosystem
Ecosyste
models (2)
m
model
models
(1)
Spatial models
Depth
No
Primary
production
Ice cover
Spatialized dataset
Jackknife
models:
OK?
Multiple linear
regression
Yes
Catch data
(Ecopath)
Consistency
checks
Catch (by
category
and year)
Analysis
Analyses
Predict spatial
biomass (by year)
Catch
(spatialized, by year)
Fishing intensity
(by year)
SAUP
database
Compare results with
stock assessments
Results
Results
Discussion
Discussion
(t/km2)
Biomass of table fish in 1900
(map: V. Christensen, SAUP)
Biomass of table fish in 1999
(map: V. Christensen, SAUP)
Or once more…
The squashed
down food
pyramids
shown earlier
imply shorter
food chains…
The shorter food
webs, and the
reduced biomass
imply, ironically,
a greater
susceptibility to
environmental
fluctuations
(Pauly et al., Nature, Nov. 2002)
We now turn to the marine mammals
Biomass
(millions of metric tonnes)
One fact for
starters: there
are less of
them now
than before.
And hence their food consumption is less than
before…
Consumption or catch
(millions of metric tonnes/year)
..even though
they still do
consume more
than we catch
in total.
But these gross number are not very informative,
and we thus applied our mapping approach to
marine mammals, as well.
The source of the marine mammal distributions...
One first result is here: the distribution of marine
mammal food consumption in the 1990s.
[Note that, presently, our model does not include seasonal migrations, and is thus likely to
underestimate feeding rates in the Northern North Atlantic, and conversely for the Southern North
Atlantic. Correcting for this will likely add to food consumption around, e.g., Iceland
The key equation is here. It defines an overlap
index that accounts for both the diet (or catch)
composition of the 60 + species involved (and all
fisheries), and the respective amounts taken.
This is the index that we
mapped…
The feeding overlap between the fisheries and marine
mammals can also be shown by groups:
There is a better ways to look at whales and
other marine mammals, anyways:
Cited from: ‘In a Perfect Ocean’ (Island Press, 2003; D. Pauly and J. Maclean)
Here are
more
details on
this.
In conclusion: the problem of fisheries, in the North
Atlantic cannot be resolve by killing more wildlife.
Rather, we must deal with the root cause, illustrated
by this map, where areas with no fishing whatsoever
are shown in green …
Acknowledgements…
• Thanks to the Pew Charitable Trusts, Philadelphia;
• Fisheries Centre, Faculty of Graduate Studies, UBC;
• Members of the Sea Around Us project;
and many others.