Fish as feed inputs for aquaculture – Practices, sustainability and implications: a global synthesis 23
Habitats
The pelagic gear and purse seines used to target many industrial fish species such as
sprats, blue whiting and Peruvian anchovy are deployed in the water column and
have minimal contact with the sea floor. Demersal otter trawls are used to catch some
species, such as sand eel and Norway pout, and these may have more of an impact on
the sea bed and benthos. The degree of impact depends on the targeted species and
the location, as specific gears will be used to target specific species, and the impact on
the sea floor will relate to both the substrate type and the physiology of the flora and
fauna.
Typically in the North Sea sand-eel fishery, the trawl is kept close to the sea bed,
which is usually sandy (Wright, Jensen and Tuck, 2000), but actual contact is kept to
a minimum. The gear is also lighter than the gear used in other demersal trawls. The
effect of this disturbance on the more dynamic sand habitats is less significant than
disturbance in areas of lower energy such as muddy substrates and in deep water, as the
level of natural disturbance in the more dynamic areas is likely to be greater than that
caused by fishing (Kaiser et al., 1998).
Based on the results of 11 studies, six of which involved experimental trawling,
Johnson (2002) concluded that the physical effects of trawling on sand habitat include
trawl-door tracks left on the sea floor, smoothed sediments and removal of biogenic
mounds. At greater depths (>120 m), tracks were evident up to one year after trawling.
At shallow sites (< 7 m), tracks were no longer visible after a few days. Four studies
that examined the effects of chronic trawling and documented decreased abundance
and biomass of sedentary macrofauna and decreased diversity. Studies examining the
effects of short-term or pulse trawling documented changes in the abundance of some
infaunal and epifaunal taxa, such as polychaetes, nematodes and benthic diatoms. These
changes mimicked natural disturbance. Recovery ranged from weeks in intertidal areas
to possibly years at depths of 80–200 m.
Bycatch and discards
The incidental catch of non-target species, and in particular, the capture of juveniles of
commercial species, is one of the most controversial aspects of feed fisheries, as most
undersized fish are landed and processed. In North Atlantic waters, juvenile herring
are known to shoal with sprat (Hopkins, 1986), while juveniles of commercial species
such as whiting and haddock are known to shoal with industrial teleost feedfish such
as Norway pout (Huse et al., 2003; Eliasen, 2003). Bycatch levels are not necessarily
high – the bycatch in the Danish and Norwegian North Sea sand-eel fishery (mainly
herring, saithe and whiting) averaged 3.5 percent of the total catch over the period
1997–2001 (ICES, 2003a). While levels are low given the scale of the feed fisheries being
prosecuted, actual quantities of bycatch can be significant. In 2002, the Danish sand-eel
landings accounted for 622 100 tonnes, of which 3.7 percent was considered bycatch,
which is a total of 23 018 tonnes of bycatch herring, cod, haddock, whiting, saithe and
mackerel. In the same period, the sprat fishery took 27 972 tonnes of bycatch.
Globally, purse seines and other seines catch the vast majority of small pelagics.
These seine fisheries contribute over 350 000 tonnes to the global discard estimate
and have a weighted discard rate^5 of 1.6 percent (proportion of the catch discarded)
(Kelleher, 2005). Chilean fisheries harvests an average (1992–2001) of 5 million
tonnes of small pelagics – these fisheries have a low discard rate and account for less
than 40 000 tonnes of discards. Peruvian fisheries show a similar pattern of discards,
although a higher discard rate in the small pelagic fisheries (average nominal catch of 8
million tonnes, 1992–2001) generates discards of 260 000 tonnes.