64
M. Holmer et al.
vicinity of fish farms. This paradox could be attributed to the dispersive nature of
fish-farming sites, i.e., to the fact that phytoplankton cells do not stay long enough
to capitalize on nutrients (Gowen et al. 1983), or to experience rapid grazing by
zooplankton as suggested by Machias et al. (2005). Several studies have measured
significant diel changes of nutrient concentrations in the vicinity of fish farms in oligotrophic waters (Karakassis et al. 2001; Pitta et al. 2006), indicating that dispersion
is a very efficient mechanism at those sites. However, it has recently been shown
(Dalsgaard and Krause-Jensen 2006) that in situ incubation of phytoplankton and
of Ulva sp. can be used as a relatively low-cost monitoring strategy to document the
distance from the farms where pelagic primary production is affected. This method
has the advantage that it is not affected by episodic events such as those affecting concentrations of nutrients and particulate material in the water column, whereas the
incubation period of the bioassays allows for estimates based on integration of the
water quality conditions over several days. It is worth noting that even though these
bioassays have been able to detect changes up to a distance of 200–300 m from the
fish farms, the intensity of these effects decreases rapidly with distance. However,
when several farms are aggregated in a fish-farming zone producing thousands of
tonnes, it is reasonable to ask: what are the large scale effects of this aggregation
which should be detectable despite the nutrient dispersion? A recent survey in the
Mediterranean (Pitta et al. 2005) showed that most of the significant changes in
nutrients as well as chlorophyll a or PON were found at the deepest layer of the
water column below the thermocline, indicating that they are related to the remineralization of benthic organic material.
Wild fish communities are also affected by aquaculture. Partly, this effect is
related to the attraction of some fish species to the floating structures (see
Chapter 3 this volume), but fish communities can also be affected at large spatial
scales (Machias et al. 2004, 2005, 2006; Giannoulaki et al. 2005) probably
because of the changes in primary productivity in the area and the rapid transfer
of nutrients up the food web. This effect has been documented in the
Mediterranean where oligotrophic conditions and the structure of planktonic
communities seem to favour this process. In this context it has been suggested
(Machias et al. 2005) that fish communities are probably a good indicator of the
increased material flux since they are long-lived organisms integrating processes
over longer time periods, and their predators are unlikely to respond promptly to
an increase in their biomass.
The effects of fish farms on seagrass meadows have been documented by
many recent papers (Delgado et al. 1997; Holmer et al. 2003; Marbà et al. 2006;
Diaz-Almela et al. submit). In the recently finished EU-funded project MedVeg
(Effects of nutrient release from Mediterranean fish farms on benthic vegetation
in coastal ecosystems) four sites were monitored along the Mediterranean for
benthic fauna, sediment geochemistry, water quality and seagrass-related variables. The results showed that the distance of detectable effects varied greatly
among the variables used. In particular, seagrass mortality seemed to be the
indicator detected at greater distance than any of the others determined in this
project (Marbà et al. 2006; Frederiksen et al. 2007; Diaz-Almela et al. submitted).
M. Holmer et al.
vicinity of fish farms. This paradox could be attributed to the dispersive nature of
fish-farming sites, i.e., to the fact that phytoplankton cells do not stay long enough
to capitalize on nutrients (Gowen et al. 1983), or to experience rapid grazing by
zooplankton as suggested by Machias et al. (2005). Several studies have measured
significant diel changes of nutrient concentrations in the vicinity of fish farms in oligotrophic waters (Karakassis et al. 2001; Pitta et al. 2006), indicating that dispersion
is a very efficient mechanism at those sites. However, it has recently been shown
(Dalsgaard and Krause-Jensen 2006) that in situ incubation of phytoplankton and
of Ulva sp. can be used as a relatively low-cost monitoring strategy to document the
distance from the farms where pelagic primary production is affected. This method
has the advantage that it is not affected by episodic events such as those affecting concentrations of nutrients and particulate material in the water column, whereas the
incubation period of the bioassays allows for estimates based on integration of the
water quality conditions over several days. It is worth noting that even though these
bioassays have been able to detect changes up to a distance of 200–300 m from the
fish farms, the intensity of these effects decreases rapidly with distance. However,
when several farms are aggregated in a fish-farming zone producing thousands of
tonnes, it is reasonable to ask: what are the large scale effects of this aggregation
which should be detectable despite the nutrient dispersion? A recent survey in the
Mediterranean (Pitta et al. 2005) showed that most of the significant changes in
nutrients as well as chlorophyll a or PON were found at the deepest layer of the
water column below the thermocline, indicating that they are related to the remineralization of benthic organic material.
Wild fish communities are also affected by aquaculture. Partly, this effect is
related to the attraction of some fish species to the floating structures (see
Chapter 3 this volume), but fish communities can also be affected at large spatial
scales (Machias et al. 2004, 2005, 2006; Giannoulaki et al. 2005) probably
because of the changes in primary productivity in the area and the rapid transfer
of nutrients up the food web. This effect has been documented in the
Mediterranean where oligotrophic conditions and the structure of planktonic
communities seem to favour this process. In this context it has been suggested
(Machias et al. 2005) that fish communities are probably a good indicator of the
increased material flux since they are long-lived organisms integrating processes
over longer time periods, and their predators are unlikely to respond promptly to
an increase in their biomass.
The effects of fish farms on seagrass meadows have been documented by
many recent papers (Delgado et al. 1997; Holmer et al. 2003; Marbà et al. 2006;
Diaz-Almela et al. submit). In the recently finished EU-funded project MedVeg
(Effects of nutrient release from Mediterranean fish farms on benthic vegetation
in coastal ecosystems) four sites were monitored along the Mediterranean for
benthic fauna, sediment geochemistry, water quality and seagrass-related variables. The results showed that the distance of detectable effects varied greatly
among the variables used. In particular, seagrass mortality seemed to be the
indicator detected at greater distance than any of the others determined in this
project (Marbà et al. 2006; Frederiksen et al. 2007; Diaz-Almela et al. submitted).