Page 8 - m313p261

268 Mar Ecol Prog Ser 313: 261–270, 2006

60Bray-Curtis similarity In Capo D’Orlando Bay, where the MVC was high,
70 the percent contribution of fish farm waste organic
80 0m matter increased with increasing distance from the
90 500 m cages. Waste particles were influenced by a ‘sail
100 1000 m effect’, due to high currents that dispersed waste par-
H2 (0 m) ticles to farthest distances and reduced isotopic traces
H1 (500 m) at sites closer to the cages. In the sediments, some
H1 (1000 m) particles settled near the cages (about 16% of the
H2 (500 m) total), whereas most settled at greater distances,
H1 (0 m) contributing about 28% and 34% to total SOM at
H3 (1000 m) 500 m and 1000 m, respectively. Recent studies by
H3 (0 m) other researchers supports our results. Using manipu-
H2 (1000 m) lative experiments, Cromey et al. (2002) modelled the
H3 (500 m) ‘rolling’ activity of waste particles under the influence
of bottom water currents. Artificial particles that sim-
CTRL (EI) SOM A SOM B ulated the specific weight and size of common salmon
feed were moved several hundreds of meters from
Avg. dissim.: 24.3 Avg. dissim.: 20.1 cages in the presence of high velocity currents at the
bottom of the water column. This agrees with the
SM (28.7), PEL (26.9) PEL (32.9), SM (31.8) idea that resuspension movements from a bottom cur-
TER (26.4), EJE (18.1) EJE (23.4), TER (11.9) rent can be a major factor that influences the exten-
sion plume from point-source disturbance. In addi-
Avg. dissim.: 41.7 tion, our results suggested that under a given
situation (depth of ~25 to 30 m and current velocity of
PEL (32.3), SM (32.2), EJE (22.4), TER (13.1) ~18 to 40 cm s–1), sediment resuspension can enhance
the effects of fish farm waste (in the form of pellets
Fig. 7. Cluster analysis of SOM mixing model percent and ejections) by enlarging its spatial influence by
data (Bray-Curtis similarity index). See Fig. 6 for further about 10 to 20 times that observed at locations with
lower current velocities (value extrapolated from the
(equivalent) details; Fig. 2 for acronyms difference between the extension area at Gulf
of Castellammare, Menfi Bay, and Capo d’Orlando).
function of different levels of hydrodynamics. Other Cluster analyses provided important confirmation
researchers, with similar objectives, have consistently regarding the variable impact of waste from fish
emphasized the importance of hydrography on a local farms depending on distance and current velocity.
scale (Wu et al. 1994, Wu 1995). Nevertheless, previous
studies have never before shown a clear effect of waste The water column and sediments appear to react
products beyond 100 to 300 m from the centre of cage differently to fish farm waste when influenced by dif-
emissions (Pearson & Black 2000). In contrast, by using ferent hydrodynamic levels. In the last 2 decades, most
the stable isotope approach, our results enabled us to studies have focused on water column dynamics in
trace organic enrichment over larger distances. Our 3 order to detect impacts of fish farming activities. Some
study locations were characterised by remarkably have analysed the dynamics of the dissolved nutrient
different current velocities, and the resulting hydro- pool or the features of suspended organic matter (La
dynamic patterns explained the different influences of Rosa et al. 2002, Alongi et al. 2003) and the free living
fish farm waste on the surrounding environments of and attached fraction of suspended bacteria or phyto-
each location. plankton (McKinnon et al. 2002). More recently, other
authors have considered sediment dynamics in their
In the Gulf of Castellammare, low current velocities protocols, and have determined that the overall fea-
produced deposits of cage waste at a distance less than tures of SOM result in an underestimation of the real
100 m from the point-source. When moving away from importance of the role of this bulk (sensu Danovaro et
fish farm cages, the isotopic contribution of cage waste al. 2003). Nevertheless, our isotopic approach allowed
quickly decreased, and it contributed less than 10% to us to infer that the water column represented an
total POM and SOM at the farthest sites. aleatory subsystem with respect to the sediments. At
the control location (Egadi Islands), the isotopic com-
When the current velocity approached about 20 to position of the water column was naturally variable
22 cm s–1, as in Menfi Bay, the fish farm waste contri- and showed overall differences in both δ13C and δ15N
bution decreased in the water column (not more than among sites separated by a distance of 500 m. Such dif-
10 to 15%) with increasing distance from the cages; as
was also observed at the Gulf of Castellammare loca-
tion, peak contributions were apparent close to the
cages. In contrast, the SOM exhibited an inverted
pattern; cage waste contributions to SOM increased
linearly with distance. Waste contributions reached a
peak at the 1000 m distance, where it comprised about
38% of total SOM.