Page 7 - 4022883

BIOLOGICAL CONSERVATION 141 (2008) 699–709 705

Fig. 4 – Fish species response to protection, measured as the natural log ratio of biomass between reserves and fished areas,
in relation to the enforcement level. Bars indicate 95% confidence intervals. Black circles: significant ratios; grey circles: non-
significant ratios. See Section 2 for details on the analysis.

mare, Torre Guaceto, Portofino) and at Tavolara (characterized Rocky reef communities strongly impacted by fishing may
by medium enforcement). This ecological threshold, there- thus show extirpation or functional extinction of fish species
fore, seems to be attainable only where enforcement is high that have important ecological roles (e.g. predators or herbi-
or where local fishing pressure is not very strong. At Ustica Is- vores), with consequences on ecosystem functioning and ser-
land marine reserve where enforcement has been effective vices (Holmund and Hammer, 1999; Worm et al., 2006). MPAs
for many years, Diplodus fish continued to have low densities or reserves that fail to increase fish densities and sizes may
as had historically been the case. This was attributed to the also face economic losses, e.g. in terms of decreased attrac-
paucity of habitats suitable for juvenile stages around the is- tiveness of seascapes deprived of large charismatic fish for
land (Vacchi et al., 1998). The Ustica case suggests that the recreational divers or lower incomes from fishing undertaken
recovery of functionally relevant species potentially affecting in proximity to the reserves’ boundaries. A crucial point,
the whole communities does not always occur, sometimes therefore, is that the enforcement and good compliance are
regardless of good enforcement, and proper information fundamental pre-requisites for fish populations to replenish
about fishing pressure and local ecological conditions are (Guidetti and Sala, 2007), spillover of adult fish to occur (Rob-
needed to elucidate why these unpredicted responses can oc- erts et al., 2001), other community-wide effects to be felt (e.g.
cur. Herbivorous fishes (mostly Sarpa salpa) also displayed trophic cascades or barren-algal transitions; Sala et al., 1998)
slightly greater density and biomass at the reserves than in and economic initiatives to be activated (Holmund and Ham-
fished conditions, which is not consistent with results ob- mer, 1999).
tained by Guidetti and Sala (2007). Such a discrepancy could
be due to the differences in local densities of piscivorous fish With the exception of Tavolara (characterized by medium
predators and, to some extent, to different local fishing tradi- enforcement and relatively low fishing impact), a positive
tions (e.g. the use of specific gears that may impact S. salpa as relationship was found between the level of enforcement
target or by-catch). and the total fish biomass. Particularly high values were
found at the Miramare reserve, very small in size and where