444 
Fishery Bulletin 106(4) 
Shallow sites 
Filograna implexa 
Encrusting bryozoans 
0.8 -| 
0.6 - 
0.4 - 
0.2 
o 
May-96 May-97 
F dis i = 5.22*; F , 
Undisturbe^ 
May-98 May-99 
May-00 
V' 
: 0.14; F. 
dist x yr 
= 0.19 
Hydroids 
Bushy bryozoans 
Sponges 
Cultch frequency 
Figure 2 
Temporal trends in colonial epifaunal percent cover and cultch percent frequency at shallow 
sites (40-50 m). Trend lines from sites that were either disturbed or undisturbed by bottom 
fishing are shown. Shallow undisturbed data came from site 17, whereas shallow disturbed 
data were derived from sites 16, 17W, and 18. The scale of the y axis varies between graphs 
of different epifaunal taxa. Markers (i.e., black squares and black diamonds) are back-trans- 
formed means, and error bars are 95% confidence intervals. Different letters above data points 
represent annual means that differ significantly from each other in multiple comparison tests. 
F statistics from two-way analyses of variance (ANOVAs) on the effects of disturbance (dist), 
year (yr), and disturbance x year (distxyr) are shown below each graph (df=l,4,4). Significance 
levels from ANOVA tests are indicated as follows: *P<0.05; **P <0.01; ***P<0.001. 
the only taxon to have a significant year term in the 
two-way ANOVAs performed on deep sites (Fig. 4). 
The greater between-year variability at shallow sites 
can be interpreted as a sign of the effectiveness of CA- 
II. Because there can be subtle, pre-existing differences 
in habitat quality between areas inside and outside a 
marine reserve, a conclusive demonstration of reserve 
effectiveness typically requires that biological distinc- 
tions between fished and unfished areas become am- 
plified over time (Murawski et al., 2004). As a result, 
variables truly affected by the establishment of CA-II 
should have significant interaction terms in their corre- 
sponding two-way ANOVA, as well as exhibit significant 
differences between disturbance categories and among 
years. Sponges and encrusting bryozoans were the two 
colonial epifauna taxa that exhibited significant in- 
teraction terms in the ANOVAs performed for shallow 
sites. No interactions between disturbance and year 
were detected at the deep sites. This finding reflects 
the fact that bottom-fishing effort was fairly consistent 
over time at deep sites (Table 1; Collie et ah, 2005), 
whereas fishing effort at shallow sites showed a more 
distinct temporal trajectory related to the establish- 
ment of CA-II. 
