246 
Fishery Bulletin 1 14(2) 
Table 4 
Mean catch per unit of effort (CPUE), by gear type, determined 
from shore-based creel surveys conducted by the Guam Division 
of Aquatic and Wildlife Resources in 2010-2012 and estimated 
biomass of targeted species, by gear type, from visual surveys 
conducted by the Coral Reef Ecosystem Program of the NOAA 
Pacific Islands Fisheries Science Center in 2011. These values 
were used to calculate the catchability coefficient ( q ) for each 
gear type. Data for cast-net fishing were excluded from this anal- 
ysis (see the Materials and methods section). gh=gear hour. 
CPUE (kg/gh) 
Biomass 
Gear type 
2010 
2011 
2012 
(t) 
(/1000 gl 
Drag net 
0.99 
0.91 
0.31 
157 
0.0042 
Gill net 
1.26 
0.09 
0.77 
386 
0.0012 
Hook and line 
0.03 
0.09 
0.14 
324 
0.0002 
Scuba spear 
0.37 
0.40 
0.24 
544 
0.0006 
Snorkel spear 
0.06 
0.06 
0.38 
386 
0.0003 
12000 -I 
10000 - \ 
V 
8000 - 
6000 - 
4000 - 
\.\ 
A 
M 
I I 
*<! 
A 
l» 
l » 
I » 
I * 
- — Total 
3-year mean 
♦ Visual surveys 
2000 - 
/ 
• V 
1985 
1990 
1995 
~ 1 — r~ 
2000 
“i — r~ 
2005 
“I 1 1 - 
2010 
Figure 3 
Reconstructed time series of total biomass of reef fishes in 
shore-based creel surveys conducted around Guam in 1985-2012 
in relation to the estimate of total biomass from visual surveys 
conducted in 2011. The time series excludes pelagic species and 
bai tfishes and is based on annual catch composition per gear 
type with the assumption of a constant catchability coefficient 
for each gear type. The black diamond represents the estimat- 
ed biomass for 2011 computed by multiplying biomass density 
(from surveys conducted by the Coral Reef Ecosystem Program 
of the NOAA Pacific Islands Fisheries Science Center) by the 
reef area (71 km 2 ) of hard-bottom habitats at depths of 0-30 m 
around Guam. The gray line indicates the 3-year mean of catch- 
es summed over all gear types. 
the data up to the island level by combining both day 
types and all regions. With the estimated catchability 
coefficient of each gear type (Table 4), we reconstructed 
a time series for relative biomass (Fig. 3). There was 
a steep (almost 50%) downward trend in fish biomass 
from 1985 to around 1990. In the following 
decade, fish biomass leveled off but then de- 
clined again by about 2003 (Fig. 3). 
On the basis of the reconstructed time se- 
ries for total fish biomass (Fig. 3), we were able 
to estimate relative biomass of the functional 
groups targeted in the shore-based fisheries 
for reef fishes (Fig. 4). For the reconstructed 
biomass per functional group, species compo- 
sition was assumed to have remained similar 
to the species composition observed during the 
visual surveys conducted in 2011 and, there- 
fore, is an estimate of the true value. 
Discussion 
The fishery data used in this study show an 
overall reduction in catches in Guam that 
in turn indicates a general decline in recon- 
structed stock biomass from the late 1980s to 
recent years (Fig. 3) — a finding that is consis- 
tent with the results of fishery-independent 
surveys that show significant depletion of 
shallow-water reef-fish populations and large 
fishes around Guam (Richards et al., 2011; 
Williams et al., 2011, 2015). Parrotfishes were 
caught by hook-and-line fishing (33%) and by 
spearfishing (40%), and lower landings were 
reported in recent years. Assuming there have 
been no increases in recruitment and that 
the size composition of the catch reflects the 
size structure of the population, this decline 
in catches indicates a reduction in adult bio- 
mass and reproductive potential of the stock 
(Birkeland and Dayton, 2005). The decline 
in reconstructed biomass of parrotfishes and 
other herbivores (Fig. 4) observed in our study 
is also of some ecological concern. Herbivo- 
rous fishes play a critical role in maintaining 
benthic algal communities in states that are 
conducive for coral growth and recruitment 
(Mumby et al., 2007a; McClanahan, 1997), and 
herbivory is particularly important in promot- 
ing coral recovery after disturbances, such as 
hurricanes and bleaching events (Mumby et 
al., 2007b; Edwards et al., 2011; Graham et 
al., 2013). Therefore, declines in local parrot- 
fish biomass may indicate that reefs surround- 
ing Guam are less resilient to such events. 
Estimated CPUE and catches for all gear 
types were substantially lower during the re- 
cent period (2007-2012) than they were dur- 
ing the period 1985-1990 (Table 2). Because 
CPUE estimates for the hook-and-line fishery 
in our earliest period (0.15 kg/gear hour for 1985-1990; 
Table 2) are considerably lower than the estimates 
recorded in Guam during the early 1980s (0.55 kg/ 
gear hour; Katnik, 1982, cited in Dalzell et al., 1996), 
it seems likely that the evident downward trends in 
