Patrick et al.: Use of productivity and susceptibility indices to assess vulnerability of fish stocks to overfishing 
317 
Table 5 
Nonparametric statistical analysis of targeted versus non-targeted species productivity, susceptibility, and vulnerability scores 
in the highly migratory Atlantic shark complexes and Hawaii-based pelagic longline sector fisheries. 
Fishery 
Kruskal-Wallis P-values 
Number 
Productivity 
Susceptibility 
Vulnerability 
Hawaii-based pelagic longline — tuna 
33 
0.026 
0.373 
0.072 
Hawaii-based pelagic longline — swordfish 
33 
0.026 
0.153 
0.058 
Highly migratory Atlantic shark complexes 
37 
0.150 
<0.001 
0.380 
Combined 
103 
0.752 
<0.001 
0.160 
based pelagic longline (tuna sector), Hawaii-based 
pelagic longline (swordfish sector), and the highly migra- 
tory Atlantic shark complexes (nontarget stocks are 
identified in Appendix 1). Kruskall-Wallis tests revealed 
that the productivity scores were significantly different 
between the target and nontarget stocks in each of the 
two sectors of the Hawaii-based pelagic longline fishery 
(P- 0.026), whereas the susceptibility scores were sig- 
nificantly different (PcO.OOl) in the highly migratory 
Atlantic shark complexes (Table 5). None of these cases 
showed significant differences in both axes, and no sig- 
nificant differences were observed in vulnerability. Like 
others, these results indicate that nontarget stocks can 
be as vulnerable to overfishing as the target stocks of a 
fishery and reinforce the need for a careful examination 
of the vulnerability of nontarget stocks when making 
management decisions (see Alverson et al., 1994; Hall, 
1996; Kaiser and de Groot, 2000). 
Conclusions 
Although many qualitative risk analyses are used by 
fisheries scientists and managers, the PSA is a par- 
ticularly useful method for determining vulnerability 
because it permits an evaluation of both the productiv- 
ity of the stock and its susceptibility to the fishery. The 
output from this relatively simple and straightforward 
tool provides managers and scientists an index of how 
vulnerable target and nontarget stocks within a fishery 
are to becoming overfished. Even when specific values 
for many life history parameters are not well known, 
the categorical bins of low, medium, and high values are 
often distinct enough to allow scores for even the most 
data-poor species. The bins also help in determining 
the needed strength of conservation measures and the 
degree of precaution to apply in management measures. 
They can also identify those stocks or fisheries that war- 
rant further, more complicated analytical attention. 
Our analyses indicate that the PSA is generally ca- 
pable of distinguishing the vulnerability of stocks that 
experience differing levels of fishing pressure, although 
fixed thresholds separating low-, medium-, and high- 
vulnerability stocks were not developed. When fixed 
thresholds of vulnerability are desired, it is recommend- 
ed that managers and scientists determine thresholds 
between low-, medium-, and high-vulnerability stocks on 
a fishery-by-fishery basis, using cluster analysis or other 
techniques that identify groups of similar species. 
Like those of Shertzer and Williams (2008), our ex- 
ample applications showed that current stock complexes 
exhibit a wide range of vulnerabilities (e.g., highly mi- 
gratory Atlantic shark complexes). Therefore, managers 
should consider reorganizing complexes that exhibit a 
wide range of vulnerabilities, or at least consider choos- 
ing an indicator stock that represents the more vulner- 
able stock! s) within the complex. If an indicator stock is 
found to be less vulnerable than other members of the 
complex, management measures should be conservative 
so that the more vulnerable members of the complex are 
not at risk from the fishery. 
It is also important to note that PSA scores will 
likely vary between sectors of a targeted fishery (e.g., 
gear type, user group) or among fisheries that capture 
the stock as bycatch. For example, the susceptibility 
score for “survival after capture and release” may differ 
greatly between trawl and gill net gears. Thus, it is rec- 
ommended that a vulnerability evaluation be performed 
for all or a majority of sectors interacting with the stock 
when the overall vulnerability of stock is needed (e.g., 
for setting control rule buffers, identifying sectors where 
stocks are particularly vulnerable, etc.). An overarching 
vulnerability evaluation score could then be calculated 
by using a weighting system based on average landings 
by sector over some predetermined time frame. 
Scientists have begun using the PSA in developing 
control rules for fisheries management. For example, 
the South Atlantic Fishery Management Council is 
considering an acceptable biological catch control rule 
that is based on a tiered system that reduces the prob- 
ability of overfishing from 50% (i.e., the overfishing 
limit) to as low as 20% based on 1) the uncertainty in 
the stock assessment, 2) the status of the stock, and 3) 
the vulnerability score from the PSA (SAFMC 4 ). Ad- 
ditional control rule frameworks are being developed 
4 SAFMC (South Atlantic Fisheries Management Council). 
2009. Briefing book-attachment 10: Scientific and Statisti- 
cal Committee’s draft ABC control rule, lip. South Atlantic 
Fisheries Management Council Meeting, Stuart, FL. 
