126 
Fishery Bulletin 114(1) 
Use of expert opinion in reconstruction of fisheries catch 
data 
An expert is someone who possesses knowledge about 
a given topic through training, research, practicing of 
skills, or personal experience (Burgman et al., 2011). 
Elicitation of expert opinion has been used for a va- 
riety of environmental issues, including conservation 
(e.g., Murray et al., 2009), invasive species manage- 
ment (e.g., Kuhnert, 2011), climate change impacts (e.g., 
Morgan et al., 2001), and models of managed systems, 
such as logged forests (e.g., Crome et al., 1996) and da- 
ta-limited fisheries (e.g., Griffiths et al., 2007). Many of 
the parameters in our reconstruction are derived from 
the opinion of a single expert (i.e., M. Braynen, Direc- 
tor of the Department of Marine Resources) because 
of limited empirical data (Table 4). Nevertheless, we 
followed the 5 general steps described by Martin et al. 
(2012) to elicit expert knowledge: 
1. Decide how information will be used; 
2. Determine what to elicit; 
3. Design the process for eliciting information; 
4. Execute the process for eliciting information; and 
5. Translate the information for use in a model. 
We elicited expert information both directly and 
indirectly for use in our catch reconstruction. Direct 
elicitation of information involved posing questions 
that would provide quantitative responses that could 
be used directly in the reconstruction. For example, we 
asked: What percentage of commercial catch of Carib- 
bean spiny lobster do you think is unreported? 
In contrast, indirect elicitation of information in- 
volved posing questions that resulted in responses that 
were subsequently converted to quantitative values for 
use in our reconstruction. For example, to come up with 
the estimate for the subsistence catch rate of 32.55 
kg-person _1 -year _1 for 40% of residents of the Family 
Islands during 1950-1969, we asked 2 questions: 1) 
what percentage of residents of the Family Islands do 
you think relied on subsistence fishing for at least part 
of their dietary requirements from 1950 through 1969? 
and 2) Given your answer to question 1, if we estimat- 
ed the amount of seafood obtained through subsistence 
fishing from 1950 through 1969 in terms of the number 
of plate-sized snappers consumed per person per week, 
how many plate-size snappers per person per week do 
you think were obtained by subsistence fishing in the 
Family Islands? 
In all instances, information was elicited through 
one-on-one interviews with the expert either in person 
or over the telephone. It should be noted, however, that 
we did not directly elicit an estimate of uncertainty 
around a model parameter from the expert. Instead, 
we relied on the more general approach to estimating 
uncertainty involved in catch reconstructions — an ap- 
proach that was inspired by criteria used by the Inter- 
governmental Panel on Climate Change, as previously 
described (Mastrandrea et al., 2010; Table 2). 
As with any research method, there are limitations 
to expert advice. Most notable is the range of subjec- 
tive and psychological biases that experts, and indeed 
all humans, are prone to (see overview in Supporting 
Information of Martin et al., 2012). Given the above, 
we agree with the statement of Martin et al. (2012) 
that: “Expert knowledge should be regarded only as a 
snapshot of the expert’s judgments in time, and expert 
assumptions and reasoning should be documented in 
such a way that they can be updated as new empirical 
knowledge accrues.” 
Tourist demand for local fishes 
We separated tourist demand for local fishes into 2 cat- 
egories: 1) demand by stopover visitors and 2) demand 
by visitors who arrived on cruise ships. For our study, 
we define stopover visitors as tourists that spend at 
least one night in The Bahamas (Cleare, 2007). Most 
stopover visitors arrive by air, but some of them arrive 
by other means (e.g., private yacht). Stopover visitors 
fish recreationally and consume local fishes in restau- 
rants in The Bahamas. In contrast, tourists that arrive 
by cruise ship, as defined in our study, typically spend 
only a few hours ashore in The Bahamas and increase 
demand for local fishes only through seafood consump- 
tion during shore visits. 
To estimate stopover visitor demand for local fishes 
through seafood consumption in hotel restaurants, we 
designed and successfully administered a local seafood 
consumption survey with 11 major hotels on 5 differ- 
ent island groups. Our study sample represented 37% 
of all hotel rooms in The Bahamas in 2010. Hotels in 
our survey ranged in size from 19 to 2932 rooms. In all 
instances, the purchasing manager or head chef of a 
hotel restaurant completed the survey, which included 
requests for information on the type, quantity, origin 
(i.e., The Bahamas versus imported), and dollar value 
of fishes supplied to the restaurant on a yearly basis 
(Smith and Zeller 5 ). 
In our survey, there was a suspiciously large quan- 
tity of seafood that was purported to be of local ori- 
gin. We, therefore, assumed that 10% of all so-called 
local seafood was actually imported and we adjusted 
consumption accordingly. Visitors to The Bahamas con- 
sume both local and imported (e.g., salmon) seafood 
in hotel restaurants. Our study focused only on local 
seafood consumption. Overall seafood consumption (i.e., 
both local and imported products) by visitors to The 
Bahamas is therefore much greater than our estimates 
provided here. 
Then we combined tourist data (e.g., hotel occupancy 
rates, number of visitor nights per year) with results 
from our survey to estimate a consumption rate per 
stopover visitor for consumption of local seafood. Al- 
though staff at hotels provided data for a period that 
ranged from 2 through 18 years, most hotels provided 
data for only the last 2 years of our time series (i.e., 
2009-2010). We are unaware of any previous estimates 
of local seafood consumption rates in hotels in The 
