Smith and Zeller: Unreported catch and tourist demand on local fisheries in The Bahamas 
119 
A 6 
3 • 
\ / 
Visitors 
/ \ 
_Q 
E 
Residents 
1950 
1960 1970 1980 1990 2000 2010 
Figure 1 
For The Bahamas during 1950-2010, (A) number of residents com- 
pared with total number of visitors and (B) number of residents com- 
pared with total number of stopover visitors, who stay at least one 
night, and total number of cruise ship visitors, who partake in shore 
visits. 
Materials and methods 
Reconstruction of marine fisheries catches 
Reconstruction of fisheries catches is an approach to 
retroactively estimate catches when reliable time-se- 
ries data are lacking (Zeller et al., 2007, 2015). In some 
instances, this approach has involved interpolations, 
cautious extrapolation, and assumptions based on local 
expert opinion in lieu of quantitative data. The use of 
interpolations, extrapolation, and assumptions has re- 
sulted in potentially higher uncertainty in some of the 
data provided here (see also Zeller et al., 2011a, 2015), 
but this approach is justifiable because of the unaccept- 
able alternative, namely that catches for missing sec- 
tors, taxa, or time periods would be interpreted as zero 
catches — an outcome that has serious consequences for 
effective management and conservation (Pauly, 1998). 
The catch reconstruction approach 
used here consists of 7 general steps 
(Zeller et al., 2007, 2015): 
1. Identification of existing time-series 
data on catches to validate the quality 
of data transfer from national (e.g., an- 
nual reports from the Bahamas Depart- 
ment of Marine Resources, previously 
named the Department of Fisheries) to 
international (e.g., FAO reported land- 
ings data by FAO area, taxon, and year) 
levels. 
2. Identification of sectors, time periods, 
taxa, etc. not covered by step 1 through 
literature searches and local expert 
consultations. 
3. Search for available information sourc- 
es to serve as alternatives to missing 
catch data identified in step 2, through 
comprehensive literature searches and 
local expert consultations. In this step, 
we look for any source of information, 
including case studies, health studies, 
household surveys, technical reports, 
data sets, and expert opinion. 
4. Development of data anchor points in 
time for missing segments based on 
data and information sources discov- 
ered in step 3, and expansion of them 
to countrywide catch estimates for each 
sector or taxa by using clearly stated 
and conservative assumptions. 
5. Application of interpolation for time 
periods between data anchor points 
for each fishing sector or taxa, either 
linearly or on the basis of assumptions 
for commercial sectors, and application 
of interpolation, typically through per 
capita catch rates for noncommercial 
sectors, taking into account major po- 
litical, socioeconomic, and environmen- 
tal impacts. 
6. Estimation of final time series of total catch by com- 
bining reported catches (identified in step 1) and 
interpolated, country-expanded, missing data seg- 
ments (produced in step 5). The final data series 
shows catch by fisheries sector, taxon, and year. We 
define fisheries sectors using country- or regional- 
specific definitions: large-scale commercial, arti- 
sanal (small-scale commercial), subsistence (small- 
scale noncommercial), and recreational (small-scale 
noncommercial) 
7. Expression of the level of uncertainty in the data 
and information sources and in the assumptions 
made during reconstruction, by fishing sectors and 
time periods of the reconstruction. This final step is 
based on “scoring” criteria inspired by the Intergov- 
ernmental Panel on Climate Change (Mastrandrea 
et al., 2010; Table 2). Because the senior author of 
