Data Files

SERIES C
Applied Statistics

Bayesian modelling of catch in a north-west Atlantic fishery
C. Fernández, E. Ley and M. F. J. Steel
Appl. Statist., 51 (2002), 257 - 280


FILES

(1) 2LOGBK.DAT => the dataset

(2) pm5.f => code for running MC^3 and also generate coordinates for posterior densities, and other results. These later routines, as the setup() routine, are highly specialized for our application and are only provided for completeness. It is expected that you'll want to replace then w/ your own. It contains and example of parameter (driver) file, also highly specific to our application commented out at the top of the
file.

STATEMENT

The dataset and code can be freely used and distributed for
non-commercial purposes. (Please, refer to our paper: C. Fernández, E. Ley and M.F.J. Steel (2002), "Bayesian modelling of catch in a north-west Atlantic fishery," Appl. Statist., 51, 257 - 280.)

CONTACT

edu_ley@hotmail.com

A. Data

The dataset contains data on catches per species for a number of vessels collected by the European Union (EU) in the context of the Northwest Atlantic Fisheries Organization (NAFO).

The original data were gathered by the inspection vessel of the EU operating on the Grand Bank fishery. Inspectors board the fishing boats and record basic characteristics of the ship and the fishing equipment, as well as the quantities caught of different species and where and when this catch was effectuated. They use the ship's logs to collect all the
information accumulated since the last time the ship was boarded. All data correspond to 1993 and the first half of 1994, leading to 6806 observations each corresponding to a particular ship at a given day.

The file 2logbk.dat contains 1 record per day/ship.

The format is:

read(fdat,*) n, total,(species(k),k=1,6),ship,
& country,(dzone(k),k=1,4),zvisits,day,month,
& yr, solo_par, type, art, net, length,
& grt, kwh, dfish

n integer, # obs 1..6806
country 1 denmark
  2 spain
  3 portugal
(Note: There are only danish 3 ships, which are not used in the paper)
   
zone 1 3l
2 3m
3 3n
4 3o
species 1 cod -- bacalao
 
2 ghl -- fletan negro
 
3 red -- gallineta
  
4 rng -- granadero
  
5 ska -- raya
  
6 wit
706
7 cat 700
8 hal 525
9 yel 107
  10 pla 591
  11 flx 75
  12 pra 91
  13 hkw 89
  14 mon 49
  15 mnz 26
  16 scu 11
  17 tun 5
  18 skh 2
in kgs, species[6] contains species 6--18 (others).

day

year

   
     
lone/pair 1 fishing alone
2 fishing in pair		  
type 1 go -- volantero
2 sn -- seiner: cerquero
3 to -- trawler
4 ts -- side trawler
5 tt -- trawler popa
6 ttf --- trawler, freezer		  
art 1 1 gn -- red de enmalle a la deriva
2 2 gns -- red de enmalle anclada
3 3 ot -- otter trawl, arrastre
4 3 otb -- otter trawl bottom
5 4 ptb --- pair otter trawl bottom, parejas		  
(Pls note: techniques 3 and 4 are combined into 1 in the paper.)
net mesh in milimeters  
length ship length in meters  
grt gross registered tonnage  
kwh principal motor power in kwh  
dfish days fishing  
ship ship id #  

NOTE: There is an error in Table 3 in the paper (page 261), two of the reported standard deviations (species 1 and 3) are incorrectly reported. (The error was kindly pointed out by Colin Chalmers, Stats Dept, LSE.) Table 3 should read:

Table 3. Catch for various fish species

species Mean STD % zeroes % catch
1 550.20 3181.83 88.33 9.80
2 3503.48 3610.55 18.50 62.40
3 658.51 3311.26 85.73 11.73
4 213.46 502.18 43.20 3.80
5 503.39 1661.50 55.44 8.97
6 185.23 625.76 72.27 3.30
(Last modified: August 26, 2002)
B. Code

The f77 code implements the MC^3 algorithm.

 Journals

SERIES A
Statistics in Society

SERIES B
Statistical Methodology

SERIES C
Applied Statistics

SERIES D
The Statistician