248 
Fishery Bulletin 99(2) 
mercial set nets and pots are made of nonbiodegradable 
man-made materials. As a result, lost nets and pots have 
the potential to persist and continue to fish in the marine 
environment for several years depending upon the pre- 
vailing environmental conditions (Breen, 1987; Carr et ah, 
1990; Kaiser et ah, 1996). Set nets that are lost in areas 
exposed to large swell and storm activity (e.g. off the North 
west coast of Spain) are rapidly destroyed (Puente 2 ). Nets 
lost in shallow, clear water are rapidly overgrown with 
encrusting biota that makes them more visible and reduc- 
es their fishing capabilities (Erzini et ah, 1997). However, 
when static gear becomes snagged on rocks that hold it 
in place, or is lost in deep water in a relatively stable en- 
vironment, it may continue to fish for more than a year 
(Carr et ah, 1990; Kaiser et ah, 1996). 
Lost pots are likely to continue fishing for longer than stat- 
ic nets because they are constructed either entirely of metal 
or of thick net t ing attached to a rigid frame. Exact ly how long 
lost pots are likely to continue fishing remains unquantified. 
The present study was undertaken in grounds where pot 
gear is typically used — at a site within the Skomer Ma- 
rine Nature Reserve (MNR), Wales, UK. MNR staff have 
recorded the occurrence of lost pots over a number of years 
within the waters of the reserve. Most observations were of 
single pots in locations sheltered from strong water move- 
ment. Such pots were usually heavily overgrown with ses- 
sile biota, indicating that they had been submerged for 
at least 6 months. Lost pots within the MNR have been 
reported by recreational divers who have described occa- 
sional large resident catches of Crustacea and fish, such as 
ballan wrasse ( Lcibrus bergylta) and conger eels (Conger 
conger). The objectives of our study were 1) to quantify the 
number of organisms removed by a fleet of lost pots; 2) to 
describe changes in catch rate over time; and 3) to record 
any deterioration in the integrity of the fishing gear. 
Methods 
Site selection 
The study site was located in the Skomer MNR off the 
coast of Pembrokeshire, Wales (Fig. 1). This site was rep- 
resentative of those fished by locally based inshore pot 
fishermen. Maximum tidal streams ranged from 2 to 3 
knots on spring tides. The site was sheltered from heavy 
wave action and the seabed comprised mainly boulders to 
a depth of 15 m below the lowest recorded astronomical 
tidal height. Below this depth, the substratum was com- 
posed of mixed coarse sediments, cobbles and boulders, 
and occasional bedrock outcrops. 
Gear and deployment 
The gear comprised 12 parlor pots of approximate dimen- 
sions 750 mm x 500 mm x 500 mm with a top-mounted 
cylindrical entrance ring made of plastic (Fig. 1). Two sizes 
2 Puente, E. 1997. Personal commun. Isla de Txatxarramendi, 
s/n 48395 Sukarrieta (Bizkaia), Spain. 
of entrance ring were used, either 200 mm (n= 7) or 240 
mm (n=5) diameter. In our analysis we made no attempt 
to differentiate catches based on entrance ring size because 
size of trap entrance was not a major consideration in our 
experiment. The pots were attached to a mainline by using 
3.5-m lengths of polypropylene rope at intervals of 18 m. 
This configuration of the gear is the normal fishing prac- 
tice used in the UK (Fig. 1). The fleet of pots was anchored 
at each end with a 75-kg weight and marked with surface 
marker buoys (Fig. 1). Each pot was baited with the corpse 
of a thornback ray ( Raja clavata ) (its wings having been 
removed) and attached to the entrance ring with a rubber 
band. On each sampling occasion the position of the fleet 
of pots was determined by a differential global positioning 
system. The pots were deployed on 4 August 1995. During 
deployment, the mainline was checked to see that it was 
tight between each of the pots when shot from the boat. 
Gear deployment was undertaken in consultation with a 
local professional pot fisherman. Divers surveyed the fleet 
of pots immediately after it had been shot away to confirm 
that the gear was deployed on the seabed correctly and 
to record the depth at which each pot had settled. It was 
thought important to record pot depth during the study 
because pots lost in shallow water might be more prone to 
destruction by wave action, whereas progressive movement 
of the pots into deeper water might prolong their fishing 
capabilities. Although it would have been preferable to have 
deployed replicate fleets of pots in different habitats, the 
logistics of diver-based observations and the need to mini- 
mize the adverse effects of the experiment on local popula- 
tions of animals meant that this was neither possible nor 
considered ethically acceptable. Indeed, the ethical consid- 
erations of undertaking such an experiment in an MNR 
meant that 3 pots were removed from the experiment after 
88 days and a further 3 pots removed after 270 days of fish- 
ing to minimize animal deaths associated with the study. 
Data recording 
Observations on each of the pots were recorded by divers 
1, 4, 12, 27, 40, 69, 88, 101, 125, 270, 333, 369, and 398 days 
after initial deployment. Divers undertook the following 
tasks: 
1 recorded the depth (adjusted for tidal height) of pots 
throughout the experiment. 
2 recorded the identity of the catch in each pot. 
3 tagged newly captured crustaceans (brown crab [Cancer 
pcigurus ] spider crab [ Maja squinado } and the lobster 
Homarus gammarus ) on each sampling occasion with 
coded and colored cable ties on their appendages so that 
they could be distinguished from newly captured ani- 
mals (without tags) on subsequent occasions. 
The identification of previously caught individual lob- 
sters, brown crabs, and spider crabs caught in the pots 
was checked by inspection of the species-specific colored 
and coded tags placed on different body parts of each ani- 
mal during a previous sampling routine. It was not pos- 
sible to tag smaller crustaceans such as the velvet swim- 
