380 
Fishery Bulletin 113(4) 
77°0'W 
The locations of waterbodies sampled for assessing density of bycatch in derelict crab pots in North Carolina from 
April to November 2010: Bogue and Core sounds and Newport River (within the Central District of the North 
Carolina Division of Marine Fisheries) and Masonboro and Topsail sounds and Cape Fear River (within the South- 
ern District). 
We distinguished and sampled 3 habitat types in each 
waterbody: 1) marsh creeks; 2) estuarine edges (ar- 
eas of open water within 50 m of the shoreline); and 
3) areas adjacent to the ICW. Because one goal was 
to examine the interactions of diamondback terrapin 
with DCPs, we limited our surveys to nearshore wa- 
ters with depths <4 m, where diamondback terrapins 
and DCPs were most likely to co-occur. Given Avissar’s 
(2006) observation that most blue crab fishermen now 
locate their pots near to shore to avoid sea turtle in- 
terference, our focus on the nearshore habitat probably 
provides good coverage of the area of activity for crab 
pot fishing. 
With ArcMap, 8 vers. 9.3.1 (Esri, Redlands, CA), we 
used readily available maps and shapefiles to classify 
areas as adjacent to the ICW or as estuarine edge; 
however, we had to create by hand a unique shape- 
file for marsh creek areas in each waterbody, digitiz- 
ing these creeks with Google Earth software. A grid of 
1-km 2 cells was placed over the habitat-classified map, 
and then each cell in our entire study area was as- 
signed to one or more habitat types. The 1-km 2 grid 
was placed over our study area in the habitat-classi- 
fied map with Google Earth software, which allowed 
Mention of trade names or commercial companies is for iden- 
tification purposes only and does not imply endorsement by 
the National Marine Fisheries Service, NOAA. 
us to eliminate inappropriate cells (i.e., estuarine 
edge cells with insufficient water surface area, and 
marsh creek cells where the entire natural shoreline 
was hardened) from the pool of possible cells. Approxi- 
mately 20% open water was considered sufficient wa- 
ter surface area in most cases; very few marsh creek 
cells were as low as 10% open water. All sufficient 
cells had much more than the 24,000-m 2 area of wa- 
ter that we sampled within each cell. In addition, we 
eliminated cells known to be in an area included in 
the NCDMF program for removing and disposing of 
lost and abandoned crab pots. 
A stratified random sampling design was used to se- 
lect the cells to be sampled. A total of 1088 cells formed 
our sampling pool. Each cell was designated by habitat 
type within each waterbody, with the possibility that 
a given cell could represent more than one habitat 
type. Using ArcMap, we delineated the 1-km 2 cells, de- 
termined the latitude and longitude of cell centroids, 
and calculated the area of open water for each cell. We 
determined that our resources allowed us to survey 
approximately 200 cells overall, a count that could in- 
clude some of the same cells twice (as separate samples 
because they represented 2 different habitat types). 
The number of cells to be sampled in each waterbody 
was determined by a weighting method for both the 
relative numbers of fishing trips with hard-shell crab 
pots (averaging NCDMF 7 data for 2006-2008) and the 
