Voss et al.: Factors driving the density of derelict crab pots and associated bycatch in North Carolina 
381 
Table 1 
Distribution of sampling effort and raw abundance of derelict crab pots (DCPs) and actively fished crab pots (AFCPs) found 
within cells (each 1 km 2 ) sampled in 2010 during a survey of marsh creek, estuarine edge, and adjacent Atlantic Intracoastal 
Waterway (ICW) habitats in Bogue, Core, Masonboro, and Topsail sounds and Cape Fear and Newport rivers in North 
Carolina. Abundance of DCPs is the number observed in the 24,000-m 2 areas surveyed in sampled cells, and abundance of 
AFCPs is the number observed in the entirety of the 1-km 2 cells. Note that the ICW does not traverse Core Sound and was 
not sampled in Cape Fear River because depths were >4 m. 
Number of cells sampled DCP abundance AFCP abundance 
_ Total 
Marsh Estuarine Adjacent cells Marsh Estuarine Adjacent Total Marsh Estuarine Adjacent Total 
Waterbody 
creek 
edge 
ICW 
sampled 
creek 
edge 
ICW 
DCPs 
creek 
edge 
ICW 
AFCPs 
Bogue Sound 
8 
8 
8 
24 
8 
1 
1 
10 
58 
106 
94 
258 
Cape Fear River 
30 
30 
- 
60 
14 
13 
- 
27 
48 
51 
- 
99 
Core Sound 
30 
30 
- 
60 
9 
8 
- 
17 
369 
219 
- 
588 
Masonboro Sound 
9 
9 
9 
27 
20 
19 
4 
43 
20 
53 
55 
128 
Newport River 
5 
5 
5 
15 
0 
1 
0 
1 
43 
24 
20 
87 
Topsail Sound 
5 
5 
5 
15 
1 
3 
4 
8 
19 
14 
18 
51 
Totals 
87 
87 
27 
201 
52 
45 
9 
106 
557 
467 
187 
1211 
relative combined area of marsh creek, estuarine edge, 
and ICW-margin habitats within each waterbody. 
For each waterbody, we multiplied the proportion of 
total trips by the proportion of total area of the 3 habi- 
tat types combined. Sampling effort, or the number of 
cells to be sampled, was then allocated to waterbodies 
in proportion to the product of the 2 proportions. Each 
of 1088 cells was categorized by waterbody and habi- 
tat type and a random number generator was used to 
rank the priority of the cells for sampling. We deter- 
mined that a 201-cell sampling design allowed us to 
balance effort among habitat types within waterbodies. 
To distribute the 201 1-km 2 cells to be sampled across 
the 6 waterbodies, the number of cells assigned to each 
waterbody were divided equally among the 3 habitat 
types (Table 1), with the exceptions of the 2 waterbod- 
ies that lacked appropriate ICW habitat: Core Sound, 
which the ICW does not traverse, and the lower Cape 
Fear River, where the ICW is a major shipping channel 
with all depths >4 m. In those 2 waterbodies, the num- 
ber of cells to be sampled was divided evenly between 
the habitat types of marsh creek and estuarine edge 
(Table 1). 
Field sampling 
Sampling was conducted within 99 cells over 11 field 
days, between 28 April and 3 June 2010 in the Core 
Sound, Newport River, and Bogue Sound (within the 
Central District of the NCDMF) and within 102 cells 
over 12 field days, between 10 May and 10 August and 
on 11 November 2010 in Topsail Sound, Masonboro 
Sound, and Cape Fear River (within the Southern Dis- 
trict of the NCDMF) (Fig. 1). Of the 201 cells sampled, 
23 cells that were randomly selected to represent 2 
different habitat types were sampled twice, accounting 
for 46 of the 201 samples. A team from the University 
of North Carolina at Chapel Hill Institute of Marine 
Sciences conducted surveys in the Central District, 
and a team from Audubon North Carolina conducted 
surveys in the Southern District. Two scientists, one 
from each sampling team, met to discuss details of 
sampling protocols, and subsequently practiced and 
standardized techniques during a day on the water 
dedicated to standardize methods before sampling was 
conducted. 
The latitude and longitude of each target cell’s 
centroid were used to locate cells in the field with a 
handheld GPS (GPSMAP 76Cx or 60CSx, Garmin In- 
ternational, Inc., Olathe, KS). GPS also was used to 
determine the cell boundaries in relation to each cen- 
troid. This method of field demarcation enabled the 
definition of each 1-km 2 cell, within which both DCPs 
and AFCPs were counted. DCPs, which lacked surface 
buoys, were detected with the use of side-scan sonar, 
and AFCPs, which had one floating buoy per pot, were 
detected visually. 
Within each sampled 1-km 2 cell, 4 separate transects 
(30 m wide by 200 m long) were haphazardly chosen 
and surveyed by boat with no overlap of sampling area 
(for a total sampled area of 24,000 m 2 ) to detect DCPs 
with Humminbird 1197c high definition side imaging 
sonar (Johnson Outdoors Marine Electronics, Inc., Ra- 
cine, WI), combined with a GPS. Transect distances 
were verified by both one of the handheld GPSMAPs 
and the Humminbird sonar. Each team surveyed tran- 
sects in either a linear or a curved pattern that was 
tailored to the shape of estuarine contours within a 
given cell. In some cases, shallow water allowed DCPs 
to be located visually. During the same survey, all AF- 
CPs, with their required float visible on the water’s 
surface, were counted visually, within each 1-km 2 cell 
that was demarcated by using both the handheld and 
Humminbird GPSs. 
