Voss et al.: Factors driving the density of derelict crab pots and associated bycatch in North Carolina 
383 
Table 2 
Condition of derelict crab pots was ranked on a scale of 1-10. This table provides the de- 
scriptors used to rank the condition of pots. 
Condition rank 
Descriptors 
1 
Structural integrity of pot sound; some rust or discoloration may be seen 
2 
Pot retains original shape but shows wear and abrasion 
3 
Pot structure intact, with obvious flaws in wall or floor 
4 
Pot likely misshapen; walls or floor beginning to degrade 
5 
Observable decay of structure, with pits in walls or floor 
6 
Small voids in walls or floor 
7 
Walls or floor are clearly deteriorated; partial collapse likely 
8 
Holes >10 cm in walls or floor 
9 
Structural integrity of pot has failed; walls or floor well decayed 
10 
Total dilapidation; parts of walls or floor well disintegrated 
Bivalve size was measured by the perpendicular axes 
of height, from umbo to distal tip, and length, the longest 
axis from anterior to posterior ends. All size measure- 
ments were recorded to the closest 1 mm. After all data 
were collected and all animals were released from a pot, 
it was flattened and taken for proper disposal. 
To estimate the recruitment rate of DCPs in areas 
where crabbing commonly occurs in North Carolina, 
we carefully removed all DCPs from four 1-km 2 cells, 
one cell each in Bogue Sound (near Archer’s Creek) and 
in the eastern area of Newport River in the Central 
District and in Topsail Sound (behind Figure Eight Is- 
land), and Masonboro Sound (behind Masonboro Island) 
in the Southern District in spring 2010. These 4 cells 
were intensively resurveyed for newly recruited DCPs 
after 1 year (52 and 51 weeks later in the Central and 
Southern Districts, respectively). We worked with the 
North Carolina Marine Patrol of the NCDMF to ensure 
that the 4 areas used for this study component were >5 
km from state waters included in the NCDMF program 
for removing and disposing of lost and abandoned crab 
pots (Anthony and Knudsen 11 ). 
Data analyses 
Data on DCPs, AFCPs, coordinates for surveyed tran- 
sect area, and DCP bycatch and attached organisms 
for all 6 waterbodies were combined. SYSTAT, vers. 
13.00.05 (Systat Software, Inc., San Jose, CA) and 
JMP, vers. 8.0 (SAS Institute, Inc., Cary, NC) software 
packages were used for statistical analyses. Because 
the area of some sampled 1-km 2 cells included land, 
we used a conversion factor to standardize the data to 
the proportion of open water in each cell as computed 
by ArcMap. Here, the number of DCPs found in each 
24,000-m 2 area or ACFPs in a 1-km 2 cell was multi- 
n Anthony, S., and H. Knudsen. 2010. Personal commun. 
Marine Patrol, Div. Mar. Fish., North Carolina Dep. Environ. 
Nat. Resour., Morehead City, NC 28557. 
plied by the proportion-open-water conversion factor, 
with DCPs also multiplied by 41.667, to calculate the 
density of DCPs and AFCPs as the number of pots 
per square kilometer of open water. Density values for 
both DCPs (k: -0.6) and AFCPs (X\ -0.2) were Box-Cox 
transformed to preclude significant departure from ho- 
mogeneity of variance as confirmed afterward by either 
O’Brien’s or Bartlett’s tests (a=0.05). 
Two-way analyses of variance (ANOVAs) were con- 
ducted on data to test whether waterbody or habitat 
type were significant factors explaining differences in 
DCP and AFCP density and to test whether interactions 
between habitat type and waterbody existed. Tukey’s 
honestly significant difference tests were conducted post 
hoc to examine differences in DCP and AFCP densi- 
ties among the waterbodies and habitats we sampled. 
Because of the lack of a linear relationship between 
DCPs and AFCPs, analysis of covariance was unwar- 
ranted. Two-way ANOVAs were used to compare bycatch 
abundance of blue and Florida stone crabs by waterbody 
and habitat type, as well as to determine whether an 
interaction between factors existed. The bycatch abun- 
dance data for both blue and Florida stone crabs were 
Box-Cox transformed to meet normality criteria before 
ANOVAs. To investigate the effects of including cells 
that represented more than one habitat type, results 
from 2-way ANOVAs that included all (n=201) cells were 
compared with 2-way ANOVAs that excluded the cells 
that served for 2 habitat types (n=155). In cases where 
the arithmetic mean is given, the standard deviation 
(SD) is also presented in parentheses. 
Results 
Density of crab pots 
The 201 cells within the selected 6 waterbodies in 
North Carolina were scanned with a side-scan sonar 
and 106 DCPs were detected (Table 1). By extrapolat- 
