20 
Fishery Bulletin 1 12(1) 
marine mammals in order for stranding data to serve 
as reliable proxies of those source populations. 
Some degree of variability in the number of stand- 
ings is expected among years, months, and locations 
owing to the variety of factors that can affect the 
likelihood that an animal dying is beach-cast, or that 
carcasses persist on shore (without being washed out 
or buried by wave action) long enough to be observed, 
reported, and recovered. Winds and currents affect 
stranding rates (Peltier et al., 2012) and it would be 
interesting to investigate their influences on the pat- 
terns documented here. For example, southern NC had 
relatively few strandings, particularly of species other 
than coastal bottlenose dolphins. It is also farthest 
from the Gulf Stream and has a large estuary plume 
outflowing from the Cape Fear River (Xia et al., 2007), 
both of which likely impact stranding rates. 
Increased mortality due to human interactions also 
affects stranding patterns. These effects are not al- 
ways discernible because the ability to detect HI varies 
across the nature of the interaction. Evidence of fishery 
interactions, mutilation, vessel strikes, and gunshot 
wounds are relatively obvious to trained responders. In 
contrast, the detection of sonar effects on the presence 
and health of stranded marine mammals is challeng- 
ing and requires a much more sophisticated sampling 
protocol than can be implemented for most strandings 
recovered in NC. The sampling protocol requires fresh 
carcasses, expertise of the responders, and availability 
of resources for histopathology analysis and computer- 
ized tomography (CT scanning) ( see Cox et al., 2006). 
Insufficient data exist to comment on the prevalence of 
sonar exposure as a cause of strandings for the current 
study. Strandings positive for HI provide much needed 
information about the nature, timing, and frequency 
of interactions, especially in light of the limitations 
of fisheries observer programs to sufficiently cover all 
fisheries (Byrd et al., 2008). Although human interac- 
tions were detected in more than half of all species re- 
covered ( 18 of 34), the number of strandings positive 
for HI was likely underestimated because of the rela- 
tively large number of strandings assigned to HI-CBD 
as result of decomposition, scavenger damage, and a 
conservative approach to assigning HI status. Hl-no 
is the most difficult assignment to make. For example, 
strandings with questionable lesions would be assigned 
by default to HI-CBD. 
Maintaining the quality and consistency of strand- 
ing data is not a simple task. The stranding network in 
NC, with its extensive coastline, relies heavily on pub- 
lic reporting and therefore ties to state, federal, and 
local municipalities have been key to receiving reports 
of, and in some cases gaining access to, strandings. 
The availability of trained participants to respond to 
stranding reports has also been vital to the collection 
of irreplaceable data and samples. 
Conclusions 
Marine mammal strandings in NC from 1997 to 2008 
reflected the rich biodiversity occurring in waters off 
this unique location, where ‘northern’ and ‘southern’ 
species as well as coastal and pelagic species inter- 
sect. Therefore the spatial and temporal patterns de- 
tected from strandings can provide clues to the pres- 
ence of living animals occurring off the NC coast. In 
some cases, those patterns reflect what is known from 
published records of aerial and shipboard surveys. For 
other species, little is known and stranding data serve 
as a proxy for live animal distribution. In addition, the 
detection of HI, particularly FI, provides crucial infor- 
mation on the spatiotemporal patterns and relative 
mortality levels from these interactions with marine 
mammals which are otherwise difficult to obtain in situ 
(Friedlaender et al., 2001; Byrd et al., 2008). Moreover, 
changes in stranding patterns can serve as indicators 
of underlying change in source populations due to an- 
thropogenic or naturally occurring events. 
Acknowledgments 
The collection of stranding data requires assistance 
from many professionals and volunteers. We are grate- 
ful for their participation, particularly network contrib- 
utors from the authors’ institutions and the National 
Park Service, NC Division of Environment and Natural 
Resources, NC Division of Parks and Recreation, NC 
Maritime Museum, NC Sea Turtle Project, U.S. Coast 
Guard Sector NC, U.S. Fish and Wildlife Service, U.S. 
Marine Corps at Camp Lejeune, and municipalities 
in Carteret, Currituck, and Dare Counties. Statistical 
advice was provided by K. Shertzer (NMFS-Beaufort). 
The manuscript was improved by reviews of K. Craig 
(NMFS-Beaufort), S. Horstman (NMFS, Southeast Re- 
gional Office, St. Petersburg, FL), J. Litz (NMFS, South- 
east Fisheries Science Center, Miami, FL), D. A. Pabst 
(University of North Carolina Wilmington, Wilmington, 
NC), K. Shertzer, and 3 anonymous reviewers. Strand- 
ing response was authorized by the NMFS pursuant 
to Sections 109(h) and 112(c) of the Marine Mammal 
Protection Act. 
Literature cited 
Adams, L. D., and P. E. Rosei. 
2006. Population differentiation of the Atlantic spot- 
ted dolphin (Stenella frontalis) in the western North 
Atlantic, including the Gulf of Mexico. Mar. Biol. 
148:671-681. 
Baker, P, J. D. Austin, B. W. Bowen, and S. M. Baker. 
2008. Range-wide population structure and history of 
the northern quahog ( Merceneria merceneria) inferred 
from mitochondrial DNA sequence data. ICES J. Mar. 
Sci. 65:155-163. 
