Musyl et al.: Postrelease survival, vertical and horizontal movements, and thermal habitats of five species of pelagic sharks 
359 
The most striking diel vertical movement behavior 
among pelagic shark species was observed in bigeye 
threshers. Our observations were similar to those of 
Nakano et al. (2003) who acoustically tracked two im- 
mature females (175 and 124 cm precaudal length) for 
96 and 70 h, respectively; their vertical movements 
were centered between 200 and 500 m during the day 
and 80 and 130 m at night. Moreover, the diel vertical 
movement patterns we observed were comparable to the 
PSAT data reported by Weng and Block (2004). 
Movement data of the shortfin makos that we ob- 
served were similar to those recorded by Loefer et al. 
(2005) for this species in the Atlantic, in that both 
studies recorded adjustment of vertical behavior when 
the sharks entered water masses with different thermal 
characteristics. However, shortfin makos in the Atlantic 
made excursions from the surface to 556 m (tempera- 
tures from 10.4° to 28.6°C), whereas we never observed 
movements below -441 m. 
Thermal niche partitions and habitat structure 
Our results show that pelagic shark species display 
distinct thermal niche partitioning (as identified by 
UPGMA clustering) and that habitat structure for the 
epipelagic silky and oceanic whitetip sharks can be 
adequately estimated from two dimensions (these spe- 
cies spend most of their time in the warmest available 
water). By contrast, three dimensions will be required 
to describe the extended vertical habitat of the species 
that we classified as mesopelagic I (blue sharks, shortfin 
makos) and mesopelagic II (bigeye threshers). 
Except for the oceanic whitetip shark and silky shark 
clusters, which showed familial affinities based on phy- 
logeny and life history, the topology of the dendrogram 
for pelagic shark species appeared to correlate with 
body size and latitudinal gradient, but not with phy- 
logeny (Shirai, 1996), life history (Cortes, 2000), eco- 
morphotype (Compagno, 1990), neural anatomy (Lisney 
and Collin, 2006; Yopak and Montgomery, 2008; Yopak 
and Frank, 2009), relative eye size (Lisney and Collin, 
2007), or the presence of regional endothermy (Bernal 
et al., 2001; Dickson and Graham, 2004). It also does 
not appear that clustering was greatly influenced by 
the El Nino-Southern Oscillation (www.esrl.noaa.gov/ 
psd/people/klaus.wolter/MEI/, accessed November 2010) 
or Pacific Decadal Oscillation (cses.washington.edu/cig/ 
pnwc/compensopdo.shtml, accessed November 2010) 
climate patterns. 
Dickson and Graham (2004) argued that endothermy 
per se was not required for niche expansion and that 
other adaptations were necessary to allow for verti- 
cal movements below the thermocline. This hypothesis 
implies that other factors (e.g., ontogeny, latitude, lo- 
comotion, diet, and dimensionality of the environment) 
probably influence thermal niche partitions (Yopak and 
Montgomery, 2008; Yopak and Frank, 2009). Dietary 
studies based on accumulation of mercury in prey items, 
which is depth-dependent, have revealed vertical niche 
preferences among pelagic species (Choy et al., 2009). 
Dagorn et al. (2000) suggested, on the basis of their 
simulation model, that “different solutions for exploit- 
ing the same environment” had evolved among tropical 
pelagic species; their findings reflected a diverse array 
of species-specific vertical movement patterns and ver- 
tical niche partitions similar to those observed in our 
study on pelagic sharks. Numerous authors (e.g.. Brill 
et al., 2005; Bernal et al., 2009; Musyl et al. 2 ; and oth- 
ers) have suggested that evolution of the ability to make 
extensive daily vertical movements in pelagic species 
may have arisen from predator-prey dynamics. In other 
words, predator and prey may be locked in a physi- 
ological race driving the biological and physiological 
adaptations and tolerances of both and thus expanding 
their vertical niche. 
For comparative purposes, shark species from other 
locations could be analyzed with our clustering methods 
to determine thermal niche clusters. From a practical 
standpoint, pelagic shark species that form thermal 
clusters may also experience similar fishing pressures 
and this association may have direct application to miti- 
gating bycatch. For example, from longline catch data in 
the Atlantic, Rey and Munoz-Chapuli (1992) calculated 
that blue sharks were more likely to be captured in as- 
sociation with shortfin makos rather than with bigeye 
threshers and this calculation supported our groupings 
in the cluster analysis. 
Conclusions 
Results from PSAT tagging indicate that pelagic shark 
species can have high survival rates when released 
alive from longline fishing gear, and therefore catch- 
and-release may be a viable option to protect parental 
biomass in this fishery. Additional research is warranted 
to determine which biological and anthropogenic fac- 
tors correlate with at-vessel and postrelease survival. 
Furthermore, information on the temporal and spatial 
vertical distribution patterns and community structure 
of pelagic species can assist in the formulation of man- 
agement strategies to modify fishing gear, and thus 
reduce bycatch. This information should also provide 
more confidence in predicting catch rates and the species 
captured in different gear types by managers regulating 
fishing practices. As the tools and techniques for dif- 
ferentiating postrelease mortality become more refined 
allowing for larger sample sizes, it should be feasible to 
design fishing methods and practices that significantly 
reduce bycatch mortality. 
Acknowledgments 
This project was funded by Cooperative Agreements 
NA37RJ0199 and NA67RJ0154 of the National Oce- 
anic and Atmospheric Administration (NOAA) with the 
Joint Institute for Marine and Atmospheric Research 
(JIMAR), University of Hawaii. We thank crew and 
officers of the NOAA RV Townsend Cromwell and Oscar 
