Longevity of Atlantic Sharpnose Sharks Rhizoprionodon terraenovae and Blacknose Sharks Carcharhinus acronotus in the western North Atlantic Ocean based on tag-recapture data and direct age estimates

Longevity of Rhizoprionodon terraenovae and Carcharhinus acronotus in the western North Atlantic Ocean was examined using direct age estimates from vertebral sections and tag-recapture data. Time-at-liberty ranged from 7.7-14.0 years (mean =10.1) for R. terraenovae and 10.9-12.8 years (mean =11.9) for C. acronotus. Maximum estimated longevity was determined to be 19.8 years through tag-recapture data and 18.5 years from direct age estimates for R. terraenovae and 22.8 years through tag-recapture data and 20.5 years through direct age estimates for C. acronotus. These longevity estimates represent a large increase over previous estimates and may have significant effects on analyses that depend on longevity including lifetime fecundity, mortality rates, demographic analyses and stock assessments.


This article is included in the Elasmobranch
Biology & Conservation Collection Tag-recapture data grant researchers the opportunity to synthesize a vast array of information for species they are studying. Sharks, in particular, are well suited for tagging studies due to their migratory behavior, longevity, and relatively large size, which allows tagging at all life stages (Bonham et al., 1949). Valuable information gained from recapture data includes greater understanding of species-specific migratory patterns, stock structure, spatial and temporal distribution, site fidelity/residence times, and life histories (Kohler & Turner, 2001).
Among life history parameters needed to best manage populations of fishes, robust longevity estimates are of paramount importance, particularly for iteroparous species, such as sharks. For example, assuming age at maturity, reproductive periodicity, and brood size are constant, the lifetime fecundity of a given species is directly linked to its lifespan. Longevity estimates for sharks are generally derived from von Bertalanffy Growth Function (VBGF) parameter estimates (i.e. growth constant); however, VBGF parameter estimates can be heavily affected by low sample sizes, incomplete sampling among size classes and/or difficulties associated with age estimation of large individuals (Goldman, 2004;Francis et al., 2007). Obtaining accurate longevity estimates can be hampered by the low probability of catching older individuals as they represent a small portion of the entire population, difficulty associated with capturing large specimens that are more capable of escaping gear than smaller conspecifics, and the reduction of older individuals in the population due to fishing pressure (Bishop et al., 2006 Upon capture, the fork length (FL) and sex of each shark was recorded. For those sharks deemed to be in good health, a Hallprint © nylon dart tag, labeled with unique identifying numbers, was inserted into the dorsal musculature at the base of the first dorsal fin prior to release. As tagged sharks were at liberty in the wild, recaptured specimens were reported by a number of sources, including recreational anglers, commercial fishermen and SCDNR. Estimated measurements at recapture were obtained from recreational and commercial fishermen and direct measurements were recorded from specimens recaptured by SCDNR. When possible, a sample of 8-10 vertebrae were removed from the cervical region of the vertebral column from recaptured sharks.
In the case of recaptured specimens from commercial fishermen, sharks were already deceased and vertebrae were removed. In the case of recaptured specimens from SCDNR, a IACUC protocol approved for graduate students who had previously worked with SCDNR on elasmobranch studies was followed; the vertebral column was severed by serrated knife in two cervical locations.
To estimate age at recapture, vertebrae were prepared for analyses following the protocol of Frazier et al., 2014. Initial vertebral growth band counts were conducted by three unbiased readers with no knowledge of specimen length or time-at-liberty. If band counts differed among readers, sections were independently re-read until agreement was reached. Age at recapture was estimated under the assumptions: (1) the birthmark was formed prior or shortly after parturition, (2) the second band was formed 6 months later during the first winter, and (3) the third band was formed 1 year later. Therefore, we subtracted 1.5 from each total band count to calculate age in years (e.g. Loefer & Sedberry, 2003;Driggers et al., 2004).
Age estimates at recapture were also determined by adding timeat-liberty to backtransformed ages at tagging. Species and sexspecific von Bertalanffy growth function parameters from Loefer & Sedberry, 2003 (R. terraenovae) and Driggers et al., 2004 (C. acronotus) were used to backtransform age at tagging using the following equation: Where: L t = length at age t, L ∞ = theoretical maximum length, k = coefficient of growth, t o = theoretical age at which length equals zero.

Amendments from Version 1
This research note has been updated to incorporate suggestions by the reviewers. Figures have been added of the oldest aged vertebra section for each species. All mention of Faben's theoretical maximum age have also been removed from the note. An additional Rhizoprionodon terraenovae recapture was encountered and those data were added to the research note (Shark L2881). Also background information about the number of each species tagged and recaptured was added.

REVISED
A paired t-test was used to determine if age estimates from vertebral sections were significantly different than backtransformed age estimates. Statistical results were considered significant at α < 0.05.

Rhizoprionodon terraenovae
From 1993 to 1998 a total of 3,419 R. terraenovae were tagged and released, of these 155 were recaptured. Five R. terraenovae (four male and one female) were recaptured with times at liberty ranging from 7.7 to 14.0 years (mean ± S.D. = 10.1 ± 2.7). Length at initial tagging, and when available, length at recapture are listed in Table 1. Backtransformed age at tagging and recapture ranged from 3.8-10.3 years and from 11.8 to 19.8 years, respectively (Table 1). Vertebrae were sampled from Shark L5242 and the direct age estimate from the vertebral section was 18.5 years old (Figure 1). Precise measurements were taken for L5242 at tagging and recapture; over 12.1 years-at-liberty this shark grew 28 mm (2.3 mm/year). All five sharks were recaptured within 15 km of initial tagging.

Carcharhinus acronotus
From 1993 to 2013 a total of 1,537 C. acronotus were tagged and released, of these 24 were recaptured. Three C. acronotus were recaptured (two males and one female) with times at liberty ranging from 10.9 to 12.8 years (mean ± S.D. = 11.9 ± 1.0). Backtransformed age at tagging and recapture ranged from 4.2 to 11.9 years and 16.4 to 22.8 years, respectively (Table 1). Vertebrae were obtained from all recaptured C. acronotus. Age estimates from vertebral sections ranged from 14.5 to 20.5 years ( Figure 2). Precise measurements were taken at capture and recapture for sharks L2910 and L1515. L2910 was at liberty for 12.0 years and grew 84 mm (7.0 mm/year). L1515 was at liberty for 12.8 years and grew 177 mm (13.8 mm/year). All recaptured C. acronotus were recovered within 15 km of initial tagging.

Age estimates and comparisons
In recaptures with both direct and backtransformed age estimates, backtransformed age estimates were significantly larger (paired t-test, t = 4.82, P = 0.02) (Table 1). From direct age estimates, R. terraenovae observed maximum longevity increased by 9.5 years for males (previously 9+ years [Loefer & Sedberry, 2003]), and no vertebrae were available from female recaptures ( Table 2).
C. acronotus observed maximum longevity from sectioned age estimates increased by 2 years for females and 8 years for males (12.5 and 10.5 years for females and males respectively [Driggers et al., 2001]). Backtransformed maximum longevity was 4.5 years older than published estimates for females and 12.3 years older for males.

Discussion
The backtransformed and direct age estimates documented herein greatly increase the known longevity for R. terraenovae and C. acronotus, which could significantly affect population dynamics models that include longevity as a parameter. Interestingly, the maximum directly estimated ages for both species were associated with male sharks. A review of published shark age and growth studies shows that, in most studies, females have a greater or equal longevity than males (e.g. Carlson & Baremore, 2003;Carlson & Parsons, 1997;Carlson et al., 1999;Driggers et al., 2004;Drymon et al., 2006;Frazier et al., 2014). Therefore, we believe that we did not sample older females and suggest that the longevity estimates we observed for each species be applied to males and females.
Backtransformed age estimates from recaptures were in all cases larger than direct age estimates. This could be evidence that direct ages were underestimated, however direct age estimates only differed by an average of 1.8 years (range 1.0-2.5 years). Observed differences could be due to low sample size or individual variability in growth. The rearranged VBGF gives an estimate of average age-at-length based on parameter estimates. Therefore, specimens could have been younger or older than the average estimated age at time of tagging thus accounting for the discrepancy. Conversely, if the species-specific VBGF do not adequately describe the growth of R. terraenovae and C. acronotus then the observed differences would be expected. However, the growth models utilized were from studies with robust sample sizes that included all size classes (Loefer & Sedberry, 2003;Driggers et al., 2004).
The age estimates from this study greatly increase longevity for both species; however, actual life spans could be even longer. The specimens recaptured from this study were tagged in the first few years of the SCDNR longline program (1994)(1995)(1996). Given project species-specific recapture rates of less than 3% and reported tag shedding rates for nylon dart tags as high as 41% to 63% (Xiao et al., 1999), the chances of recapturing a shark at liberty for 10+ years are small. The fact that eight were encountered and all exceeded published longevity estimates lends support to this assertion. The data gathered from these recaptures also highlights the importance of continuing long term surveys and tagging efforts. While return rates of tagged sharks are notoriously low, our data demonstrate that continued tagging efforts are essential to provide the most up to date and reliable estimates of maximum longevity. Ideally, to obtain the best possible longevity estimates for a given species, future efforts should focus on tagging neonate sharks with the hope of recapturing them as they approach the end of their lifespans. Carcharhinus acronotus tag recapture data. These findings are particularly interesting given the previously assumed longevity of these species, and are clearly relevant to population models that include longevity as a parameter for these species. This note is clearly written, and conclusions from this work are appropriately presented; moreover, these data illustrate the benefit of long-term fisheries-independent monitoring and tagging programs for long-lived species such as sharks.
These data are clear, concise, and important to the continued successful management of these two species; as such, they are worthy of publication. The two tables presented are sufficient to convey the data discussed. Given the straightforward nature of these data and their appropriate interpretation by the Authors, I have no major comments. There are two typos in the first paragraph of the Discussion (detailed below) which should be addressed.
In the sentence "A review of published shark age and growth studies shows that, in most studies, females have a greater…," remove that that.
In the sentence "Therefore, we believe that we did not sample older females and suggest that the 2. In the sentence "Therefore, we believe that we did not sample older females and suggest that the we observed for each species…," revise " longevity estimates maximum observed longevity as desired. longevity estimates maximum observed longevity" I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. This research note presents data on the long-term recapture of seven individuals of two species of coastal sharks from the Northwest Atlantic; , and the .

Rhizoprionodon terranovae
Carcharhinus acronotus Based on their length at tagging, all individuals appeared to be either mature or close to maturity, but were recaptured between 7.7 and 12.8 years later. As neither species was reported to be particularly long-lived these results are interesting, and provide evidence that both species live longer than currently thought. Age was also estimated and compared in four of the samples by reading sectioned vertebrae. These data contribute to growing evidence that longevity of sharks is often underestimated when based on vertebral ageing methods, and that some aspects of elasmobranch demography that have been inferred from these studies (e.g. survival) may be biased. I'm glad the authors have taken the time and effort to document these observations, as there are relatively few long-terms studies of this kind.
While I think the data themselves are worthy of publication, I don't think any of the subsequent analysis contributed hugely to the paper. Published growth curves are used to estimate the age at tagging, which is then added to the time at liberty to get a 'backtransformed age at recapture'. Unfortunately, neither of these original growth studies presented the model variance, a detailed description of the data, or even any plots of the size at age data! Without this information, it is hard to get an idea of how variable size-at-age is for these species (and it is clearly quite variable), or the representativeness of the original sampling, etc. Personally, I think other information could be more useful, like photographs of the two oldest vertebrae sections for reference. I recommend two revisions to the research note. Firstly, some more background information is required on the tagging study itself. It is not stated how many sharks were tagged and recaptured over the duration of the whole study, nor is it clear why these seven were chosen specifically (presumably they weren't the only recaptures). Secondly, I disagree that Fabens' theoretical method for estimating longevity is the usual method of estimating longevity. I don't see the value with making comparisons to theoretical