A novel application of multi‐event modeling to estimate class segregation in a highly migratory oceanic vertebrate. Issue 12 (2nd December 2016)

• Record Type: Journal Article
• Authors: Chapple, T. K.; Chambert, T.; Kanive, P. E.; Jorgensen, S. J.; Rotella, J. J.; Anderson, S. D.; Carlisle, A. B.; Block, B. A.
• Is Part Of: Ecology. Volume 97:Issue 12(2016)
• Page Start: 3494
• Page End: 3502
• Publication Date: 2016-12-02
• Subjects: capture probability -- class segregation -- mark‐recapture -- multi‐event -- robust design -- sexual segregation -- size segregation -- temporary emigration -- white shark; Ecology -- Periodicals; Ecology -- Periodicals; Écologie -- Périodiques; Ecologie; Écologie; Écologie animale; Écologie végétale; Ecology; Periodicals; 577.05
• DOI: 10.1002/ecy.1589
• Languages: English
• ISSNs: 0012-9658

Abstract

Abstract: Spatial segregation of animals by class (i.e., maturity or sex) within a population due to differential rates of temporary emigration (TE) from study sites can be an important life history feature to consider in population assessment and management. However, such rates are poorly known; new quantitative approaches to address these knowledge gaps are needed. We present a novel application of multi‐event models that takes advantage of two sources of detections to differentiate temporary emigration from apparent absence to quantify class segregation within a study population of double‐marked (photo‐identified and tagged with coded acoustic transmitters) white sharks ( Carcharodon carcharias ) in central California. We use this model to test if sex‐specific patterns in TE result in disparate apparent capture probabilities ( p o ) between male and female white sharks, which can affect the observed sex ratio. The best‐supported model showed a contrasting pattern of Pr(TE) from coastal aggregation sites between sexes (for males Pr[TE] = 0.015 [95% CI = 0.00, 0.31] and Pr[TE]= 0.57 [0.40, 0.72] for females), but not maturity classes. Additionally, by accounting for Pr(TE) and imperfect detection, we were able to estimate class‐specific values of true capture probability ( p * ) for tagged and untagged sharks. The best‐supported model identified differences between maturity classes but no difference between sexes or tagging impacts (tagged mature sharks p * = 0.55 (0.46–0.63) and sub‐adult sharks p * = 0.36 (0.25, 0.50); and untagged mature sharks p * = 0.50 (0.39–0.61) and sub‐adults p * = 0.18 (0.10, 0.31). Estimated sex‐based differences in p o were linked to sex‐specific differences in Pr(TE) but not in p * ; once the Pr(TE) is accounted for, the p * between sexes was not different. These results indicate that the observed sex ratio is not a consequence of unequal detectability and sex‐specific values of Pr(TE) are important drivers of the observed male‐dominated sex ratio. Our modeling approach reveals complex class‐specific patterns in Pr(TE) and p * in a mark‐recapture data set, and highlights challenges for the population modeling and conservation of white sharks in central California. The model we develop here can be used to estimate rates of temporary emigration and class segregation when two detection methods are used.