Elsevier

Fisheries Research

Volume 163, March 2015, Pages 98-105
Fisheries Research

Tag shedding by tropical tunas in the Indian Ocean and other factors affecting the shedding rate

https://doi.org/10.1016/j.fishres.2014.02.025Get rights and content

Highlights

  • Estimate tag-shedding rates from double-tagging experiments.

  • Experience of taggers did not affect shedding rate estimates.

  • Constant-rate shedding model provided a better fit than the time-varying shedding rate model.

  • Difference in retention rate between the first and the second tag.

  • Exploitation rate was more biased by non-reporting than by shedding, but total mortality based on time-at-large was most impacted by continuous tag loss.

Abstract

A key objective of the Regional Tuna Tagging Project—Indian Ocean was to estimate tag-shedding rates, Type-I (immediate tag shedding) and Type-II (long-term tag shedding). To assess this, a series of double-tagging experiments (26,899 double tags released with 4555 recoveries) were conducted as part of the broader tagging program. After omitting data from tags placed by less experienced taggers, the results of our analyses did not show any evidence that individual differences between taggers (i.e., a tagger effect) impacted estimates of tag-shedding rates. However, it was shown that the probability of retaining the second tag (inserted in the left side of the fish) was larger than retaining the first tag (inserted in the right side, i.e., the side typically tagged in single-tagging experiments). We used a Bayesian model averaging approach to account for model uncertainty in the estimates of the parameters α and L used to calculate the probability of tag retention Q(t) = α e(L t) for the right tag. The parameter estimates were α = 0.993 and L (per year) = 0.030 (skipjack); α = 0.972 and L (per year) = 0.040 (yellowfin); and α = 0.990 and L (per year) = 0.021 (bigeye). These results agree with estimates obtained by other large-scale tropical tuna tagging projects. We showed that tag loss has a moderate impact on the underestimation of the exploitation rate (bias = 2–6% depending on the tuna species). However, non-reporting leads to a bias of around 7% when using the high reporting rate estimate of purse seiners. Finally, tag shedding (specifically Type-II shedding) modified the individual weights of the samples of recaptures. Consequently, the total instantaneous mortality estimates (Z; calculated from mean times-at-large) were reduced by a range of 1–3%.

Introduction

Mark-recapture techniques can facilitate the collection of useful information for stock assessments, such as stock structure, growth and mortality rates, gear selectivity, and migration patterns. Consequently, tagging studies have become one the key tools used by tuna Regional Fisheries Management Organisations (RFMOs) to improve understanding of how populations are spatially structured and the effects of fishing on these populations. Integral to the use of tagging data are standardization models, such as tag-attrition models for single release events (Kleiber et al., 1987, Hampton, 1997) or Brownie models (derived from bird-banding studies) for multiyear studies (Brownie et al., 1985, Hoenig et al., 1998, Polacheck et al., 2010). The results of tagging studies can, however, be compromised if tags or data are lost (i.e., through tag shedding and non-reporting). Both occurrences can lead to underestimations in tag-return rates, which create a negative bias in fishing mortality estimates, rates of fishery interactions, and tuna movements. Ultimately, this leads to biased estimates of stock status. Thus, the objective of this paper is to update preliminary estimates of tag-shedding rates by tropical tuna in the Indian Ocean (Gaertner and Hallier, 2008, Gaertner and Hallier, 2009).

There are two types of tag losses (Wetherall, 1982, Hampton and Kirkwood, 1990): Type-I losses, which reduce the number of tags initially put out (immediate tag shedding, immediate tagging mortality, and non-reporting), and Type-II losses which occur steadily over time (natural mortality, fishing mortality, permanent emigration, and long-term tag shedding). The current paper is only estimating the Type I and II tag shedding components of total losses. Immediate tag shedding and immediate tagging-induced mortality rates can be estimated by observing tagged fish under controlled laboratory conditions or in field cages (Pollock and Pine, 2007). However, post-release mortality estimates derived under these circumstances may be biased: in general, unlike wild fish, captive fish are not affected by post-release predation. On the other hand, the act of restraining fish in confined conditions can have lethal or sublethal effects. An alternative approach to estimating mortality that is commonly used is double-tagging experiments in which a fish is tagged with two tags simultaneously. Double tagging can also be used to estimate tag-shedding rates by identifying fish that have lost a tag.

In general, shedding rates cannot be estimated from tag-return data directly. Consequently, different methods have been proposed to estimate shedding rates from double-tagging experiments. To maximize the accuracy of these estimates, it is crucial to have a firm understanding of the functioning of the range of other variables known to impact tag shedding. The Regional Tuna Tagging Project—Indian Ocean (RTTP—IO), which focuses on tropical tuna in the Indian Ocean, has already examined some of these variables. For example, there is no evidence, or it remains unclear, whether factors such as tag length (11 cm and 14 cm length tags) or tag position (right side versus left side of the fish) influence the rate of tag returns in RTTP-IO double-tagging experiments (Gaertner and Hallier, 2008). However, multiple taggers have been used over the duration of the program, and tag-return rates are known to vary substantially between taggers. Therefore, in this context, it is desirable to estimate how shedding rates vary among taggers.

Consequently, this study focused on (1) an analysis of the tagger effect and other explanatory variables that were hypothesized a priori to influence tag loss, (2) comparing the constant-rate and time-varying approaches to modeling tag-shedding rates, (3) an analysis of how the insertion position of the tag affects shedding rates, and (4) an investigation into the consequences of ignoring tag shedding and non-reporting on the estimates of exploitation rate and total instantaneous mortality.

Section snippets

Data

Over the duration of the RTTP-IO, a number of different tag types (e.g., conventional, archival) and tag colors have been used. Tag colors are used to indicate the presence of other tag types. Conventional tags are traditionally yellow, but a white version is used if oxytetracycline is also injected into the fish, and a red version is used when an additional archival tag is inserted. In the RTTP-IO, the single- and double-tagging experiments have been alternately performed and always used the

Results

Although the beta-binomial model already naturally incorporates a certain pattern in the variances of the response, it can be necessary to also explore a variable dispersion model by incorporating an additional set of regressors in the precision sub-model. It then becomes possible to check whether the precision parameter ø is constant for all observations. With this in mind, the set of candidate models in this study were ranked on the basis of their BIC values, with the lowest value indicating

Discussion

A number of factors are known to affect tag-shedding rates. For example, it has been established that shedding rates may be impacted by the type of external tag used, with several studies having shown differences in the retention rates of dart and T-bar tags (Boucek and Adams, 2011). For tropical tunas, a comparison has been made between the conventional yellow ‘spaghetti’ tag and another tag commonly used by sport fishermen to opportunistically tag tunas and billfishes, the Betyp tag. Results

Author contributions

J-P. H. was the general coordinator of the RTTP-IO and checked the data; D.G. analyzed the data and led the writing, with J-P. H. contributing.

References (41)

  • C. Brownie et al.

    Statistical Inference from Band-Recovery Data: A Handbook

    (1985)
  • K.P. Burnham et al.

    Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach

    (2002)
  • N.G. Cadigan et al.

    Reporting and shedding rate estimates from tag-recovery experiments on Atlantic cod (Gadus morhua) in coastal Newfoundland

    Can. J. Fish. Aquat. Sci.

    (2006)
  • T.R. Carruthers et al.

    Estimating Tag Reporting Rates for Tropical Tuna Fleets of the Indian Ocean

    (2013)
  • D.G. Chapman et al.

    A method for estimating the rate of shedding of tags from yellowfin tuna

    Inter-Am. Trop. Tuna Comm. Bull.

    (1965)
  • F. Cribari-Neto et al.

    Beta regression in R

    J. Stat. Softw.

    (2010)
  • D.R. Diefenbach et al.

    Modeling and evaluation of ear tag loss in black bears

    J. Wildl. Manage.

    (1998)
  • S.L.P. Ferrari et al.

    Beta regression for modelling rates and proportions

    J. Appl. Stat.

    (2004)
  • D. Gaertner et al.

    Combining Bayesian and simulation approaches to compare the efficiency of two types of tags used in tropical tuna fisheries

    Aquat. Living Resour.

    (2004)
  • D. Gaertner et al.

    Tag shedding by tropical tunas in the Indian Ocean: explanatory analyses and first results

    Doc IOTC, WPDTA-01.

    (2008)
  • Cited by (12)

    • Approaches for estimating natural mortality in tuna stock assessments: Application to global yellowfin tuna stocks

      2023, Fisheries Research
      Citation Excerpt :

      Times at liberty of tagged tunas, in combination with approximate ages at release, can also provide information about mortality, and maximum time at liberty is in some ways analogous to maximum observed age. For recaptured yellowfin tuna, the longest periods at liberty are lower than the maximum observed ages from otoliths, at 6.5 years in the WCPO (Eveson et al., 2020), 5 years in the EPO (Schaefer and Fuller, 2022a), 9.1 years in the Atlantic (Fonteneau and Hallier, 2015), and 6.0 years in the Indian Ocean (Fonteneau and Hallier, 2015; Gaertner and Hallier, 2015) for fish mostly tagged at about age 1 year, although some tuna were tagged at older ages. The rankings by stock of maximum time at liberty match those of maximum observed age for the stocks aged using annuli (i.e. all except the EPO).

    • Tag-shedding rates for tropical tuna species in the Atlantic Ocean estimated from double-tagging data

      2022, Fisheries Research
      Citation Excerpt :

      The results of tagging studies can, however, be compromised if tags or data are lost (i.e., through tag-shedding and non-reporting). Both occurrences can lead to underestimations in tag-return rates, which create a negative bias in fishing mortality estimates, rates of fishery interactions, and tuna movements (Gaertner and Hallier, 2015). Ultimately, this leads to biased estimates of stock status.

    • Influence of oceanographic and climatic variability on the catch rate of yellowfin tuna (Thunnus albacares) cohorts in the Indian Ocean

      2020, Deep-Sea Research Part II: Topical Studies in Oceanography
      Citation Excerpt :

      In the Indian Ocean, catches of yellowfin tuna are distributed in all the tropical waters and primarily concentrated in western tropical waters, the Arabian Sea, and the north of the Mozambique Channel (Fu et al., 2018). Tag recoveries provide evidence of large movements of yellowfin tuna within the western equatorial region; however, few observations of large-scale transverse movements in the Indian Ocean have been reported (Gaertner and Hallier, 2015). This may indicate that the western and eastern regions of the Indian Ocean support relatively distinct subpopulations of yellowfin tuna.

    • Tuna aggregation and movement from tagging data: A tuna "hub" in the Indian Ocean

      2015, Fisheries Research
      Citation Excerpt :

      It was like a “Tuna Hub” similar to an airline hub with planes landing and taking off in large numbers, quickly and in different directions. High tag-induced mortality cannot be part of the explanation as Gaertner and Hallier (2015) demonstrated that immediate and continuous tag shedding and tag-induced mortality of the RTTP-IO tagged tuna are low and similar to other large-scale tuna tagging programs. Tuna dispersal and multiple recoveries of the same tuna showed some interesting points such as:

    View all citing articles on Scopus
    View full text