Jurnal Populasi Print

7/21/2019 Jurnal Populasi Print

http://slidepdf.com/reader/full/jurnal-populasi-print 1/7

Fisheries Research 129–130 (2012) 1–7

Contents lists available at SciVerse ScienceDirect

Fisheries Research

journal homepage: www.elsevier .com/ locate / f ishres

Tag loss in the lobster Palinurus elephas (Fabricius, 1787) and implications forpopulation assessment with capture-mark-recapture methods

Lucía González-Vicente∗, David Díaz, Sandra Mallol, Raquel Goni

CentroOceanográficode Baleares, IEO, Muelle de Poniente s/n07015 Palma deMallorca, Spain

a r t i c l e i n f o

Article history:

Received 3 February 2012

Received in revised form 25 May 2012Accepted 26 May 2012

Keywords:

Capture-mark-recapture

Double tagging

Tag loss

Return rate bias

Palinurus elephas

a b s t r a c t

Assessments of animal populations by mark-recapture methods rest on the assumption that all marked

animals are recognizable, yet a variety of processes, such as molting in crustaceans, can lead to tag

shedding. In this paper we estimate the shedding rate of T-bar anchor tags in the European spiny lobster

Palinuruselephas from double tagging experiments conducted in the wild. During four annual consecutive

surveys (1999–2002) lobsters were caught, double-tagged and released. The evolution of the subsequent

double- and single-tag recaptures was traced during the following ten years (2000–2009). The estimated

instantaneous rate of tag loss was 6.8% year−1 in males and 5.0% year−1 in females tagged during inter-

molt. These tag-shedding rates are low compared to most published values for similar species. Data also

suggested that there is poorer tag retention when tagging immature individuals or during premolt. We

simulate and discuss the impact of undetected recaptures on the return rate and the implications for

population assessment in the particular case of P. elephas.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Capture-mark-recapture (CMR) methods are widely used tostudy wild animal populations (Williams et al., 2002) since they

allow the estimation of population parameters such as abundance,

survival and recruitment. For marine exploited populations, CMR

techniques can be applied to determine the response of a popula-

tion to the introduction of marine reserves (or ‘marine protected

areas’, MPA), allowing the assessment the closure effectiveness on

population recovery and the estimation of movement and biomass

exportation outside the reserve (e.g. Afonso et al., 2008; Goni

et al., 2006, 2010). The accuracy of parameter estimates such as

abundance and survival obtained by CMR methods relies on the

correctness of assumptions, including the accurate identification

of tagged specimens, their traceability over time and complete tag

reporting (Lebreton et al., 1992; Seber, 1986).

However some recaptures could be missed as result of tagdetachment. Tagloss canoccur shortly after tagging as resultof tis-

sue rejection or insufficient anchoring of the tag (“type I tag loss”),

or tags maybe lost overtimeas resultof the normalaging, wear and

tear or molting (“typeII tagloss”) (Beverton and Holt, 1957; Parker,

1962). Hence, the use of CMR techniques for estimating population

parameters is hampered since this removal from the tagged pool is

indistinguishable from mortality or emigration.

∗ Corresponding author.

E-mail address: [email protected] (L. González-Vicente).

In this study we assess the rate of tag loss of T-bar anchor tags

in the European spiny lobster Palinurus elephas (Fabricius, 1787)

basedon double-taggingexperiments in the wild. The study has thedouble aimof quantifying therates of short-and long-term tagloss

andevaluatinghow tagloss affects returnratesand itsimplications

for population assessments, using P. elephas for illustration.

2. Materials and methods

2.1. Study area and species

The study area is situated in the Western Mediterranean and

includes the Columbretes islands marine reserve (hereafter “the

MPA”) and its adjacent fishing grounds. The MPA was established

in 1990 to protect species such as P. elephas, the most commer-

cially important spiny lobster species in the Northeast Atlantic and

Mediterranean.P. elephas is a long-lived, slow growing species that

in the study region matures at about 80mm carapace length (CL)

(Goni et al., 2003a) or 3–4years (Díaz, 2010). Due to its long life-

span and limited adult mobility, along with its high unit value and

excessive fishing effort, P. elephas populations are heavily over-

fished (Goni and Latrouite, 2005). For this reason, management

effectiveness of this valuable resource is crucial to its recovery and

sustainability. At present in the Spanish Mediterranean this fishery

is managed through a 6-month closed season, the prohibition of

retainingberried females, a minimum landing size (90mm CL) and

limited soak duration, in addition to area closures.

0165-7836/$ – seefrontmatter © 2012 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.fishres.2012.05.014

http://localhost/var/www/apps/conversion/tmp/scratch_4/dx.doi.org/10.1016/j.fishres.2012.05.014


http://www.sciencedirect.com/science/journal/01657836

http://www.elsevier.com/locate/fishres

mailto:[email protected]



mailto:[email protected]

http://www.elsevier.com/locate/fishres

http://www.sciencedirect.com/science/journal/01657836




2 L. González-Vicente et al./ Fisheries Research 129–130 (2012) 1–7

Table 1

Number of P. elephas double tagged and subsequently recaptured from 2000 to

2009 with twotags (double) andone tag (single) by sex. Only first time encounters

occurring earlier than thesixth year at liberty used in thestudy areshown.

Released

Female Male

1597 889

Recaptured

Female Male

648 468Double Single Double Single

45.0% 13.1% 31.7% 10.2%

Several previous studies comparing P. elephas inthe MPA and in

fished areas have shown that protection objectives are being met,

benefiting lobster in terms of abundance, mean size and reproduc-

tive potential (Díaz et al., 2011; Goni et al., 2001, 2003a,), as well

as increasing the local fishery yield as result of the outward flux of

above-average-sized lobsters (Goni et al., 2006, 2010).

2.2. Data collection

A total of 2486 lobsters (64.2% females) were captured using

trammel nets inside the MPA during annual experimental fishingsurveys conducted in September 1999 and June 2000, 2001 and

2002 (Table 1) (see Goni et al., 2003bf or details of survey methods).

All lobster caught were double tagged withcoded T-baranchortags

(Hallprint®, Australia) using an ordinary tagging gun. Tags were

inserted dorso-laterally on both sides of the midline in the first

abdominal inter-segmental membrane.

During the following ten years (2000–2009), tagged lobsters

were recaptured either inside the MPAin subsequentexperimental

fishing surveys conducted annuallyin June,or in the fishery outside

the MPA during the annual lobster fishing season (March–August).

Thanks to a close relationship with the fishermen operating in

the region and a reward offered for every recapture notified, the

detection and reporting rate from the fishermen involved in the

tag-recovery program was very high (Goni et al., 2010) and similarto that by scientific staff in the surveys inside the MPA.

2.3. Data analysis

2.3.1. Tag retentionmodel

Recapture data were aggregated by months at liberty, a time

interval (t ) at which fishing and natural mortality can be assumed

constant (Gulland, 1963). The rate of tag loss (L) and the prob-

ability of tag loss (P t ) were estimated applying a least squares

regression on the log-transformed proportion of tags retained

[2·N double/(2·N double +N single)] as a function of time (Eq. (1)).

P t = 1 − e(−Lt ) (1)

N double and N single represent the number of double and singletagged lobsters recaptured; 1- refers to the immediate tag loss

(“Type I loss”) (Bayliff and Mobrand, 1972) and L represents the

instantaneous rate of tag loss (“Type II loss”) (Bayliff and Mobrand,

1972;Chapmanet al., 1965). The number of reencountered animals

(N double +N single) at every time interval t was used as weights in the

regression.

Since the reduction of the number of reencounters over time

led to unrepresentative extreme values (0% or 100%) at the end of

the series, only the first-time encounters occurring earlier than the

sixth year at liberty were used in the study.

The probability of each of three theoretical tag-combinations

was calculated as follows: (1) P D = (1−P t )2, probability of a lob-

ster retaining two-tags; (2)P S = 2 ( 1−P t )·P t , probability of a lobster

retaining one-tag, and, (3)P 0 = (P t )2

, probability of no tagretention.

Subsequently and conditional on retention of at least one tag, the

expected probabilities of retaining either two tags (P RD) or one tag

(P RS) at recapture were computed (Eq. (2)):

P RD =

P D

1− P 0

and P RS =

P S

1 − P 0

(2)

Since this was a double tagging study, to assess the fit of the

tag loss model, the computed probabilities of being single-tagged

(P RS) at recapture (Eq. (2)) were compared with the observed per-

centage of single recaptures over time. Model selection was based

on the parameters of goodness of fit of the weighted least squares

regression and the examination of the residual plots. The level of

significance of all tests was˛= 0.05.

The following modeling assumptions were made (Hearn et al.,

1991): (1)thelossofanytagisindependentofthepositionofthetag

(left orright)andof the presenceof any other tag onthesamespec-

imen; (2)all the tags belongto therandom sample of tags available

for tagging and all of them have similar characteristics, and there-

fore similar chances to be retained; (3) the natural mortality and

catchability of an individual is independent on its tag-status; (4)

the tags are conspicuous and the same reporting reward is offered

for a single or double tagged animal, so they have equal reporting

rate; and (5) dispersal of individuals within the population is ran-

dom and independent of their tag-status. Therefore the proportionof double vs. single tagged animals reencountered is representa-

tive and independent of recapture location. Assumptions 1 and 2

relate to the independence and homogeneity of tags applied, while

assumptions 3–5 pertain to the representativeness of the tagged

population. Assumptions based on tag-status independence could

be violated when features associated with greater likelihood of tag

loss (e.g. molt state, size at tagging) are present, since lobsters lose

tags mainly through molting.

2.3.2. Factors affecting tag loss

The effect of factors time-at-liberty, size-at-tagging (log-

transformed), tagging season and sex on the probability of tag loss

was explored by multipleregression analysis.Tag losswas recorded

as a binomial response variable (retained or lost) with a logit linkfunction. The model which showed the lowest value of the Akaike

information criterion (AIC) was selected using a stepwise selection

procedure (Zuur et al., 2007).

Tag loss as a function of time-at-liberty was evaluated using

data for all mature lobsters (CL at tagging≥80mm CL) tagged and

released in June (surveys 2000–2002). Data for assessing the effect

of size at tagging on the probability of tag loss included all lobsters

tagged and released in June (surveys 2000–2002) grouped in two

size classes: small (immaturewith CL at tagging < 80mm) and large

(adults with CL at tagging≥80 mm). For assessing the effect of tag-

ging season on tag loss the dataset included recaptures grouped

according to month of tagging, June and September, the latter a

proxy for proximity to the males autumn molt.

2.3.3. Effects of ignoring tag loss on the return rate

To assess the effectof the tag loss on the return rate (the ratio of

individuals expected to be reencountered over time, relative to the

initial number of tagged animals released), we created a stochastic

model which allowed forecasting the reencounter reduction with

time. The following parameter values were used as input data: (i)

number of releasesR♀ =R♂ = 200 individuals per cohort; (ii) appar-

ent survivalPhi♂ = 0.6, Phi♀ = 0.8 (assuming lower values for males

based on their higher probability of mortality and emigration, Goni

et al., 2010), and (iii)recapture probability identical for bothgroups

p♀ = p♂ =0.5. We simulated 100 realizations from the log-normal

distribution of tag loss probabilities previously estimated (P t ±SD)

andestimated the mean proportion of tags lost over time. Then, the

expected return rate with time in the no-tag loss scenario (Phin · p)



L. González-Vicente et al. / Fisheries Research 129–130 (2012) 1–7 3

Fig. 1. Proportion of single-tag recaptures for lobster tagged in September (autumn, gray bubbles) and June (summer, white bubbles). Recaptures accumulated in 30days-

at-liberty bins. Size of the bubble is proportional to thenumber of recaptures within each bin. Forsimplicity only single-tagged data (%) are represented. Females (top) and

males (bottom).

and the expected return rate affected by tag loss (Phin · p− P t ) were

computed, where Phin is the cumulative survival up to time (n).

Thus, theexpectedfrequencies of returnedanimalsover timeunder

both scenarios were reported.

3. Results

A total of 1116 first-time recaptures were obtained from 2000

to 2009, which represented 40.6% of the tagged females and 52.6%of the tagged males (Table 1). The majority of the recaptures took

place during the first three years-at-liberty, followed by a sharp

reductionof the numberof recoveries with time (Fig.1). From these

recaptures, 502(45.0%)femalesand 354(31.7%)males retainedtwo

tags.

Data exploration suggested model assumptions were generally

met, not finding statistical differences in the number of ani-

mals reencountered depending on tag-status, neither for males or

females (2-test testing the position of the tag p-value> 0.05, n.s.;

t -test testing for dispersal of single and double tagged animals in

the MPA and in the fished area p-value> 0.05, n.s.; QAICc values

indicated models accounting for recapture probability dependent

on tag-status performed worse and showed a lower statistical sup-

port).

3.1. Factors affecting tag loss

Multiple regression analyses indicated that the model includ-

ing factors time-at-liberty, size-at-tagging, tagging season, sex and

the interaction time:size-at-tagging explained the most deviance

and that these five attributes were significant ( p< 0.05) (Table 2).

Therefore, data were split and analyzed separately.

Modeling results for June tag releases indicated that the model

accounting for immediatetag loss 1- was not well supported, sug-

gesting that tags were adequately applied and retained in the shortterm for individuals tagged during the intermolt. However there

was a statistically significant long-term tag shedding L, reaching

5.0%year−1 in females and 6.8%year−1 in males (Table 3a). When

this constant rate was back-transformed into a probability P t (Eq.

(1)), the cumulative probability of losing a tag increased monoton-

ically. As a result, a population of P. elephas tagged with a single

tag would be expected to lose 50% of the tags after 10years in

males and after 14years in females (Fig. 2). No overall pattern was

found in the residuals when comparing the expected probability

of single-tagged recaptures (P RS) and the observed percentage of

single recaptures (% S Obs) (Fig. 3).

Size-at-tagging and its interaction with time were also signifi-

cant factors affecting tag loss (Table 2) and a higher proportion of

tags were shed in the group which was tagged when undersized.




Table 2

Deviance tableof multiple regressionof predictor variables for tag loss recorded as binomial response (logit link): time-at-liberty, size-at-tagging (log-transformed), tagging

season and sex. Statistically significant p-values in bold.

Estimate SE Deviance Residual deviance F p-Value

Intercept 21.251 19.719 1211.10

Time-at-liberty −11.865 5.463 177.86 1033.24 177.862 0.000

Size-at-tagging −5.708 4.262 21.2 1012.04 21.198 0.000

Factor (tagging season) −0.446 0.190 5.77 1006.27 5.767 0.016

Factor (sex) −11.611 27.534 5.82 1000.46 5.816 0.016

Time:size 2.832 1.181 9.38 991.08 9.383 0.002Time:factor (sex) 6.568 7.856 0.99 990.08 0.992 0.319

Size:factor (sex) 2.427 5.922 3.23 986.85 3.233 0.072

Time:size:factor (Sex) −1.366 1.690 0.64 986.21 0.639 0.424

Tagging season had a significant effect on the probability of tag

shedding (Table 2) in line with the observed pattern (Fig. 1). Long

term tag loss L was sensibly higher in male lobsters tagged in

September than in June (Table 3), although immediate tag loss 1-

again was not statistically different from zero.

3.2. Effects of ignoring tag loss on the return rate

Simulations mimicking the returns under the no-tag loss and

tag-loss scenarios showed a noticeable reduction of the expectedreencounters over time in the latter, more marked in males than in

females (Fig. 4). That is, after the first year at liberty, the number of

reencounters would be reduced by 22.1% and 12.2% for males and

females, respectively.

4. Discussion

The use of CMR methods to study animal populations relies

on the assumption that all tagged animals will be accurately rec-

ognized (Lebreton et al., 1992; Seber, 1986). Since tag shedding

constitutes a removal from the tagged pool that is indistinguish-

able from mortality or emigration, its knowledge is critical. In this

double-tagging experiment of the European spiny lobster P. ele-

phas in the wild, T-bar anchor tag shedding rates were amongthe lowest estimated for lobster species marked with external

tags. In spite of this relatively low rate, in an ordinary CMR study

where a single tag is used, half of the tagged population would

not be recognized after 10–14years (male and females, respec-

tively).

Published lobster tag-loss estimates vary depending not only

on the species, but also on tag type, size-at-tagging, tag-

ging season, duration of experiment, insertion location and

experimental conditions (wild, captivity) (Melville-Smith and

Chubb, 1997; Montgomery and Brett, 1996; Taylor and Hoenig,

1991; Trendall, 1989; Scarrat, 1970), hampering comparisons and

probably explaining the high variability of results (Table 4). Tag

loss rates estimated in this study (6.8–5.0% year−1 for males and

females, respectively) were lower than expected on the basis of

studies using comparable tag methods on other lobster species,

which can be as high as 87% year−1 in Homarus, 56% year−1 in Jasus

or 54%year−1 in Panulirus (Table 4). Nevertheless, the limited infor-

mation available for comparison suggests that T-bar anchor tagsinserted dorsally provide the highest retention in spiny lobsters, as

is the case of this study and that of Jasus verreauxi in tank experi-

ments of 8%year−1 (Table 4). Field experiments on Jasus edwardsii

with T-bar anchor tags inserted ventrally, which allegedly offers a

better protection against friction and entanglement, yielded a sim-

ilar shedding rate for females (6.1% year−1) but a higher rate for

males (13.4% year−1) (Frusher et al., 2008).

Due to theirhighergrowthrates andconsequent moltfrequency

(Díaz, 2010), undersized specimens of P. elephas were more likely

to lose tags than adults, being this probability of tag loss reduced

as growth rate slows down with time, as inferred from the mul-

tiple regression analyses. Similarly, males which grow faster than

females showed poorer tag retentionthan females. This is expected

because the main cause of crustacean tag shedding is molting(Davis, 1978; Restrepo and Hoening, 1988), although other causes

have been identified, such as removal by congeners (Krouse and

Nutting, 1990) or predators(Rowe and Haedrich, 2001), and entan-

glement with surfaces (Ennis, 1986). For the same reason, tags

inserted duringpre-molt, such as those of males tagged in autumn,

showed poorer tag retention than tags inserted in summer during

post-molt (adult females) or inter-molt (males). Similarly, in other

lobster species, tagging duringthe late pre-molt hasbeen observed

Table 3

Estimatesof short-term (1-) andlong-term (L) tag loss(± SE)by sexin lobsters≥80mm CL attagging;(a) tagged inJuneand(b) inSeptember.The modelswithlowest AIC

score are highlighted in bold.

Sex Model Tag loss (year−1 ± SE) p-Value AIC

(a)

Female p= 1-e−Lt 1- 0.021 ± 0.04 0.966

L 0.040 ± 0.01 0.000 −65.48

p= 1-e−Lt L 0.050 ± 0.01 0.000 −67.48

Male p= 1-e−Lt 1- 0.021 ± 0.04 0.637

L 0.072 ± 0.01 0.000 −55.94

p= 1-e−Lt L 0.068 ± 0.01 0.000 −57.71

Female p= 1-e−Lt 1- 0.021 ± 0.036 0.567

L 0.040 ± 0.011 0.001 −19.67

p= 1-e−Lt L 0.045 ± 0.007 0.000 −21.32

Male p= 1-e−Lt 1- 0.021 ± 0.036 0.540

L 0.081 ± 0.019 0.000 5.11

p= 1-e−Lt L 0.090 ± 0.013 0.000 3.52




Table 4

Publishedestimates of lobster tag loss rates. Tag loss estimatesfrom short term studies (<1year) were scaled up to oneyear at liberty. (*)Juveniles.

Species Reference Tag type Tag loss (year−1) Experiment

Palinurus elephas This study, 2012 T-bar anchor tag 6.8%♂ Field

5.0% ♀

Panulirus cygnus Chittleborough (1974) Sphyrion tag 27% Field

Rock lobster tag 30%

Panulirus argus (*) Davis (1978) Floy FD-68B tag 45% Field

Panulirus marginatus O‘Malley (2008) Streamer tag 54% Field

Jasus novaehollandiae Winstanley (1976) Dart tag 56%♂ Field41% ♀

Jasus verreauxi Montgomery and Brett (1996) Toggle tag 6% Laboratory

T-bar anchor tag 8%

Dart tag 8%

Jasus edwardsii Frusher et al. (2008) T-bar anchor tag 13.4%♂ Field

6.1% ♀

Homarus americanus Cooper (1970) Sphyrion tag 22% Laboratory

Scarrat (1970) 32% Laboratory

Fogarty et al. (1980) 87% Field

Ennis (1986) 36% Molted Field

24% Non molted

Homarus gammarus Bennett and Lovewell (1983) Sphyrion tag 36% Laboratory

to increase the risk of tag loss (Comeau and Mallet, 2003; Moriyasu

et al., 1995).

Despite the comparably low rate of tag shedding obtained forP. elephas in this study, this loss has a significant impact on the

return rate. To that proportion of the population which is missed

with time as result of shedding itself, we have to add the effect of

the decreasing return rate with time, given that the detection of

those tagged animals is conditional on their survival up to time t

and recaptureat that particular event.This wouldbe more acute for

those strata with higher growth rates i.e. differential growth rate

Fig.2. Expected cumulativeprobabilityand 95% confidenceintervalsof tag loss(P t )

for lobsters taggedin Junemeasuring≥80mm CLat tagging.Females() and males

().

related to sex, size or molt stage.Our simulations forthe returnrate

based on constant parameters and number of single-tag releases,

forecast the failure to detect recaptured animals beyond the fourthyear at liberty as result of the impact of tag loss. Because our field

data come from a double tagging experiment, we obtained reen-

countersfar beyond this time limit (>10 years). Certainlythe steady

state simulated here is used just for illustrating the importance of

accounting for tag loss, since in fact we suspect that several fac-

tors, such as sex-related differences in movement probability or

catchability, are jointly influencing the reencounter pattern of P.

elephas in the study area in the long term (González-Vicente et al.,

2011). Therefore, P. elephas true survival may be generally higher

0%

20%

40%

60%

80%

100%

876543210Years at liberty

% S Obs PRS

0%

20%

40%

60%

80%

100%

876543210

Years at liberty

% S Obs PRS

Fig. 3. Expected frequency of single-tag lobsters at recapture (P SR ) versus observed

proportion (June releases≥80mm CL). Females () and males ().




0

20

40

60

80

100

876543210

A b

s o l u t e f r e q u e n c y

Years at liberty

Expected returns

with tag lossExpected returns

0

20

40

60

80

100

876543210

A b

s o l u t e f r e q u e n c y

Years at liberty

Expected returns

with tag lossExpected returns

Fig. 4. Expected number of returned individualsin scenariosunaffected (solid line)

and affected (dashed line) by tag loss. Females (top) and males (bottom).

than the apparent survival simulated here, and so are the return

rates, leading to this long term reencounter time series.

Although simulation studies have previously shown that the

‘type I tagloss’ does not lead to importantbiason survival estimates

(McGarvey et al., 2009), when studying the parameters involved

in the population dynamics, unaccounted ‘type II tag loss’ wouldlead to erroneous estimates of recapture and survival. Moreover,

recaptured but unidentified animals would be misinterpreted as

immigrants, biasing upwards recruitment estimates. As a result of

this, inadequate management measures to protect the population

could be enforced, while the originating cause of animal loss (tag

detachment) would remain unidentified.

Tag loss has been shown to bias abundance and survival esti-

mates in field studies (McDonald et al., 2003; Seber and Felton,

1981), a problem which is commonly remedied by adjusting sur-

vival estimates a posteriori (Pollock, 1981). Also, when population

parameters are affected by complex time-, location- or state-

dependency factors, the correction for tag loss can be incorporated

into the likelihood to obtain specific estimates for each category

and time event. For instance, Oosthuizen et al. (2010) determinedcohort- and site-specific tag loss estimates for southern elephant

seal (Mirounga leonina), in order to avoid the heterogeneity among

the different age and location strata. Similarly, it is possible to

include sampling methodology aspects into the likelihood estima-

tion. For instance, Juillet et al. (2011) combined information from

liverecaptures, deadrecoveries, and single anddouble tagged spec-

imens of greater snow goose (Chen caerulescens) to estimate tag

loss.

In conclusion, the service life of T-bar tags for this species

spans beyond 15years, validating this tagging method and provid-

ing a reasonably high retention in the long term. In view of their

long-termretention, endurance against ecdysis and easy detection,

tagging dorsally with T-bar tags anchor is a suitable method for

field studies of lobsters such as P. elephas andpresumably of others

with similar molt rates. For maximizing survival and tag retention,

tagging is advised during the inter-molt period.

Acknowledgements

The authors are grateful to the skippers and crews of the fishing

vessels operatingin thestudy areafor their invaluable collaboration

in recapture notification over the years. We also thank A. Quet-

glas and O. Renones and all the collaborators who contributed tothe data collection in the early years of the project. We would like

to thank to Dr. McGarvey and one anonymous referee for his/her

constructive comments. This work was funded by the Spanish Sec-

retaría General del Mar by means of a grant for the LANGOSTA &

ERICOL projects (1997–2012). L.G.V. was supported by a FPI-IEO

scholarship (2008–2012).

References

Afonso, P., Fontes, J., Holland, K.N., Santos, R.S., 2008. Social status determinesbehaviour andhabitat usage in a temperate parrotfish: implicationsfor marinereserve design. Mar. Ecol. Prog. Ser. 359,215–227.

Bayliff, W.H., Mobrand, L.M., 1972. Estimates of the rates of shedding of dart tags

from yellowfin tuna. Inter-Am. Trop. Tuna Commun. Bull. 15, 441–462.Bennett, D.B., Lovewell, S.R.J., 1983. Lobster (Homarus gammarus (L.)) tagging trialsin England. In: Ministry of Agriculture,Fisheries and Food, Lowestoft (England).Directorate of Fisheries Research Tech. rep. No. 71, p. 7.

Beverton, R.J.H., Holt, S.J., 1957. On the dynamics of exploited fish populations. In:Fisheries Investigations Series II (19). Chapman & Hall, London.

Chapman, D.G., Fink, B.D., Bennett, E.B., 1965. A method for estimating the rate of shedding of tags from yellowfin tuna. Inter-Am. Trop. Tuna Commun. Bull. 10,335–342.

Chittleborough, R.G., 1974. Development of a tag for the western rock lobster. In:CSIRO Division of Fisheries and Oceanography, Rep. No. 56, p. 19.

Comeau, M.,Mallet, M., 2003. Theeffect of timingof tagging on streamer-tagrecap-ture rates for American lobster (Homarus americanus). Fish. Bull. 101, 476–483.

Cooper, R.A., 1970. Retention of marks and their effects on growth behavior, andmigrations of the american lobster, Homarus americanus. Trans. Am. Fish. Soc.99, 409–417.

Davis, G.E., 1978. Field evaluation of a tag forjuvenile spiny lobster Panulirus argus.Trans. Am. Fish. Soc. 107, 100–103.

Díaz, D., Mallol, S., Parma, A.M., Goni, R., 2011. Decadal trend in lobster reproduc-

tive output from a temperate marine protected area. Mar. Ecol. Prog. Ser. 433,149–157.

Díaz, D., 2010. Ecología y dinámica de la fase juvenil de la langosta roja Palinuruselephas en el Mediterráneo noroccidental. PhD thesis. Universitat de Barcelona.321.

Ennis,G.P.,1986. Sphyrion tag loss from theAmericanlobster Homarus americanus.Trans. Am. Fish. Soc. 115, 914–917.

Fogarty, M.J., Borden, D.V.D., Russell, H.J., 1980. Movements of Tagged AmericanLobster, Homarus americanus, offRhode Island. Fish. Bull. 78 (3), 771–780.

Frusher, S., Burch, P., Mackinnon, C., Wotherspoon, S., 2008. Developing taggingmodels and validating key assumptions for estimating crucial fishery assess-ment parameters in the Tasmanian rock lobster fishery. FRDC project FRDCproject no.2003/051, p. 134.

González-Vicente, L., Goni,R., Frusher,S.D., Díaz, D.,Mallol, S.,2011.Willwe seeyouagain? Populationmonitoring of Palinurus elephas insidea marinereserve fromcapture-mark-recapture data. In: 9th International Conference and Workshopon Lobster Biology andManagement, June 19–24, Bergen (Norway),p. 73.

Goni, R., Hilborn, R., Díaz, D., Mallol, S., Adlerstein, S., 2010. Net contribution of spillover from a marine reserve to fishery catches. Mar. Ecol. Prog. Ser. 400,

233–243.Goni, R., Latrouite, D., 2005. Review of the biology, ecology and fisheries of Palin-

urus spp. species of European waters: Palinurus elephas (Fabricius, 1787) andPalinurus mauritanicus (Gruvel, 1911). Cah. Biol. Mar. 46, 127–142.

Goni, R., Quetglas,A., Renones,O., 2003a. Size at maturity, fecundityand reproduc-tive potential of a protected population of the spiny lobster Palinurus elephas(Fabricius, 1787) from theWestern Mediterranean. Mar. Biol. 143, 583–592.

Goni, R.,Quetglas,A., Renones,O., 2003b.Differentialcatchabilityof maleand femaleEuropean spinylobsterPalinurus elephas (Fabricius1787) in trapsand trammel-nets. Fish. Res. 65, 295–307.

Goni, R.,Quetglas,A., Renones,O., 2006. Spilloverof spiny lobsters Palinurus elephasfrom a marinereserveto anadjoiningfishery.Mar.Ecol. Prog.Ser.308,207–219.

Goni, R., Renones,O., Quetglas,A., 2001. Dynamics of a protectedWesternMediter-ranean population of the European spiny lobster Palinurus elephas (Fabricius1787) assessed by trap surveys. Mar. Freshwater Res. 52, 1577–1587.

Gulland, J.A.,1963. On theanalysisof double-taggingexperiments.In: NorthAtlanticFish Marking Symposium, I.C.N.A.F, pp. 228–229.

Hearn, W.S., Leigh, G.M., Beverton, R.J.H., 1991. An examination of a tag shedding

assumptionwithapplication toSouthernBluefin Tuna.ICESJ. Mar.Sci.48, 41–51.




Juillet, C., Choquet,R., Gauthier,G., Pradel, R., 2011.A Capture-Recapture model withdouble-marking, live and dead encounters, and heterogeneity of reporting dueto auxiliary mark loss. J. Agric. Biol. Environ. Stat. 16, 88–104.

Krouse, J.S., Nutting, G.E., 1990. Effectiveness of the Australian western rock lobstertagfor marking juvenile Americanlobstersalong theMaine coast.Am. Fish. Soc.Symp. 7, 94–100.

Lebreton,J.D., Burnham,K.P., Clobert, J.,Anderson, D.R.,1992. Modeling survival andtesting biological hypotheses using marked animals: a unified approach withcase studies. Ecol. Monogr. 62, 67–118.

McDonald, T.L., Amstrup, S.C., Manly, B.F.J ., 2003. Tag loss can bias Jolly-Sebercapture-recapture estimates. Wildlife Soc. Bull. 31, 814–822.

McGarvey, R., Matthews, J.M., Feenstra, J.F., 2009. Estimating mortality from times-at-large: testing accuracy and precision using simulated single tag-recoverydata. ICES J. Mar. Sci. 66, 573–581.

Melville-Smith,R., Chubb,C.F., 1997.Comparison of dorsal and ventral tag retentionin western rock lobsters Panulirus cygnus (George). Mar. Freshwater Res. 48,577–580.

Montgomery, S.S., Brett, P.A., 1996. Tagging eastern rock lobsters Jasus verreauxi:effectiveness of several types of tag. Fish. Res. 27, 141–154.

Moriyasu, M., Landsburg, W., Conan, G.Y., 1995. Sphyrion tag shedding and taginduced mortality of the American Lobster Homarus americanus H. MilneEdwards 1837 (Decapoda, Nephropidae). Crustaceana 68, 184–192.

O‘Malley,J.M., 2008.Evaluations oftag retentionand a devicefor releasingdiscardedHawaiian spiny lobsters Panulirus marginatus. N. Am. J. Fish. Manage. 28 (3),619–624.

Oosthuizen, W.C., de Bruyn,P.J.N., Bester, M.N., Girondot,M., 2010. Cohortand tag-site-specific tag-loss rates in mark-recapture studies. A southern elephant sealcautionary case. Mar. Mammal. Sci. 26, 350–369.

Parker, R.R.,1962.Tag failure associated with a netfisheryas a source ofexperimen-tal error.J. Fish. Res. Board Can. 19, 315–320.

Pollock,K.H.,1981. Capture-recapture models:a review ofcurrentmethodsassump-tions,and experimental design. Stud. Avian Biol. 6, 426–435.

Restrepo,V.R.,Hoening,J.M.,1988. Munro’smethodfor estimatingintermolt periodsof tropical decapods is robust. Bull. Mar. Sci. 42 (3), 488–492.

Rowe, S.,Haedrich, R.L., 2001. Streamer tagloss from American lobsters.Trans. Am.Fish. Soc. 130, 515–518.

Scarrat, D.J., 1970. Laboratory and field tests of modified sphyrion tags on Lobsters(Homarus americanus). J. Fish. Res. Board Can. 27, 257–264.

Seber, G.A.F., 1986. A review of estimating animal abundance. Biometrics 42,

267–292.Seber, G.A.F., Felton, R., 1981.Tag lossand thePetersenmark-recapture experiment.

Biometrika 68, 211–219.Taylor, D., Hoenig, J.M., 1991. Effect of tag anchor location on retention/survival

throughmolt in male snow crabs (Chionoecetes opilio). Fish. Bull. 89, 325–329.Trendall, J.T., 1989. Influence of tag type on recapturerates in thetropical rock lob-

ster,Panulirusornatus. In:Hancock,D.A. (Ed.), AustralianSocietyfor FishBiologyTagging Workshop. Tagging: Solution or Problem? (Sydney, NSW: July 21–22,1988) Bureau of Rural ResourcesProceedings. Canberra: A.G.P.S (5), pp. 68–76.

Williams, B.K.,Nichols,J.D., Conroy, M.J.,2002. Analysis and Management of AnimalPopulations. Modeling, Estimation, and Decision Making. Academic Press, SanDiego.

Winstanley, R.H.,1976.Markingand tagging ofthe southernrocklobster Jasusnovae-hollandiae Holthuis off Tasmania. N. Z. J. Mar. Freshwater Res. 10 (2), 355–362.

Zuur, A.F., Ieno, E.N., Smith, G.M., 2007. Analysing Ecological Data. Springer, NewYork.

Documents

Jurnal Populasi Print