An Evaluation of the Management Effectiveness of a Stock Re-Building Strategy for a Multi-Species Demersal Trap Fishery in Abu Dhabi, United Arab Emirates

doi:10.4236/ojms.2011.13009

Paper Menu >>

Journal Menu >>

Open Journal of Marine Science, 2011, 3, 82-97

doi:10.4236/ojms.2011.13009 Published Online October 2011 (http://www.SciRP.org/journal/ojms)

An Evaluation of the Management Effectiveness of a Stock

Re-Building Strategy for a Multi-Species Demersal Trap

Fishery in Abu Dhabi, United Arab Emirates

Edwin Mark Grandcourt, Thabit Zahran Al Abdessalaam, Stanley Alexander Hartmann, Franklin

Francis, Ahmed Tarish Al Shamsi

Biodiversity Management Sector, Environment Agen cy–Abu Dhabi, Abu Dhabi, United Arab Emirates

E-mail: [email protected]

Received June 27, 2011; revised September 8, 2011; accepted September 22, 2011

Abstract

The off-shore demersal trap fishery in the southern Arabian Gulf has been managed by gear regulations and

effort constraints which were aimed at rebuilding depleted stocks. In order to evaluate the success of these

regulations, a variety of selectivity and other fishery metrics were compared for the key species (Diagramma

pictum, Epinephelus coioides and Lethrinus nebulosus) before and after their introduction. With the excep-

tion of a minor increase in the mean age at first capture from 1.3 yrs to 1.9 yrs for E. coioides, there were no

significant changes in the values or trends in juvenile retention, mean size or the mean sizes and ages at first

capture. The comparison of selectivity ogives with data derived independently through an experimental fish-

ing program indicated that the failure of juvenile escape panels to modify the selectivity characteristics of the

fishery could not be attributed to a lack of compliance. Furthermore, there were no significant changes in

fishing mortality rates, harvest rates, catch, effort, yield per recruit and relative spawner biomass per recruit

following the introduction of the management regulations. Age structures were highly truncated and the

management measures had failed to increase the relative proportion of older age classes. Stock status indica-

tors suggested that all species were heavily over-exploited with evidence of both growth and recruitment

over-fishing. Given the failure of existing regulations to modify gear selectivity, reduce effort and rebuild

stocks, the results of the study suggest that management authorities should consider alternative measures in-

cluding a moratorium on the use of traps in the off-shore demersal fishery of Abu Dhabi.

Keywords: Demography, Stock Assessment, Per Recruit, Population Dynamics, Tropical Fisheries

Management

1. Introduction

The fisheries of the southern Arabian Gulf provide a

source of income, employment and recreation at the same

time as contributing to the cultural heritage and food secu-

rity of the inhabitants of the coastal states. Fishing in the

waters off the Emirate of Abu Dhabi in the United Arab

Emirates (UAE) is primarily conducted from traditional

wooden dhows and open fibreglass dories. Dome shaped

wire traps are the most commonly used fishing gear al-

though a variety of other methods exist including; fixed

inter-tidal fence nets, gillnets, hand lines and trolling

lines. Landings are typically diverse and characteristic of

a multi-species tropical fishery, with over 100 species

from 35 families being commonly caught. Target species

in the demersal fishery are primarily composed of repre-

sentatives of the families; Carangidae, Lethrinidae, Hae-

mulidae, Serranidae, Siganidae and Sparidae [1].

An extensive fisheries resource survey conducted in

UAE waters during 2002 indicated that the stocks of key

commercially exploited species had declined to 19% of

levels that were present in 1978. The reduction in abun-

dance was even more pronounced (13%) for more vul-

nerable species such as the orange spotted grouper (Epi-

nephelus coioides) [2]. Subsequent fisheries dependent

assessments corroborated these results and suggested that

71% of the resource base was over-exploited [1]. For

many of the key species, such as Lethrinus nebulosus

E. M. GRANDCOURT ET AL.83

and Diagramma pictum, the relative stock sizes were

well below limit biological reference points suggesting

that both growth and recruitment over-fishing had oc-

curred [3].

The principal critical management issue faced by the

off-shore demersal fishery in Abu Dhabi is that the ma-

jority of species are being exploited well above sustain-

able levels. The excessive fishing effort has even im-

pacted the lower trophic groups, such as the Siganidae,

which are typically more resilient [4]. Furthermore, the

selectivity characteristics of the main gear type (fish

traps) resulted in a large proportion of catches being

composed of immature fish that had not achieved their

full growth potential. More than half of the landed ca-

ches of the sparids (Acanthopagrus bifasciatus and Ar-

gyrops spinifer) for example were composed of imma-

ture fish [5].

Prior to the year 2000, the demersal fishery of Abu

Dhabi was largely open access and there were no con-

straints on fishing effort. Information on the depleted

status of the resource base derived from stock assess-

ments prompted authorities to implement a variety of

management regulations. Effort reductions were imple-

mented at the start of 2004 by limiting the number of fish

traps to 125 for each dhow and banning the use of traps

by open dories. Traditional dhows previously used an

average of 220 fish traps. The simultaneous introduction

of gear regulations included the requirement for a

stainless steel juvenile escape panel to be fitted to all

traps. Given the potential for traps to ‘ghost fish’ after

being lost [6], a magnesium/zinc alloy sacrificial anode

pin, which allows the panel to open after 2 weeks was also

mandatory.

In 2001, an on-going catch and effort data recording

system and stock monitoring programme for 3 key spe-

cies in the demersal trap fishery (D. pictum, E. coioides

and L. nebulosus) was implemented [1]. This provided

the opportunity to establish demographic and fishery me-

trics that relate to the resource status and selectivity

characteristics of the demersal trap fishery. In this con-

text, the objective of this study was to use a selection of

time series metrics to provide an objective evaluation of

the effectiveness of effort and gear regulations that were

implemented in 2004 as part of a stock re-building man-

agement strategy for the off-shore demersal fishery of

Abu Dhabi.

2. Materials and Methods

2.1. Study Site and Sampling Protocol

Size frequency data was collected from commercial

catches of D. Pictum, E. coioides and L. nebulosus made

off the coast of the Emirate of Abu Dhabi in the United

Arab Emirates (Figure 1) between January 2001 and

December 2008. Fish were selected at random from

Figure 1. Study site showing the location of the Emirate of Abu Dhabi in the southern Arabian Gulf.

E. M. GRANDCOURT ET AL.

landings, lengths were taken using a measuring board

and recorded to the nearest cm fork length (LF) for D.

Pictum and L. nebulosus and total length (LT) for E.

coioides. Biological data was collected from specimens

purchased from commercial catches over the same period.

Standard length (Ls), fork length (LF) and total length (LT)

were obtained using a measuring board and recorded to

the nearest mm. Whole wet weight was measured using

an electronic balance and recorded to the nearest g. Fish

were sexed by macroscopic examination of the gonad

which was dissected out and subsequently weighed to

0.1 g using an electronic balance. Sagittal otoliths were

extracted, cleaned in water, dried, weighed to 0.1 mg and

stored in seed envelopes. One of each pair of sagittae

was weighed to 0.1 mg and embedded in epoxy resin.

Transverse sections through the nucleus of approxi-

mately 200 to 300 m thickness were obtained using a

twin blade saw. Sections were mounted on glass slides

and examined using a low power microscope and re-

flected light.

Catch and effort data were collected through a strati-

fied catch and effort data recording system which was

implemented in 2001. Landings of the key species were

recorded to family level between 2001 and 2004 and

only to species level from 2005 onwards. Consequently,

data used in the analyses of catches were aggregated to

family level (Haemulidae, Lethrinidae and Serranidae)

for the entire time series.

2.2. Age Estimation and Length Weight

Parameters

Age was estimated from the number of opaque bands in

transverse sections of otoliths as these had been previ-

ously validated as annuli in D. Pictum and L. nebulosus

[3] and E. coioides [7]. Three independent age readings

were made and data only included in the analyses if two

or more readings were in agreement. The overall preci-

sion was established using the index of average percent

error (APE) [8].

Parameters of the length weight relationship were ob-

tained by fitting the power function to length and weight

data:

WaL

Where W is the wet weight, a is a constant, L

is the fork

length (replaced with L

for E. Coioides) and b is close

to 3.0 for species with isometric growth.

2.3. Mortality and Selectivity

Size-at-age data were used to construct age length keys

following the method described in [9], and these were

used to convert length frequency data into age frequency

distributions. The annual instantaneous rate of total mor-

tality (Z) for each species by year was subsequently de-

termined using the age based catch curve method [10].

The natural logarithm of the number of fish in each age

class was plotted against the corresponding age and Z was

estimated from the descending slope of the best fit line

using least-squares linear regression. Initial ascending

points representing fish that were not fully recruited to the

fishery were excluded from the analyses.

Backwards extrapolation of age based catch curves

was used to estimate probability of capture data based on

the method described in [11]. Selectivity curves were

generated using the logistic function fitted to plots of the

probability of capture against age and used to derive

values of the mean age at first capture for each species

by year:





1/(1exp() )Prttc

Where: P is the probability of capture, tc is the mean age

at first capture and r is a constant which increases in

value with the steepness of the selection curve.

Length frequency samples for each year were con-

verted into relative age frequency distributions using

parameters of the von Bertalanffy growth function [7,3]

following the method of [12]. The natural logarithm of

the number of fish in each relative age group divided by

the change in relative age was plotted against the relative

age. Backwards extrapolation of the descending limb of

these length converted catch curves was used to estimate

probability of capture data using the method of [11]. Se-

lectivity curves were generated by fitting the logistic

function to probability of capture and size data which

were used to derive values of the mean size at first cap-

ture (Lc50) for each species. The lengths giving the high-

est yields (Lopt) were estimated for the study species us-

ing the empirical relationship of [13] where Lopt =

L(3/(3+M/k)).

Selectivity curves were also generated for pooled size

frequency data collected before (2001-2003) and after

(2004-2008) the introduction of juvenile escape panel

regulations. These were compared to selectivity curves

for the study species that were derived from an experi-

mental fishing program in order to provide an independ-

ent assessment of selectivity and gauge compliance. The

mean size was also calculated for each species by year.

The annual instantaneous rate of fishing mortality (F)

was calculated for each species and year by subtracting

the natural mortality rate (M), estimated as 0.13 and 0.20

for D. Pictum and L. nebulosus respectively [3] and 0.19

for E. coioides [7], from the total mortality rate (Z) de-

rived from age based catch curves. Juvenile retention (J)

was calculated as the proportion of fish in landings that

were below the mean size at first sexual maturity of fe-

E. M. GRANDCOURT ET AL.85

males given as 31.8 cm LF and 27.6 cm LF for D. Pictum

and L. nebulosus respectively [3] and 42.6 cm LT for E.

coioides [14].

The annual harvest rate (H), or percentage removal by

the fishery was estimated as:



H -e00





2.4. Per-recruit Analyses

A yield per recruit (YPR) model [10] was used to esti-

mate YPR and the relative spawner biomass per recruit

(SBR) every year for each species. Published estimates

of natural mortality rates (M) and growth parameters (k

and L) of the von Bertalanffy growth function for the

same stocks [3,7], were used as inputs to the model along

with the selectivity parameters (Lc50) estimated here. The

target biological reference point was defined as the fish-

ing mortality rate at which the slope of the YPR curve

was 1/10th of the value at its origin (F0.1). The limit bio-

logical reference point was defined as the fishing mortal-

ity rate at which YPR would be maximized (Fmax). Addi-

tional target (Fopt) and limit (Flimit) biological reference

points of 0.5M and 2/3M respectively were calculated for

each species for comparison following [15].

2.5. Analyses of Metrics and Age Structures

Selectivity (J, mean size, Lc50, tc) and other fishery met-

rics (F, YPR, SBR, H, catch, effort) were compared be-

fore (2001-2003) and after (2004-2008) the introduction

of juvenile escape panels and effort restrictions using

Mann-Whitney U tests [16]. Partial correlations [16] were

also used to evaluate the impact of the introduction of

juvenile escape panels and effort restrictions on time se-

ries trends of these metrics.

The proportion of fish in each age class was calculated

from pooled data from the same before (2001-2003) and

after (2004-2008) periods. Age structures were examined

using 2 goodness of fit tests. Independent tests were

conducted to determine whether there was a significant

difference from unity for individual age classes follow-

ing the introduction of management measures in 2004.

The probability level was set at .05 ( 0.5) and Yates's

correction factor was used on account of there being only

1 degree of freedom for each comparison.

3. Results

3.1. Age Estimation and Length Weight

Parameters

Transverse sections of sagittal otoliths were character-

ised by well defined alternating translucent and opaque

bands when viewed with reflected light under low power

magnification (Figure 2).

A total of 1,082 otoliths were used to age D. Pictum of

which 83 samples were rejected due to a lack of agree-

ment between readings. Fish ranged in size from 18.9 to

63.0 cm LF and age from 1 to 13 yrs. A high level of re-

peatability was achieved in age readings with an APE

value of 5.7%. A total of 17,273 size measurements were

recorded and the length weight relationship (y = 0.014

x2.99) provided a good fit to length and weight data (r2 =

0.99).

A total of 1,706 otoliths were used to age E. coioides

of which 26 samples were rejected due to a lack of

agreement between readings. Fish ranged in size from

20.6 to 100.2 cm LT and age from 1 to 12 yrs. A high

level of repeatability was achieved in age readings with

an APE value of 3.5%. A total of 36,528 size measure-

ments were recorded and the length weight relationship

(y = 0.006 x3.22) provided a good fit to length and weight

data (r2 = 0.99).

A total of 1,827 otoliths were used to age L. nebulosus

Figure 2. Photomicrographs of transverse sections through

the sagittal otoliths of (a) D. pictum (b) E. coioides and (c) L.

nebulosus viewed with reflected light. Black dots indicate

annually deposited opaque bands and the axes along which

age readings were made (scale bar = 1 mm).

E. M. GRANDCOURT ET AL.

of which 48 samples were rejected due to a lack of agree-

ment between readings. Fish ranged in size from 19.0 to

57.8 cm LF and age from 1 to 14 yrs. A high level of re-

peatability was achieved in age readings with an APE

value of 5.2%. A total of 25,432 size measurements were

recorded and the length weight relationship (y = 0.026

x2.89) provided a good fit to length and weight data (r2 =

0.98).

3.2. Selectivity

For D. Pictum, harvest rates ranged from 43%-53% and

the level of juvenile retention reached 64.9% in 2004.

Furthermore, the mean size at first capture (24.7-31.6 cm

LF) was below the mean size at first sexual maturity

(31.8 cm LF) for all years (Table 1). The estimate of the

length giving the highest yield (Lopt) of 53.4 cm LF was

considerably greater than the mean size at first capture

Table 1. Sample sizes (n), total mortality rate (Z), fishing mortality rate (F), target fishing mortality rate (F

0.1

), limit fishing

mortality rate (F

max

), harvest rate (H), juvenile retention (J) and selectivity parameters (Lc

, tc) for (a) D. Pictum (b) E.

Coioides and (c) L. nebulosus in the southern Arabian Gulf from 2001 to 2008.

(a)

H Lc

Year n Z F (95% CI) F

0.1

max

(%) J (cm L

) (yrs)

2001 - - - - - - - - -

2002 3567 0.74 0.61 (0.68-0.53) 0.07 0.08 43.0 34.5 28.4 1.1

2003 2919 0.80 0.67 (0.77-0.57) 0.07 0.08 46.0 45.8 24.7 0.7

2004 1653 0.89 0.76 (0.88-0.64) 0.07 0.08 50.4 64.9 24.9 0.7

2005 3108 0.81 0.68 (0.80-0.56) 0.07 0.08 46.7 36.6 28.8 1.1

2006 1117 0.86 0.73 (0.83-0.62) 0.07 0.08 48.7 29.3 31.6 1.5

2007 2602 0.95 0.82 (0.96-0.68) 0.07 0.08 53.0 35.4 31.0 1.4

2008 2309 0.90 0.77 (0.91-0.62) 0.07 0.08 50.6 45.6 29.2 1.2

Mean 2468 0.85 0.72 (0.83-0.60) 0.07 0.08 48.3 41.7 28.4 1.1

(b)

H Lc

Year n Z F (95% CI) F

opt

limit

(%) J (cm L

) (yrs)

2001 19543 1.00 0.81 (1.05-0.56) 0.10 0.12 51.0 31.1 40.4 1.1

2002 5595 0.96 0.77 (1.02-0.51) 0.12 0.13 49.4 30.4 48.4 1.4

2003 3074 1.15 0.96 (1.25-0.66) 0.09 0.11 56.9 44.2 36.6 1.4

2004 1856 1.12 0.93 (1.29-0.58) 0.09 0.11 56.1 53.9 36.3 1.5

2005 3096 0.98 0.79 (1.05-0.54) 0.12 0.13 50.5 23.3 48.3 1.8

2006 866 0.95 0.76 (1.14-0.39) 0.12 0.14 49.2 12.8 50.3 2.5

2007 2658 1.14 0.95 (1.22-0.67) 0.12 0.13 56.5 23.5 48.9 2.0

2008 2361 0.96 0.77 (1.00-0.55) 0.11 0.13 49.7 23.2 47.7 1.8

Mean 4881 1.03 0.84 (1.13-0.56) 0.11 0.12 52.4 30.3 44.6 1.7

(c)

H Lc

Year n Z F (95% CI) F

opt

limit

(%) J (cm L

) (yrs)

2001 5346 0.6639 0.47 (0.57-0.36) 0.160.1834.0 8.1 38.9 1.0

2002 5604 0.6752 0.48 (0.59-0.36) 0.160.1934.6 13.1 39.6 0.7

2003 3194 0.6432 0.44 (0.53-0.36) 0.120.1432.8 16.9 30.1 0.6

2004 1835 0.6819 0.48 (0.60-0.37) 0.110.1435.0 25.1 29.6 1.0

2005 3469 0.6299 0.43 (0.55-0.31) 0.120.1532.0 14.2 32.2 0.7

2006 830 0.6974 0.50 (0.61-0.39) 0.140.1635.9 8.2 35.5 1.5

2007 2753 0.5692 0.37 (0.50-0.24) 0.130.1628.2 6.5 34.6 1.1

2008 2447 0.6583 0.46 (0.58-0.34) 0.140.1733.7 9.6 35.6 1.1

Mean 3185 0.65 0.45 (0.57-0.34) 0.140.1633.3 12.7 34.5 0.9

E. M. GRANDCOURT ET AL. 87

for all years. The results of Mann-Whitney U tests to com-

pare selectivity metrics indicated that there were no sig-

nificant differences in juvenile retention (P = 1.0), mean

size (P = 0.7), mean size at first capture (P =0.12)

and mean age at first capture (P = 0.24) following the

introduction of the mandatory escape panel regulation

(Figure 3). The results of partial correlations to evaluate

the impact of the introduction of the mandatory escape-

panel regulations on trends also indicated that there were

no significant changes in juvenile retention (P = 0.48),

mean size (P = 0.50), mean size at first capture (P = 0.35)

and mean age at first capture (P = 0.32) (Figure 4).

For E. coioides, harvest rates ranged from 49.2%-

56.9% and the level of juvenile retention reached 53.9%

in 2004. Furthermore, the mean size at first capture

(36.3-50.3 cm LT) was below the mean size at first sexual

maturity (42.6 cm LT) for some years (Table 1). The

estimate of the length giving the highest yields (Lopt) of

67.4 cm LT was considerably greater than the mean sizes

at first capture for all years. The results of

Mann-Whitney U tests to compare selectivity metrics

indicated that there were no significant differences in

juvenile retention (P=0.18), mean size (P=0.23) and the

mean size at first capture (P=0.46) following the intro-

D. pict umE. coioidesL. nebulosus

Before

After

Juvenile retention (%)

D. pict umE. coioidesL. nebulosus

Before

After

Meansize (cm)

(b)

D. pict u

E. co io id e

L. nebulosu

Before

After

c50

(cm)

(c)

0.0

0.5

1.0

1.5

2.0

D. pict u

E. co io id e

L. nebulosu

Before

After

(yrs)

(d)

(a )

Figure 3. A comparison of (a) juvenile retention (b) mean size (c) mean size at first capture (L

c50

) and (d) mean age at first

capture (t

) for D. pictum, E. coioides and L. nebulosus before (2001-2003) and after (2004-2008) the implementation of juve-

nile escape panel regulations (± SE).

E. M. GRANDCOURT ET AL.

2000 2002 2004 2006 2008

D. pict umE. coio id e sL. nebulosus

Year

Juvenile retention (%)

2000 2002 2004 2006 2008

D. pict umE. coio id e sL. nebulosus

Year

Mean size (cm)

(b )

2000 2002 2004 2006 2008

D. pict umE. coio id e sL. nebulosus

Year

c50

(cm)

(c)

2000 2002 2004 2006 2008

D. pict umE. c o io id e sL. nebulosus

Year

(yrs)

(d)

(a)

Figure 4. Trends in selectivity metrics between 2001 and 2008 (a) juvenile retention (b) mean size (c) mean size at first cap-

ture (L

c50

) and (d) mean age at first capture (t

) for D. pictum, E. coioides and L. nebulosus in the southern Arabian Gulf.

duction of the mandatory escape panel regulation. There

was, however, a significant increase (P=0.03) in the mean

age at first capture for E. Coioides from 1.3 yrs to 1.9 yrs

following the introduction of the mandatory escape panel

regulation (Figure 3). The results of partial correlations

to evaluate the impact of the introduction of the manda-

tory escape panel regulations on trends indicated that

there were no significant changes in juvenile retention

E. M. GRANDCOURT ET AL.89

(P= 0.35), mean size (P = 0.47), mean size at first cap-

ture (P = 0.38) and mean age at first capture (P = 0.29)

(Figure 4).

For L. nebulosus, harvest rates ranged from 28.2%-

35.9% and the level of juvenile retention reached 25.1%

in 2004. The mean size at first capture (29.6-39.6 cm LF)

was above the mean size at first sexual maturity (27.6 cm

LT) for all years (Table 1). The estimate of the length

giving the highest yield (Lopt) of 41.2 cm LF was consid-

erably greater than the mean sizes at first capture for all

years. The results of Mann-Whitney U tests to compare

selectivity metrics indicated that there were no signifi-

cant differences in juvenile retention (P = 0.88), mean

size (P = 0.30), mean size at first capture (P = 0.30) and

mean age at first capture (P = 0.07) following the intro-

duction of the mandatory escape panel regulation (Fig-

ure 3). The results of partial correlations to evaluate the

impact of the introduction of the mandatory escape panel

regulations on trends indicated that there were no sig-

nificant changes in juvenile retention (P = 0.22), mean

size (P = 0.30), mean size at first capture (P = 0.70) and

mean age at first capture (P = 0.74) (Figure 4).

Comparisons of selectivity curves generated using

pooled size frequency data collected before and after the

introduction of juvenile escape panels with selectivity

curves generated with data collected from an experimen-

tal fishing program, indicated that there was a negligible

change in the selectivity characteristics of traps fitted

with escape panels by comparison with traps having

regular mesh for D. Pictum and E. Coioides (Figure 5a

&b). For L. nebulosus, there was a detectable reduction

in the selectivity characteristics of traps fitted with es-

cape panels (Figure 5c) although this could not be veri-

fied independently as there was insufficient data from the

experimental fishing program for this species.

3.3. Catch and Effort

Whilst reductions in catches occurred for the families

Haemulidae (P = 0.46), Serranidae (P = 0.10) and Leth-

rinidae (P = 0.53) following the introduction of effort

reductions (Figure 6a), none were significant. The mean

annual number of dhow trips reduced from 7,372 to

5,944 following the introduction of effort reductions

(Figure 6b), although the drop was not significant (P =

0.10). The results of partial correlations to evaluate the

impact of effort reductions on trends in catch and effort

indicated that there were no significant changes in catch

trends for the Haemulidae (P = 0.65), Serranidae (P =

0.80) and Lethrinidae (P = 0.80) (Figure 7a). There were

also no significant changes in effort trends in terms of

of the number of dhow trips per year (P=0.71) (Figure

7b).

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 1020304050607

Probability of capture

Fork length (cm)

Experimental

Before

After

(a)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 102030405060708090

Probability of capture

Total length (cm)

Experimental

Before

After

(b)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 1020304050607

Pr obabilit y of c apt ure

Fork length (cm)

Before

After

(c)

Figure 5. Selectivity curves for (a) D. pictum (b) E. coioides

and (c) L. nebulosus before (2001-2003) and after (2004-

2008) the implementation of juvenile escape panel regula-

tions. The dashed line ‘Experimental’ indicates independent

selectivity curves established using data collected from an

experimental fishing program. Note there was insufficient

data from the experimental fishing program to be able to

generate a selectivity curve for L. nebulosus.

E. M. GRANDCOURT ET AL.

500

1000

1500

2000

2500

HaemulidaeSerranidaeLethrinidae

Before

After

Mean annual catch (mt)

2000

4000

6000

8000

Before After

Mean annual effort

(no.dhow trips)

(b)

(a)

Figure 6. Comparisons of (a) the mean annual catch of the families Haemulidae, Serranidae and Lethrinidae (b) mean annual

effort in terms of the number of dhow trips and (b) mean annual effort in terms of the number of trap sets before (2001-2003)

and after (2004-2008) the implementation of fishing effort reductions.

1000

2000

3000

4000

2000 2002 2004 2006 2008

Haemulidae Serranidae Lethrinidae

Year

Totalcatch (mt )

2000

4000

6000

8000

10000

2000 2002 2004 2006 2008

Year

Total effor t ( no. dhow trips)

(b)

(a)

Figure 7. Trends in (a) the total catch of the families Haemulidae, Serranidae and Lethrinidae (b) total effort in terms of the

number of dhow trips and (b) total effort in terms of the number of trap sets in the southern Arabian Gulf between 2001 and

2008.

E. M. GRANDCOURT ET AL. 91

3.4. Mortality, Harvest Rates, Yield and

Spawner Biomass Recruit

For D. pictum, the fishing mortality rate estimates

(0.61-0.82) were considerably greater than both the tar-

get (F0.1) and limit (Fmax) biological reference points for

all years (Table 1), indicating that the population is heav-

ily overexploited. Consequently, estimates of the relative

spawner biomass per recruit, which ranged from 3.2%-

6.6%, were particularly low (Table 2). Values of the

alternative target and limit biological reference points of

0.5M (0.07) and 2/3M (0.09) respectfully, closely resem-

bled mean values of F0.1 (0.07) and Fmax (0.09). The re-

sults of Mann-Whitney U tests to compare fishery met-

rics indicated that there were no significant differences in

fishing mortality rates (P = 0.053), YPR (P = 1.00), SBR

(P = 0.70) and harvest rates (P = 0.53) following the

introduction of effort reductions (Figure 8). The results

of partial correlations to evaluate the impact of effort

reductions on trends also indicated that there were no

significant changes in the fishing mortality rate (P =

0.49), YPR (P = 0.65), SBR (P = 0.81) and harvest rate

(P = 0.32) (Figure 9).

For E. coioides, the fishing mortality rate estimates

(0.76-0.98) were considerably greater than the target

(F0.1) (0.09-0.12) and limit (Fmax) (0.11-0.14) biological

reference points for all years (Table 1), indicating that

the population is heavily overexploited. Consequently,

estimates of the relative spawner biomass per recruit,

which ranged from 1.4%-5.9%, were particularly low

(Table 2). Values of the alternative target and limit bio-

logical reference points of 0.5M (0.10) and 2/3M (0.13)

respectfully, closely resembled mean values of F0.1 (0.11)

and Fmax (0.12). The results of Mann-Whitney U tests to

compare fishery metrics indicated that there were no

significant differences in fishing mortality rates (P=0.55),

YPR (P = 0.053), SBR (P = 0.18) and harvest rates

(P=0.66) following the introduction of effort reductions

(Figure 8). The results of partial correlations to evaluate

the impact of effort reductions on trends also indicated

that there were no significant changes in the fishing

mortality rate (P = 0.75), YPR (P = 0.33), SBR (P = 0.56)

and harvest rate (P = 0.95) (Fi gur e 9).

For L. nebulosus, the fishing mortality rate estimates

(0.37-0.50) were considerably greater than the target

(F0.1) (0.11 - 0.16) and limit (Fmax) (0.14 - 0.19) biologi-

Table 2. Relative spawner biomass per recruit (SBR) and yield per recruit (YPR) for D. Pictum, E. Coioides and L. nebulosus

in the southern Arabian Gulf from 2001 to 2008.

Year Species SBR (%) (95% CI) YPR (g)

D. pictum - -

2001 E. coioides 1.8 (4.9-0.7) 890.8

L. nebulosus 6.9 (10.8-4.7) 119.9

D. pictum 6.6 (8.3-5.3) 387.2

2002 E. coioides 2.5 (6.9-1.0) 1059.0

L. nebulosus 5.9 (9.8-3.8) 99.3

D. pictum 4.2 (5.9-3.1) 266.8

2003 E. coioides 1.4 (4.0-0.6) 985.8

L. nebulosus 6.3 (9.6-4.3) 97.1

D. pictum 3.2 (4.6-2.3) 231.4

2004 E. coioides 1.5 (5.5-0.5) 997.3

L. nebulosus 6.4 (10.5-4.2) 115.0

D. pictum 5.4 (7.5-4.0) 350.6

2005 E. coioides 3.2 (7.7-1.5) 1295.6

L. nebulosus 7.1 (12.8-4.3) 111.8

D. pictum 6.5 (8.2-5.3) 438.4

2006 E. coioides 5.9 (8.0-2.5) 1787.0

L. nebulosus 7.8 (11.8-5.5) 152.4

D. pictum 5.0 (6.8-3.9) 379.4

2007 E. coioides 2.4 (5.4-1.1) 1300.2

L. nebulosus 11.0 (20.4-6.7) 166.5

D. pictum 4.6 (6.7-3.4) 336.9

2008 E. coioides 3.7 (7.8-1.9) 1365.6

L. nebulosus 7.6 (12.9-4.9) 134.7

E. M. GRANDCOURT ET AL.

0.0

0.2

0.4

0.6

0.8

1.0

D. pictumE. coioidesL. nebulosus

Before

After

Fishingmortality rate (F)

0.0

0.4

0.8

1.2

1.6

D. pic t umE. coioidesL. nebulosus

Before

After

Yield per recruit (kg)

(b)

D. pic t umE. coioidesL. nebulosus

Before

After

Spawner biomass

per recruit (%)

(c)

(a)

D. pictu

E. c o ioide

L. nebulosu

Before

After

Harvest rate (%)

(d)

Figure 8. A comparison of (a) fishing mortality rates (b) YPR (c) SBR and (d) harvest rates for D. pictum, E. coioides and L.

nebulosus before (2001-2003) and after (2004-2008) the implementation of juvenile escape panel regulations and effort reduc-

tions (± SE).

cal reference points for all years (Table 1), indicating

that the population is heavily overexploited. Conse-

quently, estimates of the relative spawner biomass per

recruit, which ranged from 5.9%-11.0%, were particu-

larly low (Table 2). Values of the alternative target and

limit biological reference points of 0.5M (0.10) and 2/3M

(0.13) respectfully, were less than the mean values of F0.1

(0.14) and Fmax (0.16). The results of Mann-Whitney U

tests to compare fishery metrics also indicated that there

were no significant differences in fishing mortality rates

(P = 0.76), YPR (P = 0.10), SBR (P = 0.053) and harvest

rates (P = 0.88) following the introduction of effort re-

ductions (Figure 8). The results of partial correlations to

evaluate the impact of effort reductions on trends indi-

cated that there were no significant changes in the fishing

mortality rate (P = 0.65), YPR (P = 0.57), SBR (P = 0.31)

and harvest rate (P = 0.44) (Fig ure 9).

3.5. Age Structure

Analysis of the age structure of D. pictum indicated that

truncation was exacerbated following the introduction of

management measures in 2004. While there were no sig-

nificant differences in the relative proportions of age

E. M. GRANDCOURT ET AL. 93

0.0

0.2

0.4

0.6

0.8

1.0

1.2

2000 2002 2004 2006 2008

D. pict umE. c o io id e sL. nebulosus

Year

Fishing mortality rate (F)

0.0

0.5

1.0

1.5

2.0

2000 2002 2004 2006 2008

D. pict umE. co io id e sL. nebulosus

Year

Yield per recruit (kg)

(b)

Year

(b)

2000 2001 2002 2003 2004 2005 2006 2007 2008

D. pict umE. c o ioid e sL. nebulosus

Year

Spawner biomass per recruit (%)

(c)

2000 2002 2004 2006 2008

D. pict umE. c o ioidesL. nebulosus

Year

Harvest rate (%)

(d)

(a)

Figure 9. Trends in (a) fishing mortality rates (b) YPR (c) SBR and (d) harvest rates for D. pic tum , E. coioides and L. nebulo-

sus, between 2001 and 2008 in the southern Arabian Gulf.

classes up to 9 yrs, there was a significant reduction in

the proportion of the oldest (10 to 13 yr) age classes

(Figure 10a). The same pattern was identified for E.

coioides which also had no significant differences in the

relative proportions of age classes up to 9 yrs and a sig-

nificant reduction in the proportion of the oldest (10 and

12 yr) age classes (Fi gure 10b). For L. nebulosus, there

were no significant differences between the 1 yr to 13 yr

E. M. GRANDCOURT ET AL.

0.52

0.63

0.85

0.84

0.83

0.74 0.58 0.13 0.04* 0.01*

123456789101112

Before

After

Age Class (yrs)

Proportion (%)

(b)

0.87 0.67

0.87

0.65 0.610.560.29 0.200.07 0.01*0.01*0.01*0.01*

12345678910111213

Before

After

Age Class (yrs)

Proportion (%)

(a)

0.70

0.62 0.52

0.77

0.84

0.66

0.67

0.54

0.490.490.38 0.380.060.01*

1234567891011121314

Before

After

Age Class (yrs)

Proportion (%)

(c)

Figure 10. Comparison of age structures before (2001-2003) and after (2004-2008) the introduction of management regula-

tions for (a) D. pictum (b) E. coioides and (c) L. nebulosus. Data labels above bars show the results of 

goodness of fit tests (P

values) for individual age classes, * indicates a significant difference.

age classes and a significant reduction in the oldest (14

yr) age class following the introduction of management

regulations (Figure 10c).

4. Discussion

The principal conclusion of this study is that gear regula-

tions and effort reductions implemented between 2004

and 2008 have failed to achieve the desired management

targets for the key species in the demersal trap fishery of

the southern Arabian Gulf. With the exception of a minor

increase in the mean age at first capture from 1.3 yrs to

1.9 yrs for E. coioides, there were no significant changes

in the values or trends of any of the selectivity and other

fishery metrics in the 5 yr period following the introduc-

tion of juvenile escape panels and limits on the number

E. M. GRANDCOURT ET AL.95

of traps used per vessel.

The use of selectivity ogives established through an

experimental fishing program, being independent of

those derived from sampled fisheries data were of tre-

mendous utility to the analyses, particularly from a

management perspective. The ogives for E. coioides and

D. pictum showed that there was a negligible change in

the selectivity characteristics of traps fitted with escape

panels by comparison with traps having a regular mesh

and no panel. This independently corroborates the results

of comparisons of selectivity parameters before and after

the introduction of the gear regulations and clearly indi-

cates that the failure of escape panels to modify the se-

lectivity characteristics of the fishery cannot be attributed

to a lack of compliance by fishers.

The large differences between estimates of the sizes at

first capture with those which would give the highest

yields, emphasizes the need to revise existing gear regu-

lations for the trap fishery. This is of critical importance

to a stock rebuilding strategy, particularly given the det-

rimental impact that the depletion of the older age classes

can have on reproductive capacity and output, as has

been demonstrated for E. coioides [14]. However, be-

cause of the differences among species, it is not possible

to have a gear regulation that would select all species at

the optimum sizes and ages at first capture. For this rea-

son, alternative management regulations also need to be

considered.

The utility of the catch and effort data was limited as

there were no reliable estimates of the total number of

traps used before and after the introduction of effort re-

strictions. Furthermore, species specific catch data across

the entire time series and catches made by vessels from

other emirates were not available. Nevertheless, as there

were no significant changes in the values or trends in

catch and effort or any of the other fishery metrics (F,

YPR, SBR, H) following the introduction of effort con-

straints, this management regulation has also clearly

failed to achieve any progress in terms of stock rebuild-

ing.

A more detailed understanding of recruitment would

have helped to elucidate the variability in relative spaw-

ner biomass per recruit over the study period. An associ-

ated limitation of the YPR analyses is the assumption

that there is no relationship between the size of the

spawning stock biomass and subsequent recruitment over

the range of fishing mortality rates used [17]. This is

particularly restrictive for small, fast growing tropical

species with high rates of natural mortality, for which

predictions suggest that high yields per recruit may be

achieved at levels of exploitation where the remaining

spawning stock biomass may not be capable of sustain-

ing recruitment [11]. As values of the target and limit

biological reference points (F0.1 and Fmax) approximated

the precautionary alternatives (0.5M and 2/3M) for D.

pictum and E. coioides, the failure of the model was not

apparent for these species. Conversely, the target and

limit biological reference points (F0.1 and Fmax) for L.

nebulosus were greater than the precautionary alterna-

tives (0.5M and 2/3M) suggesting that the YPR analyses

may have over-estimated sustainable fishing mortality

rates for this species.

Other potential sources of error include the use of a

single age-length key and the associated assumption that

there is no significant inter-annual variability in growth.

This would be particularly important in the developmen-

tal phases of a new fishery where density dependant ef-

fects may occur. Given that the existing fishery is well

established, this assumption may have been less of a

limitation to the analyses than the absence of fish in

samples that were at the extremes of the size and age

ranges.

Whilst increments in the sagittal otolith sections of the

study species have been shown to be formed annually

[3,7], it is important to distinguish between the validation

of increment periodicity and absolute age [18]. As the

absolute age of the study species was not validated, im-

provements in the age data and associated analyses could

have been achieved using an independent validation

means such as a mark recapture study using chemical

marking of juvenile fish with a known age.

The maximum age estimate of D. pictum (13 yrs) was

considerably less than the maximum age of 31 yrs esti-

mated by [19] for this species in New Caledonia. Like-

wise, our longevity estimate of 14 yrs for L. nebulosus

was less than those of [20] (20 yrs) and (21 yrs) [21] for

this species in the northern Arabian Gulf and Gulf of

Aden respectively. Additionally, the maximum age of E.

coioides here (12 yrs) was considerably less than the

maximum age of 22 yrs [20]. The large discrepancy be-

tween the maximum observed ages in our study by com-

parison with those recorded elsewhere can be attributed

to the truncation of the age structures through intensive

fishing. Whilst size specific selectivity cannot be dis-

carded as a possible explanation for the small proportion

of larger and older fish in size frequency and biological

samples, the impact of fishing on the size and age struc-

ture of the respective populations is considered a more

likely explanation, particularly given the historic absence

of regulation in the fishery. The analyses of age struc-

tures of the key species indicated that truncation was

exacerbated following the introduction of management

measures in 2004, further highlighting the failure of ex-

isting regulations to rebuild stocks. Whilst sample sizes

of the oldest age classes were small, significant reduc-

tions were observed for all key species following the

E. M. GRANDCOURT ET AL.

introduction of regulations, in contrast to the intended

impact.

YPR and SBR analyses indicated that the study spe-

cies were all heavily over-exploited. As the existing fi-

shing mortality rates were all well in excess of the limit

biological reference points, growth over-fishing had

clearly occurred. Furthermore, with current values of the

relative spawner biomass per recruit being less than 10%

in all cases, recruitment over-fishing would also have

occurred, based on meta-analyses [22]. The critical ma-

nagement issues that relate to the stock status established

in earlier assessments [3,7] are clearly still pertinent.

There is widespread recognition that target species are

over-exploited in places where wire-mesh fish traps have

been used extensively [23,24]. Given the failure of ex-

isting regulations to modify gear selectivity, reduce ef-

fort and rebuild stocks, management authorities should

consider alternative measures including a moratorium on

the use of traps in the off-shore demersal fishery of Abu

Dhabi.

5. Acknowledgements

This study forms part of the activities of the ‘Fish Land-

ings and Population Dynamics Project’ (project # 02-

23-0008-09) which is implemented by the Biodiversity

Management Sector―Marine, of the Environment Agen-

cy―Abu Dhabi. The management of the Environment

Agency―Abu Dhabi is thanked for their support for this

work. The Abu Dhabi Fishermen’s Cooperative Society

facilitated size frequency sampling and catch data collec-

tion. The UAE Coast Guard and later CNIA (Critical

National Infrastructure Authority) provided effort re-

cords. All the EAD enumerators are thanked for their

contribution to data collection; Khalid Al Ali, Sultan Al

Ali, Hamad Al Shamsi, Mohamed Al Zaabi, Sabah Has-

san, Jihad Hassan, Mohamed Ahmed Hassan, Abdulla Al

Blooki, Abdulla Muhyiddeen, Eid Al Romaithi, Yaqoob

Yousif Al Hammadi.

6. References

[1] E. M. Grandcourt, “Fish and Fisheries,” In: T. Z. Al Ab-

dessalaam, Ed., Marine Environment and Resources of

Abu Dhabi, Motivate Publishing, Dubai, 2008, pp. 200-

225.

[2] B. Shallard, “Distribution and Abundance of Demersal

Fish Stocks in the UAE. Technical Report 1. Fish Re-

sources Assessment Survey Project of Abu Dhabi and

UAE Waters,” Environmental Research and Wildlife

Development Agency, Government of Abu Dhabi, Abu

Dhabi, 2003, p. 211.

[3] E. M. Grandcourt, T. Z. Al Abdessalaam, A. T. Al Shamsi

and F. Francis, “Biology and Assessment of the Painted

Sweetlips (Diagramma pictum (Thunberg, 1792)) and

Spangled Emperor (Lethrinus nebulosus (Forsskål, 1775))

in the Southern Arabian Gulf,” Fisheries Bulletin, Vol.

104, No. 1, 2006, pp. 75-88.

[4] E. M. Grandcourt, T. Z. Al Abdessalaam, F. Francis and

A. Al Shamsi, “Population Biology and Assessment of

the White-Spotted Spinefoot, Siganus Canaliculatus (Park,

1797), in the southern Arabian Gulf,” Journal of Applied

Ichthyology, Vol. 23, No. 1, 2007, pp. 53-59.

doi:10.1111/j.1439-0426.2006.00796.x

[5] E. M. Grandcourt, T. Z. Al Abdessalaam, F. Francis and

A. Al Shamsi, “Biology and Stock Assessment of the

Sparids, Acanthopagrus Bifasciatus and Argyrops Spini-

fer (Forsskål, 1775), in the Southern Arabian Gulf,” Fi-

sheries Research, Vol. 69, No. 1, 2004, pp. 7-20.

doi:10.1016/j.fishres.2004.04.006

[6] H. Al Masroori, H. Al Oufi, J. McIlwain and E. McLean,

“Catches of Lost Fish Traps (ghost fishing) from Fishing

Grounds near Muscat, Sultanate of Oman,” Fisheries Re-

search, Vol. 69, No. 3, 2004, pp. 407-414.

doi:10.1016/j.fishres.2004.05.014

[7] E. M. Grandcourt, T. Z. Al Abdessalaam, A. T. Al Shamsi

and F. Francis, “Population Biology and Assessment of

the Orange-Spotted Grouper, Epinephelus coioides (Ha-

milton, 1822), in the Southern Arabian Gulf,” Fisheries

Research, Vol. 74, No. 1, 2005, pp. 55-68.

doi:10.1016/j.fishres.2005.04.009

[8] R. J. Beamish and D. A. Fournier, “A Method for Com-

paring the Precision of a Set of Age Determinations,”

Canadian Journal of Fisheries and Aquatic Sciences, Vol.

38, No. 8, 1981, pp. 982-983. doi:10.1139/f81-132

[9] W. E. Ricker, “Computation and Interpretation of Bio-

logical Statistics of Fish Populations,” Bulletin of the

Fisheries Research Board of Canada, Vol. 191, 1975, pp.

2-6.

[10] R. J. H. Beverton and S. J. Holt, “On the Dynamics of

Exploited Fish Populations,” Chapman and Hall, London,

1957, p. 179.

[11] F. C. Gayanilo Jr. and D. Pauly, “FAO-ICLARM Stock

Assessment Tools, Reference Manual, ICLARM,” Food

and Agricultural Organization of the United Nations, Rome,

1997.

[12] D. Pauly, “Length-Converted Catch Curves: A Powerful

Tool for Fisheries Research in the Tropics (Part 1),”

Fishbyte, Vol. 1, No. 2, 1983, pp. 9-13.

[13] R. J. H. Beverton, “Patterns of Reproductive Strategy

Parameters in Some Marine Teleost Fish,” Journal of

Fish Biology, Vol. 41, No. B, 1992, pp. 137-160.

[14] E. M. Grandcourt, T. Z. Al Abdessalaam, F. Francis, A. T.

Al Shamsi and S. A. Hartmann, “Reproductive Biology

and Implications for Management of the Orange-Spotted

Grouper, Epinephelus coioides (Hamilton, 1822), in the

southern Arabian Gulf,” Journal of Fish Biology, Vol. 74,

2009, pp. 820-841.

doi:10.1111/j.1095-8649.2008.02163.x

[15] K. Patterson, “Fisheries for Small Pelagic Species: An

Empirical Approach to Management Targets,” Reviews in

Fish Biology and Fisheries, Vol. 2, No. 4, 1992, pp.

E. M. GRANDCOURT ET AL.

321-338.

doi:10.1007/BF00043521

[16] R. R. Sokal and F. J. Rohlf, “Biometry,” The Principles

and Practice of Statistics in Biological Research, 3rd Edi-

tion, W. H. Freeman and Co, New York, 1995.

[17] C. D. Buxton, “The Application of Yield-per-Recruit Mod-

els to Two South African Sparid Reef Species, with Special

Consideration to Sex Change,” Fisheries Research, Vol. 15,

No. 1-2, 1992, pp. 1-16.

doi:10.1016/0165-7836(92)90002-B

[18] S. E. Campana, “Accuracy, Precision and Quality Control

in Age Determination, Including a Review of the Use and

Abuse of Age Validation Methods,” Journal of Fish Bi-

ology, Vol. 59, No. 2, 2001, pp. 197-242.

doi:10.1111/j.1095-8649.2001.tb00127.x

[19] G. Loubens, “Biologie de Quelques Espèces de Poissons

du Lagon Néo-Calédonien. III. Croissance,” Cahiers Indo-

Pacific, Vol. 2, No. 2, 1980, pp. 101-153.

[20] C. P. Mathews and M. Samuel, “Growth, Mortality and

Length-Weight Parameters for Some Kuwaiti Fish and

Shrimp,” Fishbyte, Vol. 9, No. 2, 1991, pp. 30-33.

[21] R. R. C. Edwards and S. Shaher, “The Biometrics of Ma-

rine Fishes from the Gulf of Aden,” Fishbyte, Vol. 9, No.

2, 1991, pp. 27-29.

[22] P. M. Mace, “Relationships between Common Biological

Reference Points Used as Thresholds and Targets of Fish-

eries Management Strategies,” Canadian Journal of Fish-

eries and Aquatic Sciences, Vol. 51, No. 1, 1994, pp.

110-122.

doi:10.1139/f94-013

[23] R. Mahon and W. Hunte, “Trap Mesh Selectivity and the

Management of Reef Fishes,” Fish and Fisheries, Vol. 2,

No. 4, 2001, pp. 356-375.

doi:10.1046/j.1467-2960.2001.00054.x

[24] J. Stewart and D. J. Ferrell, “Mesh Selectivity in the

NSW Demersal Trap Fishery,” Fisheries Research, Vol.

59, No. 3, 2003, pp. 379-392.

doi:10.1016/S0165-7836(02)00024-3