B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

. sc iencedi rec t . com

ava i lab le a t www

journal homepage: www.elsevier .com/ locate /b iocon

Populations of inshore serranids across theCanarian Archipelago: Relationships with humanpressure and implications for conservation

Fernando Tuyaa,*, Pablo Sanchez-Jerezb, Ricardo J. Harouna

aBIOGES, Department of Biology, Marine Sciences Faculty, Campus Tafira, University of Las Palmas de G.C., 35017, Las

Palmas, Canary Islands, SpainbDeparment of Marine Sciences, University of Alicante, P.O. Box 99 E-03080, Alicante, Spain

A R T I C L E I N F O

Article history:

Received 24 December 2004

Received in revised form 5 April 2005

Accepted 20 September 2005

Available online 2 November 2005

Keywords:

Fishing pressure

Groupers

Combers

Hierarchical design

Canary Islands

0006-3207/$ - see front matter � 2005 Elsevidoi:10.1016/j.biocon.2005.09.012

* Corresponding author: Tel.: +34 928457456;E-mail address: ftuya@yahoo.es (F. Tuya).

A B S T R A C T

We investigated spatio-temporal variability in the abundances and biomasses of four spe-

cies of inshore serranids (the dusky grouper Ephinephelus marginatus, the island grouper

Mycteroperca fusca, the painted comber Serranus scriba, and the blacktail comber S. atricauda)

throughout the Canarian Archipelago (central-east Atlantic Ocean) with underwater visual

transects. By means of a multiscaled sampling design spanning three orders of magnitude

of spatial variability (from 10 s of meters among replicated 100 m2 transects to 100 s of kilo-

metres among islands) and four sampling periods, we related differences in the distribu-

tions of serranids to differences in the degree of human pressure, such as fishing

intensity and human population. Differences in human pressure among islands provide

the most parsimonious explanation for many of the consistent inter-island differences in

the abundance and biomass of the analyzed species. Larger-bodied serranids (E. marginatus

and M. fusca) are more vulnerable than the smaller species (S. scriba and S. atricauda). In

fact, the larger, more vulnerable species have been almost completely extirpated from

the most intensely fished islands. Our results show that the larger groupers have been over-

exploited throughout the Canary Islands, and highlight the urgent need for stringent man-

agement measures and better control of littoral reef fish resources.

� 2005 Elsevier Ltd. All rights reserved.

1. Introduction

Coastal development, pollution, and principally fisheries, ex-

ert the most pervasive influences on inshore natural

resources and ecosystems (Beger et al., 2003; Jackson 2001).

Fisheries remove large, long-lived, or slow-growing fish that

are frequently replaced by those with higher turnover rates

(Pitcher, 2001; Friedlander and De Martini, 2002, and refer-

ences therein). Serranids are a family of sedentary, long-

lived, and slow-growing reef fish that include, among others,

groupers, combers, and basses. These top-level predators,

er Ltd. All rights reserved

fax: +34 928457457.

especially the different species of groupers, are among the

most highly valued of all reef-associated fish, and are being

increasingly targeted for human consumption by both pro-

fessional and recreational fishers in temperate and tropical

seas throughout the world (Bohnsack et al., 1994; Carter

et al., 1994; Chiappone et al., 2000). Even artisanal fisheries

have reached levels of intensity that are unsustainable for

this vulnerable group of fish (Gobert, 1994; Huntsman

et al., 1994; Sadovy, 1994; Chiappone et al., 2000). The various

pressures on this family have led to marked declines in sev-

eral species (Chiappone et al., 2000) and, in 1996; groupers

.

14 B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

represented an important proportion of all marine fishes in-

cluded in the IUCN Red List (www.redlist.org). Consequently,

we urgently need to assess the current status and the

exploitation patterns of these animals worldwide, as the

Species Survival Commission of the IUCN have recently

highlighted.

Several species of the family Serranidae are found along

shallow hard-bottoms in the warm-temperate waters of the

Canary Islands, central east-Atlantic (www.pescabase.org;

Brito et al., 2002). Most of these species are prized food fishes

in both recreational and commercial fisheries (Bas et al.,

1995), especially large-sized groupers such as Ephinephelus

marginatus and Mycteroperca fusca. Other small or medium-

sized serranid species, such as those belonging to the genus

Serranus, are less targeted. In addition, these species have

higher population turnover rates and are consequently less

vulnerable to overfishing than groupers. Some studies have

investigated the decline in the coastal fisheries resources of

the Canarian Archipelago as a consequence of several dec-

ades of overexploitation (Bas et al., 1995; Falcon et al., 1996).

However, no study has yet empirically assessed the spatial

distribution patterns of serranids across the Canarian Archi-

pelago and related distributions to the degree of fishing and

human development across the islands.

The goal of the present paper was to investigate, through

underwater visual transects, spatio-temporal variability in

the population structure (abundance and biomass) of four

species of inshore serranids (the dusky grouper E. margina-

tus, the island grouper M. fusca, the painted comber

Serranus scriba, and the blacktail comber S. atricauda)

throughout the Canarian Archipelago. We used a multi-

scaled sampling design spanning three orders of magnitude

of spatial variability (from 10 s of meters among replicated

transects to 100 s of kilometres among surveyed islands)

and four sampling periods during the years 2003 and

2004. Specifically, we hypothesized that (1) significant differ-

ences in serranid abundances and biomasses among

surveyed islands are attributable to differences in human

pressure (i.e. fishing intensity and human population), with

the almost complete extirpation of these species from the

most intensely perturbed islands; (2) the impact of human

pressure depends on the life-history characteristics and

the commercial interest of each species; and (3) variability

among islands is consistent through time.

Fig. 1 – Map of

2. Methods

2.1. Sampling design and study locations

We adopted a hierarchical sampling design with randomly

positioned study locations throughout the Canarian Archipel-

ago (Underwood, 1997). We randomly selected 3 locations of

rocky substrate in each of the 7 islands of the Archipelago,

as well as in a group of small islets at the north of Lanzarote

Island, called the Chinijo Archipelago (Fig. 1), at each sam-

pling period. This protocol was repeated in four periods:

March 2003, October 2003, March 2004 and October 2004. All

subtidal sampling was conducted between 10 and 18 m depth,

along rocky-bottoms with similar slopes to minimize the ef-

fect of habitat type on the distribution and patchiness of

assemblages (Garcıa-Charton and Perez-Ruzafa, 1999). All

sampled locations were outside areas completely closed to

fishing (so-called integral no-take reserves). However, some

locations were subject to special fishery regulations and man-

agement, with limitations on the type and amount of fishing

gear (e.g. traps are not allowed).

Information on fishing pressure levels is pivotal to the

interpretation of spatial patterns of reef fish (Ferreira et al.,

2004). We were unable to calculate absolute levels of fishing

pressure among islands because of the lack of fisheries data

and the difficulty of standardizing measures of fishing across

the different gear types in use (Bas et al., 1995). Nevertheless,

using published information in environmental and govern-

mental databases, personal observations, and communica-

tion with fisheries officials and managers, we were able to

rank (i) fishing intensity among islands, as well as (ii) human

population. We consider both indices to be good descriptors

of human pressure on inshore fish resources, as several

investigations have considered (Jennings et al., 1995; Dulvy

et al., 2004; Friedlander and Brown, 2004; Hawkins and Rob-

erts, 2004). The islands constituting the Canarian Archipel-

ago were thus classified into four groups (Table 1) according

to the indices of fishing and human pressure on natural

coastal resources. Although this assumption is simplistic, it

represents an easy way to interpret and discuss our data,

and enables comparisons to be made with the conservation

status and large-scale spatial variability of fish populations

in other oceanic archipelagos and their relationships with

gradients of fishing intensity (Jennings et al., 1995;

study area.

Table 1 – Fishing pressure and human population per coastal perimeter at each island of the Canarian Archipelago

Islands Number offishing ships(official)a

Number offishing ships(observed)b

Number of shipsper coastal perimeter(No. km�1) (observedships in brackets)

Main geartypes

Human population(No. of islanders percoastal perimeter,

No. ind km�1) (No. of touristskm�1 y�1 in brackets)

% Of coastal perimeterwith some form of fishery

regulationc (e.g. trapsare not allowed)

Observations

Highly fished (>0.5 ships per km of island perimeter)/no management

Gran Canaria 375 240 1.54 (0.98) Hook-and-line, traps 3251.31 (11934.15) 0

Tenerife 265 272 0.74 (0.76) Hook-and-line, traps 2234.32 (9776.53) 0

Moderate-to-highly fished (0.15–0.5 ships per km of island perimeter)/no management

Lanzarote 52 95 0.21 (0.38) Hook-and-line, traps 463.51 (7487.17) 0

Fuerteventura 51 523 0.15 (1.53) Hook-and-line, traps 220.65 (4161.02) 0

Gomera 23 78 0.23 (0.79) Hook-and-line, traps 200.92 (2052.33) 0 Targeted by fishermen from

nearby Tenerife island

Moderate-to-highly fished (0.15–0.5 ships per km of island perimeter/little management (<8%)

La Palma 36 110 0.23 (0.70) Hook-and-line, traps 550.5 (784.31) 7

Lightly-to-moderate fished (<0.15 ships per km of island perimeter/moderate management (>8%)

El Hierro 15 59 0.14 (0.55) Hook-and-line, 96.32 (284.36) 9

Chinijo 8 100 0.15 (1.02) Hook-and-line 11.26 (178.57) 100 High number of retired

fishermen fish as recreational

fishers

Fishery characteristics and existing management regulations (as % of coastal perimeter with some form of fishery regulations) are also shown.

a Canarian Government.

b Bas et al. (1995).

c National Spanish network of Marine Protected Areas (www.mapya.es/rmarinas/).

BIO

LO

GICAL

CO

NSERVATIO

N128

(2006)13–24

15

16 B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

Chiappone et al., 2000; Hawkins and Roberts, 2003; Dulvy

et al., 2004; Friedlander and Brown, 2004; Hawkins and Rob-

erts, 2004).

2.2. Fish surveys

Adult and sub-adult fish populations were sampled by means

of visual census techniques. At each sampling location, 8 rep-

licates of 25 m long transects were haphazardly laid during

daylight hours. The abundance and size (total length to the

nearest 2 cm) of each fish species was recorded on waterproof

paper by a SCUBA diver within 2 m of either side of the tran-

sects (100 m2), according to standard procedures (Brock, 1982;

Lincoln-Smith, 1988, 1989; Kingsford and Battershill, 1998;

Hawkins et al., 1999; Denny and Babcock, 2004; Tuya et al.,

2004). Fish abundance was estimated based on a modification

of the method presented by Harmelin-Vivien et al. (1985).

Therefore, when fishes were grouped in schools larger than

20 individuals, their numbers were estimated according to

six abundance classes (20–40, 40–70, 70–150, 150–300, 300–

700, >700). Biomass of each serranid species was further cal-

culated using available length–weight relationships for the

Canarian Archipelago, and from other published and Web-

based sources (Froese and Pauly, 2005). Lengthswere first esti-

mated by assigning each fish to the mid-point of its observed

size range (Miller and Gerstner, 2002). All measured biotic

variables (abundance and biomass per species) were stan-

dardized to an area of 100 m2.

2.3. Data analysis

Abundance and biomass data for each species were analyzed

by means of ANOVA models (Underwood, 1997) to test for dif-

ferences among surveyed islands, sampled locations within

islands and the four sampling periods. The model therefore

incorporated the following experimental factors: (1) ‘‘Island’’

(fixed factor with eight levels corresponding to the seven

main islands plus Chinijo Archipelago), (2) ‘‘Period’’ (random

factor with four levels, and orthogonal to the previous factor),

and (3) ‘‘Locations’’ (random factor nested within islands and

periods, with three levels). Before analysis, Cochran�s test was

used to check for homogeneity of variances. Because fish-

count data contains many zeros, they are usually non-normal

and cannot be effectively transformed (Anderson and Millar,

2004; Hawkins and Roberts, 2004). No transformation ren-

dered homogeneous variances for some descriptors (Coch-

ran�s test, p < 0.01), so the significance level was set at the

0.01 level instead of 0.05, as ANOVA is robust to heterogeneity

of variances, particularly for large balanced experiments

(Underwood, 1997). If ANOVA detected significant differences

for the factor ‘‘Island’’, further analyses were done as inde-

pendent comparisons by using the SNK a-posteriori multiple

comparison test (Underwood, 1997).

Correlation models were used to find significant relation-

ships between the mean abundance and biomass per island

of each fish species and the indices of fishing intensity and

human population shown in Table 1. Moreover, we added

the average number between the official and observed num-

ber of fishing vessels operating per island, as an averaged in-

dex of fishing intensity.

3. Results

3.1. E. marginatus

The dusky grouper, E. marginatus, was the least abundant ser-

ranid species throughout the Canarian Archipelago, with a

mean abundance for the total study of 0.03 ± 0.18 ind

100 m�2 (mean ± SD, n = 768 transects), and a total mean bio-

mass of 168.88 ± 1126.09 g 100 m�2 (mean ± SD, n = 768). The

majority of the specimens (92%, n = 22 individuals) were ob-

served at El Hierro Island and the Chinijo Archipelago

(Fig. 2), while we found only one individual at both Tenerife

and Lanzarote Islands during the study. We did not find E.

marginatus at any of the other islands (Fig. 2). Clear differences

therefore existed in the mean abundances and biomasses of

this species among the surveyed islands (Fig. 2 and Table 2),

which the ANOVA indicated were consistent through time

(non-significance of the interaction term ‘‘P · I’’, Table 2).

Mean abundances and biomasses at El Hierro Island and the

Chinijo Archipelago were similar (p > 0.01, SNK test), with

both islands having greater abundances and biomasses of

dusky groupers than the rest of the islands (p < 0.01, SNK

tests; Fig. 2). Significant intra-island spatial variability (differ-

ences among locations within each island at each sampling

period) was not detected (Table 2).

Although no significant relationships were determined for

mean biomass, themean abundance per island of this grouper

species was significantly correlated with some indices of both

the degree of fishing intensity (rs = �0.38, p < 0.05), and the

human population (rs = �0.40, p < 0.05) (Table 3). Conse-

quently, the islandswith the highest levels of human pressure

on this resource, as indicated by the indices shown in Table 1,

had the lowest abundances, and vice versa.

3.2. M. fusca

The island grouper, M. fusca, was the species that showed the

greatest mean biomass for the entire study (464.63 ±

1464.25 g 100 m�2, mean ± SD, n = 768 transects), while the

mean abundance was 0.32 ± 0.76 ind 100 m�2 (mean ± SD,

n = 768). This serranid was unique among the four species

studied in that its mean biomasses differed among the four

sampling periods (Table 2 and Fig. 3). However, differences

in the mean abundance among periods were not detected

(Table 2). Significant differences in abundances and

biomasses among islands were strong (Table 2); we found

the greatest mean abundance and biomass at El Hierro Island

for the entire study (p < 0.01, SNK tests; Fig. 3). The ANOVA

also detected significant differences among locations within

islands in each period (Table 2), reflecting differences in the

spatial patchiness of this species at this spatial scale (Fig. 3).

The mean abundance and biomass of this species showed

negative significant correlations with some of the indices of

human pressure (Table 3).

3.3. S. scriba

The painted comber, S. scriba, was only found across the east-

ern islands of the Canarian Archipelago, possibly reflecting a

biogeographic distribution pattern between the eastern and

B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4 17

western islands. The mean abundance for the overall study

was 0.05 ± 0.21 ind 100 m�2 (mean ± SD, n = 768 transects),

while the mean biomass was 11.93 ± 57.40 g 100 m�2

(mean ± SD, n = 768). The clear inter-island variability was

confirmed by the ANOVA with the significance of the factor

‘‘Island’’ (Fig. 4 and Table 2). However, we did not record differ-

Fig. 2 – Ephinephelus marginatus. Mean abundance and biomass

values. HI: Hierro, LP: La Palma, GO: Gomera, TF: Tenerife, GC: G

Archipelago.

ences among the islands where this species was found

(p > 0.01, SNK tests; Fig. 4) through time. Additionally, this re-

sult was corroborated by a lack of significance (p > 0.05, Table

3) in the correlation analyses. However, differences in abun-

dances and biomasses of this species did exist at the smallest

spatial scale, with the ANOVA models detecting significant

throughout the study. Error bars are standard errors of mean

ran Canaria, FV: Fuerteventura, LZ: Lanzarote, CH: Chinijo

Ta

ble

2–

An

aly

sis

on

the

eff

ect

so

fp

eri

od

(ra

nd

om

),is

lan

d(fi

xed

an

do

rth

og

on

al)

,a

nd

loca

tio

ns

(ra

nd

om

an

dn

est

ed

wit

hin

isla

nd

sa

nd

sam

pli

ng

peri

od

s)o

nth

em

ea

na

bu

nd

an

cea

nd

bio

ma

sso

fea

chsp

eci

es

Sourceof

variation

df

Ephinep

helusmarginatus

Mycteroperca

fusca

Serranusscriba

Serranusatricauda

Abundance

Biomass

Abundance

Biomass

Abundance

Biomass

Abundance

Biomass

MS

FMS

FMS

FMS

FMS

FMS

FMS

FMS

F

Period=P

30.00

0.18

0.40

0.23

0.07

0.81

83.27

10.16*

0.00

0.06

0.60

0.03

0.06

0.75

31.80

0.34

Island=I

70.05

21.81*

21.34

24.61*

2.45

24.51*

247.93

33.97*

0.05

28.90*

67.11

39.67*

1.37

67.17*

1480.53

66.50*

Loca

tions(P

·I)

64

0.004

0.87

1.75

0.84

0.08

2.23*

8.19

1.99*

0.01

2.19*

21.48

2.48*

0.08

1.78*

94.21

1.89*

P·I

21

0.002

0.53

0.86

0.49

0.10

1.12

7.29

0.89

0.001

0.11

1.69

0.08

0.02

0.25

22.26

0.24

Residual

672

0.005

2.08

0.04

4.12

0.007

8.67

0.04

49.84

*p<0.01.

18 B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

differences among locations within islands at each sampling

period (Table 2).

3.4. S. atriacuda

The blacktail comber, S. atricauda, was the most abundant ser-

ranid observed throughout this study (0.71 ± 0.18 ind 100 m�2,

mean ± SD, n = 768 transects), with a mean biomass of

88.58 ± 231.46 g 100 m�2 (mean ± SD, n = 768). The large-scale

spatial distribution pattern of this species across the Canary

Islands was completely different to those recorded for both

grouper species. That is, the greatest mean abundance and

mean biomass per island were detected in islands with mod-

erate-to-high fishing pressure (Fig. 5). In fact, the mean abun-

dances and biomasses of S. atriacuda at Gomera and La Palma

islands were greater than those recorded at the rest of the is-

lands through time (p < 0.01, SNK tests; Fig. 5). Further, we de-

tected significant within-island variability in both the mean

abundance and biomass of this species (Table 2 and Fig. 5).

A significant correlation between both the mean abundance

and biomass per island of this species and the index of hu-

man population was detected (Table 3).

4. Discussion

The importance of human perturbations as mechanisms that

shape contemporaneous shallow reef fish communities has

been emphasized in the recent literature (e.g. Jackson, 2001;

Dulvy et al., 2004; Ferreira et al., 2004, and references therein).

Although there was evidence to suggest that human pressure

on inshore resources of the Canary Islands were responsible

for some of the significant trends we observed in serranid dis-

tributions, the interpretation of the results cannot be general-

ized across the four serranid species.

The most striking results were found for the two grouper

species (E. marginatus and M. fusca), which showed the stron-

gest responses to variations among islands in both fishing

intensity and human population. This result supports our

first hypothesis: that major human intervention has affected

abundances and biomasses of these two species across the

Canarian Archipelago. This outcome is not surprising. These

species are slow growing, large-sized, have low population

turnover rates (Zabala et al., 1997; La Mesa et al., 2002; Bodilis

et al., 2003), and are heavily targeted by fishermen in the

Canaries (Bas et al., 1995; Falcon et al., 1996). Both species

are hence highly vulnerable to overfishing. Similar results

have been obtained elsewhere, such as in the Mediterranean

Sea, by comparing fish populations between areas protected

from fishing and areas open to fishing (La Mesa et al., 2002;

Garcıa-Charton et al., 2004; and references therein). Addition-

ally, the distribution patterns for these two groupers support

our second hypothesis: within the serranid family, larger-bod-

ied species are more vulnerable to human overexploitation

than smaller congeners.

Both comber species (S. scriba and S. atricauda) are med-

ium-sized fish which are of lower commercial interest and

have higher growth rates (Froese and Pauly, 2005) than the

groupers E. marginatus and M. fusca. As expected, the effects

of fishing pressure on these two species are less notable than

the effects on groupers. Indeed, the painted comber (S. scriba)

Table 3 – Significant correlation coefficients between the selected indices of human and fishing pressure and the meanabundance and biomass of the four serranid species for the entire study

Indices Ephinephelus marginatus Mycteroperca fusca Serranus scriba Serranus atricauda

Abundance Biomass Abundance Biomass Abundance Biomass Abundance Biomass

Number of fishing ships per

coastal perimeter (official)

– – – – – – – –

Number of fishing ships per

coastal perimeter (observed)

– – – – – – – –

Average number of official

ships� number of observed ships

�0.38* – �0.37* 0.38* – – – –

Number of islanders per coastal

perimeter

�0.40* – – – – – – –

Number of tourists per coastal

perimeter

– – �0.54** �0.51** – – �0.44* �0.48**

* p < 0.05.

** p < 0.01.

B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4 19

showed a clear geographical distribution pattern, similar to

those reported in other studies (Falcon et al., 1996; Brito

et al., 2003), with a clear lack of differences among islands

subjected to a gradient of human pressure. In contrast, the

blacktail comber, S. atricauda, did not show the lowest abun-

dance and biomass in the most heavily populated islands,

as we expected. It is possible that other factors beyond the

scope of this work (see La Mesa et al., 2002 for a discussion),

such as habitat availability, recruitment and settlement, are

more important than inter-island differences in human pres-

sure in determining the patterns of distribution of these two

species across the Canarian Archipelago. The possibility of

competition release/relaxation by the removal of the top pre-

dators (groupers) as a result of overfishing could be also

responsible for this pattern.

Existing knowledge of fishing effects on littoral resources

has traditionally been gained by two main approaches: tem-

poral comparisons of fish populations in an area subject to

changing rates of exploitation (e.g. before and after an area

is protected from fishing), and spatial comparisons in areas

subjected to different levels of fishing intensity (Jennings

et al., 1995; McClanahan et al., 1999; Friedlander and Brown,

2004; Hawkins and Roberts, 2004). The few studies that have

dealt with the ecology of coastal fish populations in the

Canary Islands (Bortone et al., 1991; Falcon et al., 1996) were

made within the last two decades, long after the enormous

changes in the exploitation of marine resources of the

Canarian Archipelago had occurred. Consequently, there is

no before-impact baseline data prior to the period of rela-

tively high fishing pressure to compare our results with.

We therefore adopted the latter approach to assess the role

of human pressure in the distribution, abundance and bio-

mass of serranids among the various islands. Contrasting

islands and locations spanning a gradient in which human

pressure has differed for a long time can potentially offer

new insights into the effects of human pressure (Jennings

et al., 1995; Friedlander and Brown, 2004; Hawkins and

Roberts, 2004). However, the drawback of this approach is

that islands may differ in respects other than human pres-

sure (e.g. habitat structure). Hence, caution is necessary in

ascribing differences in the observed fish populations to hu-

man exploitation. In addition, we have used a correlational

approach to infer that human pressure is the underlying

causal factor. It should be borne in mind that causality

can only be determined through experimental manipulation

(Dulvy et al., 2004).

Coastal fisheries in the Canarian Archipelago have been

facing overexploitation and severe depletion of fish popula-

tions over the last few decades (Aguilera et al., 1994; Bas

et al., 1995; Falcon et al., 1996). The decline in abundance

and biomass of large-sized serranids, particularly around

the more populated islands (Bortone et al., 1991), is likely

the cumulative result of years of chronic overfishing. In this

context, groupers are more abundant around islands where

fishing activities are limited, there is some form of manage-

ment and fishing limitations, and human population is not

so high. El Hierro Island and Chinijo Archipelago are the

two clearest examples of relatively low fishing and human

pressure. Thus, the majority of specimens of the dusky-

grouper, E. marginatus, were observed at these areas. This

species is rare at depths accessible by SCUBA diving at the

rest of the islands, but was frequently observed at El Hierro

Island and more rarely recorded at the Chinijo Archipelago,

giving additional support for our first stated hypothesis. This

observation fits with several contemporary examples from

around the world that have demonstrated how some of the

largest fish species appeared to have been extirpated by hu-

mans in some of the most heavily perturbed islands of oce-

anic archipelagos, such as the Caribbean (Hawkins and

Roberts, 2004), the Indo-Pacific (Dulvy et al., 2004), the Hawai-

ian Archipelago (Friedlander and De Martini, 2002; Fried-

lander and Brown, 2004) and the Seychelles (Jennings et al.,

1995). Moreover, our results are similar with those reported

from the Mediterranean Sea, where the effects of protection

from fishing in certain areas of the coast have lead to incre-

ments in dusky grouper abundance and biomass (Zabala

et al., 1997; La Mesa et al., 2002; Garcıa-Charton et al., 2004,

and references therein).

With regard to our third hypothesis, results of our work have

highlighted the relatively small role of time in determining

Fig. 3 – Mycteroperca fusca. Mean abundance and biomass throughout the study. Error bars are standard errors of mean

values. HI: Hierro, LP: La Palma, GO: Gomera, TF: Tenerife, GC: Gran Canaria, FV: Fuerteventura, LZ: Lanzarote, CH: Chinijo

Archipelago.

20 B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

the abundance and biomass patterns of the observed fish

populations. In other words, differences among islands for

the four serranid populations were generally persistent

through time. Falcon et al. (1996) suggested that the impact

of the very high fishing pressure on the fish population

dynamics along the Canary Islands may exceed the effects

that other more natural factors related to the season of the

year, such as food availability, competition or productivity.

Further, coastal sea water temperatures only range from 18

to 24 �C in the Canary Islands, in contrast for example, to a

Fig. 4 – Serranus scriba. Mean abundance and biomass throughout the study. Error bars are standard errors of mean values. HI:

Hierro, LP: La Palma, GO: Gomera, TF: Tenerife, GC: Gran Canaria, FV: Fuerteventura, LZ: Lanzarote, CH: Chinijo Archipelago.

B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4 21

variation from 12 to 25 �C in the Mediterranean Sea (Francour

et al., 1994).

Intensive human pressure on vulnerable fish species raises

concerns about the long-term sustainability of these popula-

tions. Consequently, we urge national and regional authorities

to establish management measures, backed by a solid and

effective legal framework, to better control inshore fishery re-

sources in the Canarian Archipelago. In this sense, successful

Fig. 5 – Serranus atriacuda. Mean abundance and biomass throughout the study. Error bars are standard errors of mean

values. HI: Hierro, LP: La Palma, GO: Gomera, TF: Tenerife, GC: Gran Canaria, FV: Fuerteventura, LZ: Lanzarote, CH: Chinijo

Archipelago.

22 B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

enhancement of adult spawning populations by establishing a

network of marine reserves and areaswith fishing restrictions

is highly encouraged.

Acknowledgements

Research was economically supported by the Spanish ‘‘Minis-

terio de Medio Ambiente’’ in the framework of the ‘‘Canarias,

por una costa viva’’ project (www.canariasporunacosta-

viva.org). We gratefully thank A. Boyra, L. Ortega, N. Monte-

sdeoca, I. Blanch, C. Garcia, E. Falcon, P. Gonzalez-Navarro, O.

Bergasa, T. Sanchez, O. Tavio, A. Iglesias, A. Lopez, N. Rodriguez,

A. Del Rosario, F. Del Rosario, G. Herrera, R. Herrera and F. Espino

for helping us with the underwater data collection and process-

ing. Special thanks go to F. Espino for ideas, comments and

information about human pressure on the study area.

B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4 23

R E F E R E N C E S

Aguilera, F., Brito, A., Castilla, C., Dıaz, A., Fdez-Palacios, J.M.,Rodrıguez, A., Sabate, F., Sanchez, J., 1994. Canarias: Economı´ a,Ecologıa y Medio Ambiente. Francisco Lemus, Las Palmas, IslasCanarias.

Anderson, M.J., Millar, R.B, 2004. Spatial variation and effects ofhabitat on temperate reef assemblages in north eastern NewZealand. Journal of Experimental Marine Biology and Ecology305, 191–221.

Bas, C., Castro, J.J., Hernandez-Garcıa, V., Lorenzo, J.M., Moreno, T.,Pajuelo, J.G., Gonzalez-Ramos, A.J., 1995. La Pesca en Canariasy areas de influencia. Ediciones del Cabildo Insular de GranCanaria, Las Palmas, Islas Canarias.

Beger, M., Jones, G.P., Munday, P.L., 2003. Conservation of coral reefbiodiversity: a comparison of reserve selection procedures forcorals and fishes. Biological Conservation 111, 53–62.

Bodilis, P., Ganteaume, A., Francour, P., 2003. Recruitment of thedusky grouper Epinephelus marginatus in the North-WesternMediterranean Sea. Cybium 27, 123–129.

Bohnsack, J.A., Harper, D.E., McClellan, D.B., 1994. Fisheries trendsfrom Monroe County, Florida. Bulletin of Marine Science 54,982–1018.

Bortone, S.A., Van Tasell, J., Brito, A., Falcon, J.M., Bundrick, C.M.,1991. A visual assessment of the inshore fishes and fisheryresources off El Hierro, Canary Islands: a baseline survey.Scientia Marina 55, 529–541.

Brito, A., Pascual, P.J., Falcon, J.M., Sancho, A., Gonzalez, G., 2002.Peces de las Islas Canarias. Catalogo comentado e ilustrado,Francisco Lemus, Islas Canarias.

Brito, A., Falcon, J., Aguilar, N., Pascual, P., 2003. Fauna vertebradamarina. In: Fernandez-Palacios, J.M., Martın-Esquivel, J.L. (Ed.),Naturaleza de las Islas Canarias. Editorial Turquesa, Tenerife,Islas Canarias.

Brock, R.E., 1982. A critique of the visual census method forassessing coral reef fish populations. Bulletin of MarineScience 32, 269–276.

Carter, J., Marrow, G.J., Prior, V., 1994. Aspects of the ecology andreproduction of Nassau grouper, Ephinephelus striatus, off thecoast of Belize, Central America. Proceedings of the GulfCaribbean Fish Institute 43, 65–111.

Chiappone, M., Sluka, R., Sullivan Sealey, K., 2000. Groupers(Pisces: Serranidae) in fished and protected areas on theFlorida Keys, Bahamas and northern Caribbean. MarineEcology Progress Series 198, 261–272.

Denny, C.M., Babcock, R.C., 2004. Do partial marine reservesprotect reef fish assemblages?. Biological Conservation 116,119–129.

Dulvy, N.K., Freckleton, R.P., Polunin, N.V.C., 2004. Coral reefcascades and the indirect effects of predator removal byexploitation. Ecology Letters 7, 410–416.

Falcon, J.M., Bortone, S.A., Brito, A., Bundrick, C.M., 1996.Structure and relationships within and between the littoralrock-substrate fish communities off four islands in theCanarian Archipelago. Marine Biology 125, 215–231.

Ferreira, C.E.L., Floeter, S.R., Gasparini, J.L., Ferreira, B.P., Joyeux,J.C., 2004. Trophic structure patterns of Brazilian reef fishes: alatitudinal comparison. Journal of Biogeography 31, 1093–1106.

Francour, P., Boudouresque, C.F., Harmelin, J.G., Harmelin-Vivien,M.L., Quignard, J.P., 1994. Are the Mediterranean watersbecoming warmer? Information from biological indicators.Marine Pollution Bulletin 28, 523–526.

Friedlander, A.M., Brown, E., 2004. Marine protected areas andcommunity-based fisheries management in Hawaii. In:Friedlander, A.M. (Ed.), Hawaii Coastal Fisheries in the NewMillennium, Proceedings of the 2001 Fisheries Symposium ofthe American Fisheries Society, Hawaii Chapter, pp. 208–230.

Friedlander, A.M., De Martini, E.E., 2002. Contrast in density, size,and biomass of reef fishes between the northwestern and themain Hawaiian Islands: the effects of fishing down apexpredators. Marine Ecology Progress Series 230, 253–264.

Froese, R., Pauly, D., 2005. FishBase. World Wide Web electronicpublication. Available from: www.fishbase.org, version(02/2005).

Garcıa-Charton, J.A., Perez-Ruzafa, A., 1999. Ecologicalheterogeneity and the evaluation of the effects of marinereserves. Fisheries Research 42, 1–20.

Garcıa-Charton, J.A., Perez-Ruzafa, A., Sanchez-Jerez, P.,Bayle-Sempere, J.T., Renones, O., Moreno, D., 2004. Multi-scalespatial heterogeneity, habitat structure, and the effect ofmarine reserves on Western Mediterranean rocky reef fishassemblages. Marine Biology 144, 161–182.

Gobert, B., 1994. Preliminary analysis of the exploitation ofgroupers in Martinique. Proceedings of the Gulf Caribbean FishInstitute 43, 446–455.

Harmelin-Vivien, M.L., Harmelin–Vivien, J.G., Chauvet, C., Duval,C., Galzin, R., Lejeune, P., Barnabe, G., Blanc, F., Chevalier, J.,Duclerc, J., Lasserre, G., 1985. Evaluation visuelle despeuplements et populations de poissons: methodes etproblemes. Terre Vie 40, 467–539.

Hawkins, J.E., Roberts, C.M., 2003. Effects of fishing onsex-changing Caribbean parrotfishes. Biological Conservation115, 213–226.

Hawkins, J.E., Roberts, C.M., 2004. Effects of artisanal fishing onCaribbean coral reefs. Conservation Biology 18,215–226.

Hawkins, J.P., Roberts, C.M., Van�T Hof, T., De Meyer, K., Tratalos, J.,Aldam, C., 1999. Effects of recreational scuba diving onCaribbean coral and fish communities. Conservation Biology13, 888–902.

Huntsman, G.R., Potts, J., Mays, R.W., 1994. A preliminaryassessment of the populations of seven species of grouper(Serranidae, Ephinephelidae) in the western Atlantic Oceanfrom Cape Hatteras, North Carolina to Dry Tortugas, Florida.Proceedings of the Gulf Caribbean Fish Institute 43, 193–213.

Jackson, J.B.C., 2001. What was natural in the coastal oceans?.Proceedings of the National Academy of Sciences 98,5411–5418.

Jennings, S., Grandcourt, E.M., Polunin, N.V.C., 1995. The effects offishing on the diversity, biomass and trophic structure ofSeychelles� reef fish communities. Coral Reefs 14, 225–235.

Kingsford, M., Battershill, C., 1998. Studying Marine TemperateEnvironments: A Handbook for Ecologists. CanterburyUniversity Press, Christchurch, New Zealand.

La Mesa, G., Louisy, P., Vacchi, M., 2002. Assessment ofmicrohabitat preferences in juvenile dusky grouper(Ephinephelus marginatus) by visual sampling. Marine Biology,175–185.

Lincoln-Smith, M.P., 1988. Effects of observer swimming speed onsample counts of temperate rocky reef fish assemblages.Marine Ecology Progress Series 43, 223–231.

Lincoln-Smith, M.P., 1989. Improving multispecies rocky reef fishcensuses by counting different groupsof species using differentprocedures. Environmental Biology of Fish 26, 29–37.

McClanahan, T.R., Muthiga, N.A., Kamukuru, A.T., Machano, H.,Kiambo, R.W., 1999. The effects of marine parks and fishing oncoral reefs of northern Tanzania. Biological Conservation 89,161–182.

Miller, M.W., Gerstner, C.L., 2002. Reefs of an uninhabitedCaribbean island: fishes, benthic habitat, and opportunities todiscern reef fishery impact. Biological Conservation 106,37–44.

Pitcher, T.J., 2001. Fisheries managed to rebuild ecosystems?Reconstructing the past to salvage the future. EcologicalApplications 11, 601–617.

24 B I O L O G I C A L C O N S E R VAT I O N 1 2 8 ( 2 0 0 6 ) 1 3 –2 4

Sadovy, Y., 1994. Grouper stocks on the western centralAtlantic: the need for management and management needs.Proceedings of the Gulf Caribbean Fish Institute 43,43–64.

Tuya, F., Boyra, A., Sanchez-Jerez, P., Barbera, C., Haroun, R.J., 2004.Relationships among fishes, the long-spined sea urchinDiadema antillarum and algae throughout the CanarianArchipelago. Marine Ecology Progress Series 278, 157–169.

Underwood, A.J., 1997. Experiments in Ecology: their LogicalDesign and Interpretation Using Analysis of Variance.Cambridge University Press, Cambridge.

Zabala, M., Garcıa-Rubies, A., Louisy, P., Sala, E., 1997. Spawningbehaviour of the Mediterranean dusky grouper Ephinephelusmarginatus (Lowe, 1834) (Pisces, Serranidae) in the MedesIslands Marine Reserve (NW Mediterranean, Spain). ScientiaMarina 61, 65–77.