Mar Biol (2007) 151:1849–1861

DOI 10.1007/s00227-007-0623-2

RESEARCH ARTICLE

Spawning patterns in the leopard grouper, Mycteroperca rosacea, in comparison with other aggregating groupers

Brad E. Erisman · Michele L. Buckhorn · Philip A. Hastings

Received: 12 May 2006 / Accepted: 12 January 2007 / Published online: 6 March 2007© Springer-Verlag 2007

Abstract We documented the spawning patterns ofthe leopard grouper, Mycteroperca rosacea, from Aprilto June 2005 in the central Gulf of California, Mexicoto draw comparisons with other aggregate-spawninggroupers and provide information useful for manage-ment of their Wshery. Adults formed spawning aggrega-tions of 150 to >700 individuals at speciWc sites, andspawning occurred daily at these sites from late Aprilthrough early June. Courtship occurred throughout theday, but spawning was restricted to the evening hours.Adults spawned in groups of 6–40 Wsh, and pair-spawn-ing was not observed. The group-spawning behavior ofadults and the gonosomatic indices of mature males(maximum = 7.2%) suggest that sperm competitionwas present. The site-speciWcity of leopard grouperspawning aggregations and diel spawning period weretypical of most aggregating groupers, and the size andstructure of these aggregations was similar to otherspecies in the genus Mycteroperca. Leopard grouperbehavior patterns were unusual in that spawning aggre-gations persisted for extended periods, spawning wasnot synchronized with the lunar cycle, and adultsaggregated during non-spawning periods. The extensive

duration and site-speciWcity of spawning aggregationsand the propensity of M. rosacea to form aggregationsyear-round increases the vulnerability of the species tooverWshing. Policies that limit harvest from theseaggregations are needed to improve the managementof leopard grouper Wsheries in the Gulf of California.

Introduction

Epinepheline serranid Wshes (groupers) exhibit aremarkable diversity of reproductive patterns (Sadovy1996); however, several trends have been identiWedamong species that form spawning aggregations(reviewed by Sadovy et al. 1994a; Domeier and Colin1997). Fish migrate considerable distances and formspawning aggregations at speciWc sites (Aguilar-Pereraand Aguilar-Dávila 1996; Zeller 1998). All or mostspawning activity happens within these aggregations,where it happens in the evening near dusk and is syn-chronized with the lunar cycle (Smith 1972; Thresher1984; Colin 1992). Spawning aggregations are seasonaland persist only for several days to 2 weeks before theWsh disperse (Shapiro 1987; Levin and Grimes 2002;Rhodes and Sadovy 2002). During non-reproductiveperiods, Wsh are usually solitary and do not aggregate(Sadovy and Eklund 1999; Domeier and Colin 1997).Exceptions to these trends are rare but do exist. Forinstance, Epinephelus guttatus lives in overlappinghome ranges during non-reproductive periods andspawning is not lunar-related near its northern rangelimit (Burnett-Herkes 1975; Sadovy et al. 1994a). Allmedium and large-sized groupers (maximum TL of¸50 cm) aggregate to spawn, but this behavior isuncommon among smaller species (Sadovy 1996).

Communicated by P.W. Sammarco.

B. E. Erisman (&) · P. A. HastingsMarine Biology Research Division, Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA 92093-0208, USAe-mail: berisman@ucsd.edu

M. L. BuckhornDepartment of Environmental Science and Policy, University of California, Davis, CA 95616, USA

123

1850 Mar Biol (2007) 151:1849–1861

Other characteristics of aggregating groupers aremore variable. Spawning aggregations of some speciesare patchily distributed over broad physical areas,whereas others are highly localized and discrete (Colin1992; Coleman et al. 1996). Mating systems also diVeramong groupers. Spawning occurs in pairs within singlemale, multiple female clusters in E. guttatus and severalCephalopholis species, and this behavior coupled withthe small size of ripe testes (·2% of body wt.) suggeststhat sperm competition is absent (Sadovy et al. 1994a;Donald 1995; Chan and Sadovy 2002). Conversely, E.striatus spawns in the aggregates of female led, multi-ple male groups, and sperm competition is evidentfrom this behavior and the relatively large size of testes(¸10% of body wt: Sadovy and Colin 1995).

The leopard grouper, Mycteroperca rosacea, is oneof several groupers that form spawning aggregations inthe Gulf of California, Mexico. There is little publisheddata on their reproductive biology, but existing studiesindicate that M. rosacea behavior patterns are diVerentfrom that of other aggregating groupers. Sala et al.(2003) observed courtship during both new and fullmoons within aggregations of hundreds of individualsand theorized that temporal spawning patterns of thisspecies may be unrelated to the lunar cycle. Also, leop-ard groupers are social during non-reproductive peri-ods and aggregate in considerable numbers to feed onschools of small Wshes and mysids (Hobson 1965; Par-rish 1992: Sala et al. 2003). Finally, leopard groupersare gonochoric (Erisman et al. 2007); individuals func-tion as either males or females, and post-maturationalsex change does not occur. Gonochorism has beenreported only in one other grouper species (Epinephe-lus striatus: Sadovy and Colin 1995). For these reasons,the behavior and spawning patterns of leopard group-ers warrant further research and may provide a betterunderstanding of the diversity of reproductive patternsamong groupers and related Wshes.

The leopard grouper is also of interest because of itsecological and economic importance. It is the mostabundant grouper of the central and lower regions ofthe Gulf and the most heavily targeted grouper bycommercial, recreational, and artisanal Wsheries (Ram-írez and Rodríguez 1990; Sala et al. 2003). IncreasedWshing eVorts on leopard groupers over the past severaldecades are correlated with signiWcant declines inannual landings and sizes of landed Wsh (Sala et al.2004), and these declines are attributed to intense Wsh-ing pressure on both spawning and feeding aggrega-tions. Consequently, the World Conservation Union(IUCN) currently lists M. rosacea as a “Threatened”species (http://www.iucnredlist.org). Information onthe spawning patterns of leopard groupers is needed to

determine which management policies are necessary topromote population recovery and create a more sus-tainable Wshery.

The purpose of this study was to describe the repro-ductive patterns of M. rosacea in comparison withother aggregate-spawning groupers, including detailson: (1) the mating system, (2) sperm competition, (3)site-speciWcity and duration of spawning aggregations,(4) and spawning periodicity.

Materials and methods

Field observations

To document the behavior patterns of adult M. rosacea,observations using SCUBA were conducted daily at 24sites near Loreto, Baja California Sur, Mexico(26°00�N, 111°21�W) from 12 April to 12 June 2005(Fig. 1). Observations of spawning behavior weremade primarily at two sites, Punta Lobos (20–23 April,6–9 May, and 23–26 May 2005) and Punta Coyote(weekly from 18 April to 13 May and 28 May to 12June, daily from 14 to 22 May 2005). These sites werechosen because they are known spawning aggregationsites for M. rosacea and are heavily exploited by com-mercial and recreational Wshers (R. Murillo and F.Arcas, personal communication). Observations wereconducted at 22 additional sites (e.g. Isla Danzante,Isla Coronado) to compare behaviors and abundancesof adult leopard groupers between aggregation andnon-aggregation sites.

Dives were made throughout the day, but most werecarried out during the late afternoon and evening

Fig. 1 Map of Loreto, BCS, Mexico, showing dive sites (Wlled cir-cle) and location on the Baja peninsula. The majority of diveswere conducted at Punta Lobos and Punta Coyote

123

Mar Biol (2007) 151:1849–1861 1851

(1,600–2,100 h). Dives ranged from 30 to 120 min induration and 3 to 30 m in depth. Visibility Xuctuatedbetween 3–15 m and water temperatures varied from16 to 23.5°C. Aggregation sizes were estimated at eachsite by counting the number of adults present whileswimming parallel to the reef. Consistently poor visi-bility and strong currents precluded the use of transectsor other quantitative methods for estimating aggre-gation size and Wsh density. Observations focused ondocumenting details of courtship, spawning, andaggressive behaviors. Data were recorded with pencilon plastic slates or by underwater digital video cam-eras. Frame by frame analyses of video were used toanalyze behaviors in greater detail. The sex of individ-uals was determined by a combination of behaviors,color patterns, and body shape, and veriWed by histo-logical analyses of Wsh collected by BE during divesmade at Punta Lobos and Punta Coyote (see below).

Terminology and descriptions of behaviors followedthose reported for the serranid Paralabrax clathratus(Erisman and Allen 2006). Behaviors included: follow-ing, courtship chase, mobbing, female darting, spawn-ing rush, rubbing, bumping, lateral display, gaping,nipping, dorsal Xare, aggressive chase, and spiraling.Other behaviors were described previously by Salaet al. (2003) for M. rosacea, and they included: headshake = a male shook the anterior portion of his bodyand head repeatedly upon approach to a female, and:burst rise = a male swam rapidly from the reef in anupward vertical direction for several m and then swamrapidly back down to the reef.

Collection and preparation of samples

Standard length (to the nearest mm) and weight (to thenearest g) data were collected daily from leopardgroupers caught at or near Punta Lobos by recreationalWshers (hook and line; n = 203) from 14 April to 10June 2005. All Wsh were sampled in the early afternoon(1,200–1,400 h) on the day of capture. Additional spec-imens were collected by speargun at Punta Lobos(n = 39; 22–23 April, 6–9 May, 23–25 May 2005) andPunta Coyote (n = 6; 16–17 May 2005). Sampling wasopportunistic, as data were gathered for all M. rosaceaprovided by Wshers or collected by BE. Gonads wereremoved from the body cavity and preserved in a 10%formalin-seawater solution. The gonads of 75 femalesand 100 males were weighed to the nearest 0.1 g priorto preservation, and gonosomatic indices [GSI =(gonad weight/body weight) £ 100] were calculated forthese Wsh. The relationship between gonad weight,GSI, and standard length (SL) were analyzed by Spear-man rank-correlation.

The sex and developmental class of all collectedindividuals (n = 248) were assessed by macroscopic andmicroscopic evaluations of gonads (Table 1). Criteriaused in the classiWcation of gonad class were adaptedfrom Moe (1969), Sadovy et al. (1994a), and Rhodesand Sadovy (2002). Stages of gametogenesis followedthe deWnitions of Moe (1969) and Nagahama (1983).Standard histological techniques were used for micro-scopic evaluations. Pieces of gonad tissue were embed-ded in paraYn, sectioned at 6 �m, mounted, stainedwith Mayer’s hemotoxylin-eosin, and observed under acompound microscope (Humason 1972). Tissue wastaken from the central portion of one lobe and sec-tioned transversely in larger gonads, while smallergonads were sectioned longitudinally in their entirety.Immature individuals were collected (see Tables 1, 2)but were excluded from all subsequent analyses.

Mating system and sperm competition

Behavioral observations were used to describe themating system (i.e., group vs. pair-spawning) of M. ros-acea. The number of Wsh that released gametes duringspawning events could not be determined from obser-vations. Therefore, the intensity of sperm competitionwas inferred by the number of male participants inobserved spawns and the GSI of mature males (M2—M4; n = 100) from sample collections. In broadcast-spawning Wshes, GSI is indicative of male investment ingamete production, which in turn is correlated with theintensity of sperm competition (Warner 1997; Petersenand Warner 1998; Taborsky 1998).

Spawning periodicity

Methods used to estimate spawning periodicity anddetermine if spawning was synchronized with the lunarcycle included observations of spawning behavior inthe Weld and temporal changes in the proportion ofspawning females. Collected females with evidence ofimminent or recent spawning (F4–F6) were classiWedas spawning females, and these data were organizedweekly by lunar phase (Wrst quarter, full, last quarter,new) and pooled by lunar phase only. The relationshipbetween lunar phase and the proportion of spawningfemales (vs. non-spawning, mature) was tested using aPearson �2 goodness of Wt test. Collected females withovaries that contained both hydrated oocytes andpost-ovulatory follicles (F5) were used as evidence ofdaily spawning (Hunter et al. 1985; Hunter et al.1986).

Analyses of daily and weekly Xuctuations in GSI ofmature females (F2–F6; n = 75) provided additional

123

1852 Mar Biol (2007) 151:1849–1861

Tab

le1

Mac

rosc

opic

and

mic

rosc

opic

des

crip

tion

s of

gon

adal

dev

elop

men

t cla

sses

in m

ale

and

fem

ale

leop

ard

grou

per,

Myc

tero

perc

a ro

sace

a

Gon

adal

dev

elop

men

t cla

ssM

acro

scop

icM

icro

scop

ic

Ova

ries

F1

(im

mat

ure)

Gon

ad s

mal

l and

thin

; cle

ar, p

ink

or c

ream

in c

olor

: oo

cyte

s in

disc

erni

ble;

indi

stin

guis

habl

e fr

om M

1 m

ales

Thi

n go

nad

wal

l; ti

ghtl

y pa

cked

pri

mar

y gr

owth

sta

te

oocy

tes

(chr

omat

in n

ucle

olar

, per

inuc

leol

ar)

and

oogo

nia;

no

evi

denc

e of

pre

viou

s sp

awni

ngF

2 (m

atur

e, in

acti

ve)

Gon

ad s

mal

l but

rou

nded

; gra

y in

col

or w

ith

thic

kene

d go

nad

wal

l; oo

cyte

s sm

all o

r in

disc

erni

ble;

oft

en

indi

stin

guis

habl

e fr

om M

1 m

ales

Thi

ck g

onad

wal

l; pr

imar

y gr

owth

sta

ge a

nd c

orti

cal a

lveo

lar

oocy

tes

abun

dant

; pre

viou

s sp

awni

ng in

dica

ted

by t

he p

rese

nce

of d

isti

nct

mus

cle

bund

les

or v

itel

loge

nic

oocy

tes

in la

te s

tage

s of

atr

esia

F3

(mat

ure,

act

ive)

Gon

ad la

rge;

cre

am in

col

or w

ith

tran

spar

ent

gona

d w

all;

larg

e vi

tello

geni

c oo

cyte

s vi

sibl

e an

d ti

ghtl

y pa

cked

thr

ough

out o

vary

Thi

n go

nad

wal

l; la

rge

vite

lloge

nic

oocy

tes

abun

dant

; pri

mar

y gr

owth

st

age

oocy

tes

pres

ent

in v

aria

ble

abun

danc

e; h

ydra

ted

oocy

tes

and

post

-ovu

lato

ry fo

llicl

es a

bsen

t; n

o la

rge

scal

e at

resi

aF

4 (m

atur

e, r

ipe)

Gon

ad la

rge

wit

h cl

ear,

hyd

rate

d oo

cyte

s; e

ggs

rele

ased

w

ith

appl

icat

ion

of li

ght a

bdom

inal

pre

ssur

eA

ll oo

cyte

dev

elop

men

tal s

tage

s pr

esen

t w

ith

the

exce

ptio

n of

pos

t-ov

ulat

ory

folli

cles

; hyd

rate

d oo

cyte

s fe

w to

abu

ndan

t; at

reti

c vi

tello

geni

c oo

cyte

s m

ay b

e pr

esen

tF

5 (m

atur

e, r

ipe

and

post

-spa

wn)

Gon

ad r

esem

bles

F4

or F

6 fe

mal

esA

ll oo

cyte

s de

velo

pmen

tal s

tage

s pr

esen

t, in

clud

ing

both

hyd

rate

d oo

ctye

s an

d po

st-o

vula

tory

fol

licle

sF

6 (p

ost-

spaw

n)G

onad

Xac

cid

wit

h vi

sibl

e ca

pilla

ries

; vis

ible

ooc

tyes

less

ab

unda

nt t

han

in F

3 or

F4

fem

ales

; hyd

rate

d oo

cyte

s ab

sent

Thi

ck g

onad

wal

l; m

ost

oocy

te d

evel

opm

enta

l sta

ges

pres

ent w

ith

the

exce

ptio

n of

hyd

rate

d oo

cyte

s; p

ost-

ovul

ator

y fo

llicl

es a

nd m

uscl

e bu

ndle

s pr

esen

t; at

reti

c vi

tello

geni

c oo

cyte

s pr

esen

t and

oft

en a

bund

ant;

in

tral

amel

lar

debr

is a

nd c

ytop

lasm

ic s

tran

ds p

rese

ntT

este

s

M1

(im

mat

ure

or in

acti

ve)

Gon

ad s

mal

l and

thin

; ind

isti

ngui

shab

le fr

om F

1 an

d F

2 fe

mal

esSt

rom

a pr

esen

t in

var

iabl

e am

ount

s; e

arly

pro

lifer

atio

n of

spe

rmat

ocyt

es

evid

ent;

spe

rm s

inus

es a

nd s

emin

ifer

ous

lobu

les

may

be

pres

ent;

pr

evit

ello

geni

c oo

cyte

s m

ay b

e pr

esen

t an

d va

riab

le in

abu

ndan

ceM

2 (m

atur

e, a

ctiv

e)G

onad

larg

er a

nd m

ore

roun

ded

than

M1

mal

es;

gray

ish

in a

ppea

ranc

e; s

ome

milt

may

be

visi

ble

Mos

t st

ages

of s

perm

dev

elop

men

t pr

esen

t; s

perm

atoz

oa a

bund

ant

wit

hin

sem

inif

erou

s lo

bule

s an

d sp

erm

sin

uses

M3

(mat

ure,

rip

e)G

onad

larg

e an

d w

hite

cop

ious

am

ount

s of

vis

ible

milt

; m

ilt r

elea

se p

ossi

ble

wit

h lig

ht a

bdom

inal

pre

ssur

eSe

min

ifer

ous

lobu

les

enla

rged

, coa

lesc

ed a

nd W

lled

wit

h m

atur

e sp

erm

atoz

oa; s

perm

sin

uses

larg

e an

d fu

ll of

mat

ure

sper

mat

ozoa

; ea

rly

stag

es o

f sp

erm

atog

enes

is r

are

or a

bsen

tM

4 (p

ost-

spaw

n)G

onad

Xac

cid

wit

h vi

sibl

e bl

ood;

milt

rel

ease

m

ay s

till

occu

r w

ith

abdo

min

al p

ress

ure

Thi

ck g

onad

wal

l; se

min

ifer

ous

lobu

les

are

sepa

rate

d an

d sm

alle

r th

an

in M

3 m

ales

; lob

ules

and

spe

rm s

inus

es m

ostl

y em

pty

and

cont

ain

very

lit

tle

or n

o m

atur

e sp

erm

atoz

oa; s

trom

al ti

ssue

and

spe

rmat

ogon

ia p

rese

nt

123

Mar Biol (2007) 151:1849–1861 1853

evidence of spawning periodicity. Variations in thedaily mean GSI of females were examined by a Lomb-Scargle periodogram (Press et al. 2002). The averageweekly GSI of females grouped by lunar phase wasalso calculated, and diVerences were tested using theKruskal–Wallis test. Pairwise comparisons of weeklyGSI were analyzed by Dunn’s Multiple ComparisonTest.

Results

Spawning aggregations

Spawning aggregations were identiWed at Punta Lobosand Punta Coyote (Fig. 1). The aggregation at PuntaLobos contained 600–700 adult leopard groupers scat-tered over a distance of several hundred meters alongthe northwest side of the point. These Wsh were abun-dant from the bottom to the surface and from the reef-sand interface (20 m) to the subtidal zone (1–2 m). AtPunta Coyote, the aggregation consisted of 150–200Wsh spread out along the southernmost portion of thepoint. Individuals were most common in the openwater at depths of 8–15 m, and 2–10 m oVshore of thereef. Adult leopard groupers were absent or present inlow numbers (<15 Wsh) at all other sites visited fromApril through June 2005. Spawning was not observedoutside these spawning aggregation sites, althoughputative courtship was observed once at Isla Coronadoon 13 May 2005 between a female and two males.

IdentiWcation of sexes and color patterns

Females and males diVered in their behavior and bodyshape. Females were less active than males and manyhad enlarged abdomens full of ripe eggs. Gravidfemales with enlarged abdomens also led spawninggroups and spawning rushes. When not engaged incourtship with males, females rested on or just abovethe reef, or they hovered in the water column. Maleslacked the enlarged abdomens of spawning females, as

all Wsh collected by BE with this feature (n = 10) werefemales. Males swam more vigorously and were morecurious of divers than females. This behavior was mostintense in the late afternoon and evening, when malesswam rapidly in all directions to court females, joinspawning groups, or engage in brief aggressive interac-tions with other males. Unlike females, males oftenperformed Burst Rise and Head Shake behaviors dur-ing courtship.

The color patterns of males and females wereusually similar during courtship and spawning events,but sex-related diVerences in color were observed in afew individuals. For both sexes, the posterior half andventral portions of the body were pale, the anteriorportions of the upper body and head consisted of largeblack blotches surrounded by light, conspicuous rings,and the caudal Wns were completely black (see Salaet al. 2003 for pictures of color patterns in M. rosacea).Of six Wsh with these color patterns collected by BE,four were male and two were female. However, somefemales were dark gray or black during spawningevents, and the blotches on their heads and bodieswere less distinct. Two dark colored Wsh leadingspawning groups were collected by BE, and both werefemales. Both male and female golden phase leopardgroupers (Thomson et al. 2000) were present in spawn-ing aggregations and participated in spawning events.Of seven golden phase Wsh collected by BE, three weremale and four were female.

Sample collections

Mature females (F2–F6; n = 113) ranged in size from245 to 744 mm SL and weight from 355 to 8,800 g(Fig. 2). Mature males (M2–M4; n = 124) ranged inlength from 288 to 708 mm SL and weight from 475 to6,850 g. The mean length (§SD) of females (461.76 §95.82 mm SL) and males (470.22 § 87.71 mm SL) weresimilar (unpaired t test, t = 0.709, P = 0.479). The meanweight (§SD) of females (2461.95 § 1513.30 g) andmales (2505.52 § 1281.90 g) were also similar(unpaired t test, t = 0.209, P = 0.835).

GSI of mature females (n = 75) ranged from 0.39 to11.51%, and the mean GSI (§SD) of mature femaleswas 4.37 § 2.18%. Gonad weight increased with stan-dard length in mature females (Fig. 3a, Spearmanrank-correlation, rs = 0.645, P < 0.001), but GSI didnot change with standard length (Fig. 3b, Spearmanrank-correlation, rs = 0.009, P = 0.942). GSI of maturemales (n = 100) ranged from 0.20 to 7.21%, and themean GSI of mature males was 2.34 § 1.48%. Gonadweight increased with standard length in maturemales (Fig. 4a, Spearman rank-correlation, rs = 0.391,

Table 2 Number of sampled Wsh arranged by gonadal develop-ment class and month for the leopard grouper, Mycteropercarosacea

Month Females Males Total

F1 F2 F3 F4 F5 F6 M1 M2 M3 M4

April 3 0 13 6 2 1 1 7 17 0 50May 3 2 21 24 19 9 1 6 52 16 153June 1 1 3 6 5 1 2 6 15 5 45Total 7 3 37 36 26 11 4 19 84 21 248

123

1854 Mar Biol (2007) 151:1849–1861

P < 0.001), and GSI decreased with standard length(Fig. 4b, Spearman rank-correlation, rs = ¡0.339,P < 0.001).

Mating system

Thirty-two spawning rushes were observed during thestudy period (Table 3). Spawning rushes occurred ingroups of 6–40 Wsh (mean 17.50, SD 10.86). Spawninggroups at Punta Lobos were larger than at Punta Coy-ote (unpaired t test, t = 4.398, P < 0.001). Paired, sneak,or other types of spawning were not observed at either

site. Spawning groups led by large females (>50 cm)contained more Wsh than those led by small females atPunta Lobos (Table 3, unpaired t test, t = 7.704,P < 0.001) and Punta Coyote (unpaired t test, t = 2.931,P = 0.017).

Courtship behaviors such as Rubbing, Bumping,Lateral Display, and Following occurred throughoutthe day, but behaviors associated with imminent spawning(chasing, mobbing, spawning rush) were limited to thelate afternoon and evening hours (1,645–2,020 h).Spawning began approximately two hours before sunset(1,755–1,815 h), peaked at sunset (1,950–2,010 h), and

Fig. 2 Mycteroperca rosacea. Size-frequency distributions of collected females and males arranged by 20 mm SL size classes

Fig. 3 Mycteroperca rosacea. Gonadal investment and body sizerelationships in 75 mature females. a Gonad weight (g) versusstandard length (mm). b Gonosomatic index (GSI%) versus stan-dard length (mm)

Fig. 4 Mycteroperca rosacea. Gonadal investment and body sizerelationships in 100 mature males. a Gonad weight (g) versusstandard length (mm). b Gonosomatic index (GSI%) versus stan-dard length (mm)

123

Mar Biol (2007) 151:1849–1861 1855

continued until dark when adults settled on the reef.Courtship events that ended with spawning lasted from15 s to several minutes.

Spawning groups formed when several males visitedfemales and performed various combinations of lateraldisplay, rubbing, and bumping. These behaviors led tofollowing and chasing involving one lead female andseveral males. When females leading chasing eventsdecreased swimming speed, mobbing occurred. Mob-bing was often interrupted when the lead female madean abrupt turn and darted away from the group(female darting). Chasing and then mobbing followed,as males continued to pursue the lead female. Aspawning rush began when the female slowed down,turned upwards towards the surface, and rose slowlyupwards. Males responded immediately by mobbingand forming a tight “ball” around the female with theirvents facing her. Spawning groups often increased insize at this time, as additional males (and possiblyfemales) joined. Led by the female, the entire groupperformed a spawning rush, a rapid vertical burst swimand release of gametes, then separated quickly in alldirections (Fig. 5).

Spawning rushes varied in direction, location, andbehavior. Most involved a rapid vertical ascent of thegroup for 1–4 m; however, horizontal rushes of similardistances were also observed. Gamete release usuallyoccurred 3–8 m below the surface, although spawningadults were observed breaching the surface at PuntaLobos during the evening of 25 May 2005. Groups ofsmaller adults often spawned near the surface abovethe shallow fore-reef, whereas larger adults usuallyspawned above the reef slope or oV the reef entirely.There was anecdotal evidence of assortative mating.At both sites, large males (>50 cm) courted andspawned only with females of equal or larger sizes, butsmaller males courted and spawned with all females.

Brief aggressive encounters among males wereobserved infrequently throughout the day and con-sisted of one or more of the following behaviors: lateraldisplay, dorsal Xare, gaping, nipping, bumping, aggres-

sive chase, burst rise, and spiraling. Males did notdefend discrete territories or guard females duringcourtship and spawning.

Spawning periodicity

Courtship and spawning were observed at PuntaLobos on 6–9 and 23–26 May 2005 and at Punta Coy-ote on 15–22 May, 29–30 May, and 8–9 June 2005.Seventy-three females (65% of mature females) withmicroscopic evidence of imminent or recent (F4–F6)spawning were collected daily from 18 April to 8 June2005 (Fig. 6), and 53 of these females had ovaries withvisibly hydrated oocytes (Fig. 7). Twenty-six females(23% of mature females; 35% of spawning females) with

Table 3 Abundances of Wsh in spawning groups by site and size of the lead female for the leopard grouper, Mycte-roperca rosacea

Site Lead female Observations Range (# Wsh) Mean (# Wsh) SD

Punta Lobos All females 21 8–40 21.56 11.31Large (>50 cm) 12 20–40 29.58 7.87Small (<50 cm) 9 8–15 10.78 2.68

Punta Coyote All females 11 6–20 9.82 3.31Large 6 8–20 11.83 2.48Small 5 6–10 7.40 2.51

Pooled All females 32 6–40 17.50 10.86Large 18 8–40 23.67 10.77Small 14 6–15 9.57 3.03

Fig. 5 Mycteroperca rosacea. Illustration showing the typical ori-entation of adult Wsh during a spawning rush. The lead female isdark gray with an enlarged abdomen. Surrounding individuals aremales

123

1856 Mar Biol (2007) 151:1849–1861

evidence of daily spawning (F5) were collected weeklyfrom 27 April to 7 June 2005. The monthly frequency ofF5 females increased from April to June (Table 2;April = 9%, May = 25%, June = 31%).

During the Wrst lunar quarter of April, only 20% ofmature females were actively spawning (Fig. 8). Theproportion of spawning females increased signiWcantlyfrom the Wrst lunar quarter to the full moon in April(Pearson �2 = 5.000, df = 1, P = 0.025). For the rest ofthe collection period, the weekly proportion of spawn-ing females ranged from 50 to 88%, and diVerenceswere not signiWcant (Pearson �2 = 6.751, df = 6,P = 0.344). When these data were pooled by lunarphase only, there were no signiWcant diVerences in theproportion of spawning females among lunar phases(Pearson �2 = 3.536, df = 3, P = 0.316).

Weekly mean female GSI increased steadily fromthe Wrst quarter moon in April (3.44%) to the newmoon in May (6.07%) and decreased through the Wrstquarter moon in May (4.17%, Fig. 8). A second peak inmean female GSI occurred during the last quartermoon at the end of May (5.19%) but was followed by adecline. DiVerences in weekly female GSI from Aprilto June were signiWcant (Kruskal–Wallis, H = 15.050,df = 7, P = 0.035), but all pairwise comparisons offemale GSI over this period were non-signiWcant(Dunn’s Multiple Comparison, P > 0.05). Female GSIdata pooled by lunar phase only showed no signiWcantdiVerences (Kruskal–Wallis, H = 5.159, P = 0.171).Changes in the daily mean GSI of mature females fromApril to June were not signiWcant (Lomb-Scargle peri-odogram, n = 27, � = 0.277).

Fig. 7 Mycteroperca rosacea. Macroscopic evidence of imminentspawning in the ovaries of mature females. a Ovaries dissectedfrom the body cavity of a mature, ripe female with visibly hydratedoocytes. b Close-up view of the ovary of a mature, ripe female stillintact in the body cavity but with the abdominal wall removed. Theventral portion of the ovary is Wlled with clear, ovulated oocytes,and small portions of oocytes are leaking from the urogenital pore

Fig. 6 Mycteroperca rosacea. Transverse sections of female gonads. a Mature, active (F3) female; b mature, ripe (F4) female; c mature, ripe and post-spawn (F5) female; d post-spawn (F6) female. ho hydrated oocytes; pof post-ovulatory follicles; vo vitello-genic oocytes. Scale bar 0.5 mm

123

Mar Biol (2007) 151:1849–1861 1857

Discussion

Mating system and sperm competition

Leopard groupers in the Loreto area exhibited agroup-spawning mating system, and there was no evi-dence of pair-spawning or alternative mating behaviorssuch as sneaking, territoriality, or female defense(Taborsky 1994; Gross 1996). Spawning groups con-sisted of a lead female and Wve to forty surroundingindividuals, most of which were presumed to be males.Overall, the mating system of this species is similar tothat of the Nassau grouper, E. striatus, and the kelpbass, P. clathratus (Colin 1992; Erisman and Allen2006); all three species spawn in groups within largeraggregations and are non-territorial. Conversely, M.rosacea behaved diVerently than many other aggregat-ing groupers including closely related species such asM. tigris, M. phenax, and M. microlepis, where malesare territorial and Wsh spawn in pairs (Gilmore andJones 1992; Coleman et al. 1996; Sadovy et al. 1994b;Samoilys and Squire 1994; Zabala et al. 1997).

In reef Wshes, male mating behavior often varies withlocal population density. At very high densities, thereproductive success of territorial males may be reducedsubstantially by competition with other males, becauseterritorial males must spend more time in defense andhave less time available for mating (Warner and HoV-man 1980a, b; Warner 1982, 1984). As a consequence,territorial behavior is rare in Wshes that spawn in largegroups (Warner 1982; Erisman and Allen 2006). Thebehavioral patterns of male leopard groupers were simi-lar to those of other group-spawning Wshes; aggressive

interactions among males were infrequent and short induration, and males did not guard females or defend ter-ritories from other males. For leopard groupers, occa-sional aggression may reinforce dominance hierarchiesamong rival males and determine the arrangement ofmales around a spawning female (Hutchings et al. 1999;Erisman and Allen 2006), with dominant males gainingprimary access to females during spawning. By occupy-ing the closest position to the female, dominant malesmay fertilize more eggs and realize a higher reproduc-tive success than peripheral males (Warner and Harlan1982; Petersen and Warner 1998).

The large number of male leopard groupers inspawning groups indicates that sperm competition waspresent, and GSI of mature males was consistent withpredictions of sperm competition theory. In general,testes are larger and GSI values are higher in group-spawning Wshes characterized by sperm competitionthan in pair-spawning Wshes where sperm competitionis absent or low (van den Berghe and Warner 1989;Taborsky 1994; Warner 1997). In fact, male GSI mayreach values of 7–12% in serranid Wshes that spawn ingroups (Sadovy and Colin 1995; Erisman and Allen2006). Accordingly, testes sizes and GSI of male leop-ard groupers were similar to other group-spawning ser-ranids (up to 7.2%) and considerably higher thanclosely related species that spawn in pairs. For exam-ple, Mycteroperca microlepis, M. phenax, and M. tigrisall pair-spawn, have small testes, and maximum maleGSI values range from 0.6 to 3.0% (Coleman et al.1996; Sadovy et al. 1994b; Brulé et al. 2003).

The decline in male leopard grouper GSI with bodysize is a pattern typically associated with Wshes thatexhibit multiple male reproductive strategies (Choatand Robertson 1975; Robertson and Warner 1978; Pet-ersen and Warner 1998). In these Wshes, small malesgroup or sneak spawn and overcome sperm competi-tion by investing in gamete production. Conversely,large males pair-spawn, and they invest in territorialdefense and female attraction rather than gamete pro-duction (Taborsky 1994, 1998; Warner 1997). Weobserved no diVerences in mating strategies with size inmale leopard groupers, as both small and large malesspawned in groups. However, it is possible that multi-ple mating strategies do exist, with pair-spawning andterritorial behavior by large males occurring outsideaggregations. Sadovy and Colin (1995) found small tes-tes in a ripe male E. striatus taken from outside aspawning aggregation, and they oVered a similar expla-nation. The low abundance of adult leopard groupersobserved outside spawning aggregations suggests thatterritorial mating behavior would be a successfulreproductive strategy for males at these sites.

Fig. 8 Mycteroperca rosacea. Temporal changes in spawningactivity of mature females collected from April through June2005. Bars indicate the weekly proportion of collected femaleswith evidence of imminent or recent spawning (F4–F6), and num-bers at bottom of bars indicate sample sizes. Line indicates weeklymean (§SD) gonosomatic indices (GSI%) of females, and num-bers above error bars indicate sample sizes

123

1858 Mar Biol (2007) 151:1849–1861

Spawning aggregation characteristics

Spawning aggregations are an important life historytrait for M. rosacea, because much or all of their repro-ductive activity occurs within these aggregations. Sev-eral features of leopard grouper spawning aggregationswere analogous to those of other aggregating groupers.First, aggregations were site-speciWc, as adults wereabundant only at certain locations and were rare at allother visited sites in the Loreto area. Next, there wasevidence of migration to spawning aggregations. Dur-ing non-spawning periods (summer, fall, winter), adultleopard groupers were common at most visited sites inLoreto (BE and MB, unpublished data), which sug-gests that some Wsh travel signiWcant distances to reachspawning aggregation sites. The diel spawning periodof Mycteroperca rosacea was also consistent with othergroupers in that it occurred in the evening near dusk(Thresher 1984; Samoilys and Squire 1994; Sadovy1996; Levin and Grimes 2002). The structure of leop-ard grouper aggregations was similar to that of M. phe-nax and M. microlepis, in that aggregations consisted oftens to hundreds of individuals spread out over a largearea (Coleman et al. 1996).

Mycteroperca rosacea spawning aggregations inLoreto were unusual in that they persisted forextended periods, and this is supported by behavioralobservations and data acquired from Wshers. Fish wereaggregated at Punta Lobos during all our observations,and Wshers collected large numbers of adults daily fromthat site from early April through June. Thus, theaggregation at Punta Lobos likely formed sometime inApril and remained intact for the entire spawning sea-son. Also, the aggregation at Punta Coyote persistedfor over 3 weeks and likely longer. This pattern is dis-similar to most grouper spawning aggregations, whichare typically shorter in duration (Shapiro 1987; Sadovy1996; Rhodes and Sadovy 2002).

The short duration of most grouper spawning aggre-gations coincides with spawning activity that follows alunar or semi-lunar rhythm (Thresher 1984; Samoilysand Squire 1994; Rhodes and Sadovy 2002; Lee et al.2002). Our results indicate that spawning in leopardgroupers was not lunar-related, but rather it occurreddaily from late April through early June. First, court-ship and spawning behavior were observed duringmost dives at Punta Lobos and Punta Coyote in Mayand June. Next, spawning females (F4–F6) were col-lected daily and their occurrence was not synchronizedwith the lunar cycle. In fact, 35% of spawning females,were capable of spawning on consecutive days (i.e.daily spawning in F5 females), and these females werecollected weekly throughout the study period. Finally,

female GSI patterns were also inconsistent with a lunarspawning periodicity. Fishes that spawn in concert withthe lunar cycle typically show drastic changes in femaleGSI during speciWc lunar phases when spawning activ-ity occurs (Hoque et al. 1999; Lee et al. 2002; Rhodesand Sadovy 2002; Takemura et al. 2004). Female GSIdid Xuctuate daily and weekly, but these Xuctuationswere relatively small and did not occur in rhythm withthe lunar cycle.

Factors that inXuence the temporal spawning pat-terns of Mycteroperca rosacea were not investigated,and we can only speculate as to why spawning did notfollow the lunar cycle. For aggregating groupers thatoccur in low densities or are solitary during non-repro-ductive periods, lunar-related spawning may enhancereproductive output via increased mate encounterrates and synchronized gonadal maturation (Colinet al. 1987; Takemura et al. 2004). Leopard groupersmay not require these cues to form spawning aggrega-tions, because adults aggregate throughout the year.

The seasonality of phyto- and zooplankton bloomsstrongly inXuences the reproductive timing of Wshesthat that live in highly productive environments.Spawning is timed so that larvae are exposed to peakfood densities, which results in increased growth andsurvival of larvae (Hjort 1914; Cushing 1975; Platt et al.2003). There is evidence that the reproductive timingof leopard groupers is inXuenced by seasonal patternsin productivity. The Gulf of California is an area char-acterized by exceptionally high rates of primaryproductivity (Zeitzschel 1969), and plankton concen-trations reach maximum levels in the central Gulf dur-ing the spring spawning season of M. rosacea (Kahruet al. 2004). During our observations at Punta Lobosand Punta Coyote, visibility was often poor due to highdensities of phytoplankton and large swarms of larvalWshes and invertebrates in the water column. The tim-ing of large plankton blooms likely inXuences thereproductive patterns of other reef Wshes in the Gulf,as several species engaged in courtship behavior atPunta Lobos and Punta Coyote during the study period(Lutjanus argentiventris, Bodianus diplotaenia, Kypho-sus elegans, Sphoeroides annulatus, Epinephelus labri-formis).

Non-spawning aggregations

Aggregating groupers are often solitary during thenon-spawning times of the year (Domeier and Colin1997; Sadovy and Eklund 1999); however, the leopardgrouper does not follow this pattern. Adults occur insmall to large groups throughout the year, and Wsh arerarely solitary (B.E. Erisman, personal observation).

123

Mar Biol (2007) 151:1849–1861 1859

Non-spawning aggregations usually contain fewer Wshthan spawning aggregations, although feeding aggrega-tions may be quite large. Feeding aggregations ofMycteroperca rosacea may include more than one hun-dred individuals and form in areas characterized byhigh densities of small, schooling Wshes such as Haren-gula thrissina and Cetengraulis mysticetus (Hobson1965; Parrish 1992). The temporal and spatial dynamicsof leopard grouper feeding aggregations have not beenstudied but are necessary to fully understand theirsocial system and to compare with spawning aggrega-tions. Both feeding and spawning aggregations of M.rosacea have been observed at Punta Coyote (Parrish1992; B.E. Erisman, personal observation), which sug-gests that they may be present at speciWc sites forextended periods and use such sites for multiple pur-poses.

Management considerations

The extensive duration and site-speciWcity of spawningaggregations and the propensity to form feeding aggre-gations year-round make leopard groupers particularlyvulnerable to Wshing pressure. The eVects of Wshing onthe population structure and reproductive activities arewell documented for many reef Wshes that form spawningaggregations, including several species of Mycteropercain the western Atlantic. Such eVects include disrup-tions of spawning activity, decreases in egg production,shifts in sex ratios and population structure, reductionsin the sizes and ages at Wrst maturity, and the localextirpation of aggregation sites (Coleman et al. 1996;McGovern et al. 1998; Harris et al. 2002; Levin andGrimes 2002; Matos-Caraballo et al. 2006; Sadovy andDomeier 2005). In turn, these changes may reducepopulation growth rates and slow recovery followingexploitation (Hutchings and Reynolds 2004). OverWsh-ing of aggregations also inXuences ecosystem stabilityby causing major shifts in Wsh community structure(Pauly et al. 1998; Sala et al. 2004).

OverWshing of leopard grouper spawning and feed-ing aggregations has resulted in substantial declines inlandings and CPUE of this species by regional Wsheriesin the last decade (Sala et al. 2004), and Wsheries man-agement policies have done little to change this trend(Hernandez and Kempton 2003). Current regulationson Mycteroperca rosacea Wsheries in the Gulf of Cali-fornia do not set catch quotas for commercial Wshers,specify daily catch limits for recreational Wshers thatare unreasonably high, and regulations are notenforced (Sala et al. 2004). Conservation and manage-ment policies that protect M. rosacea spawning andfeeding aggregations and limit harvest are needed to

create a more sustainable Wshery and stimulate popula-tion recovery. Potentially eVective strategies includeseasonal bans on Wshing and trading during spawning,inclusion of spawning and feeding aggregation sitesinto no-take marine reserves, Wshing quotas and gearrestrictions for commercial Wshers, and reduced dailycatch limits for recreational Wshers (Sala et al. 2002;Roberts et al. 2001; Colin et al. 2003; Nemeth 2005;Sadovy and Domeier 2005). Of course, any attempts toimprove management of leopard grouper Wsherieswould require continuous monitoring of populationsand the Wshery, as well as strict enforcement of regula-tions (Sala et al. 2003; Sadovy and Domeier 2005).

Acknowledgments We are grateful to R. Lopez for grantingpermission to work within the Loreto Marine Park. We thank R.Murillo and the Dolphin Dive Center for transport to Pt. Lobosand Arturo’s SportWshing for permission to collect data fromsportWshers. We thank J. Sanchez, M. Craig, and J. Hyde for assis-tance in the Weld, B. Macewicz and N. Holland for help with his-tological preparations, H.J. Walker and C. Klepadlo forlaboratory support, Chih-hao Hsieh for help with statistical anal-yses, F. Arcas and Grupo Ecologista Antares (GEA) for informa-tion on spawning behavior, and J. Graham, R. Rosenblatt,L. Allen, N. Holland, J. Moore, R. Warner, C. Petersen, J. Ro-sales-Casián, J. Torre (Comunidad y Biodiversidad, A.C.), andK. Rhodes for their valuable suggestions and comments on thisresearch. This study was supported by research grants to B.E.E.from the PADI Foundation (Proposal # 48), the SIO Center forMarine Biodiversity and Conservation and the National ScienceFoundation (Grant # 0333444), Maxwell Fenmore Fellowship,Carl Hubbs Fellowship, California Seagrant, SIO GraduateDepartment, and SIO Marine Vertebrate Collection, and from aUC MEXUS Small Grant to P.A.H. and J. Rosales-Casián. Addi-tional support was provided by research grants to M.L.B. fromthe National Fish and Wildlife Foundation/Anheuser Busch Con-servation Scholarship, UC MEXUS Dissertation Grant, and theUC Davis Extension University Research Expedition Program(UREP).

References

Aguilar-Perera A, Aguilar-Dávila W (1996) A spawning aggrega-tion of Nassau grouper, Epinephelus striatus (Pisces: Serrani-dae), in the Mexican Caribbean. Environ Biol Fish 43:363–370

Brulé T, Déniel C, Colás-Marrufo T, Renán X (2003) Reproduc-tive biology of gag in southern Gulf of Mexico. J Fish Biol63:1505–1520

Burnett-Herkes J (1975) Contribution to the biology of the redhind, Epinephelus guttatus, a commercially important serra-nid Wsh from the tropical western Atlantic. PhD dissertation,University of Miami, Coral Gables, FL

Chan TTC, Sadovy Y (2002) Reproductive biology, age andgrowth in the chocolate hind, Cephalopholis boenack (Bloch,1790), in Hong Kong. Mar Freshw Res 53:791–803

Choat JH, Robertson DR (1975) Protogynous hermaphroditismin Wshes of the family Scaridae. In: Reinboth R (eds) Inter-sexuality in the animal kingdom. Springer, Heidelberg, pp263–283

Coleman FC, Koenig CC, Collins LA (1996) Reproductive stylesof shallow water groupers of the northeastern Gulf of

123

1860 Mar Biol (2007) 151:1849–1861

Mexico and the consequences of Wshing spawning aggrega-tions. Environ Biol Fish 47:129–141

Colin PL (1992) Reproduction of the Nassau grouper, Epinephe-lus striatus, (Pisces: Serranidae) and its relationship to envi-ronmental conditions. Environ Biol Fish 34:357–377

Colin PL, Shapiro DY, Weiler D (1987) Aspects of the reproduc-tion of two groupers, Epinephelus guttatus and E. striatus inthe West Indies. Bull Mar Sci 40:220–230

Colin PL, Sadovy YJ, Domeier ML (2003) Manual for the studyand conservation of reef Wsh spawning aggregations. SocConserv Reef Fish Aggreg Spec Publ 1:1–99

Cushing DH (1975) Marine ecology and Wsheries. CambridgeUniversity Press, Cambridge

Domeier ML, Colin PL (1997) Tropical reef Wsh spawning aggre-gations: deWned and reviewed. Bull Mar Sci 60:698–726

Donald TJ (1995) Courtship and spawning behavior of the pygmygrouper, Cephalopholis spiloparaea (Serranidae: Epinephe-linae), with notes of C. argus and C. urodeta. Env Biol Fish43:363–370

Erisman BE, Allen LG (2006) Reproductive behavior of a tem-perate serranid Wsh, Paralabrax clathratus (Girard), fromSanta Catalina Island, California, USA. J Fish Biol 68:157–184

Erisman BE, Rosales-Casián JA, Hastings PA (2007) Evidence ofgonochorism in a grouper, Mycteroperca rosacea, from theGulf of California, Mexico. Env Biol Fish (in press)

Gilmore RG, Jones RS (1992) Color variation and associatedbehavior in the Epinepheline groupers, Mycteroperca mi-crolepis (Goode & Bean) and M. phenax (Jordan & Swain).Bull Mar Sci 51:83–103

Gross MR (1996) Alternative reproductive strategies and tactics:diversity within sexes. Trends Ecol Evol 11:92–98

Harris PJ, Wyanski DM, White DB, Moore JL (2002) Age,growth, and reproduction of scamp, Mycteroperca phenax, inthe southwestern north Atlantic, 1979–1997. Bull Mar Sci70:113–132

Hernandez A, Kempton W (2003) Changes in Wsheries manage-ment in Mexico: eVects of increasing scientiWc input and pub-lic participation. Ocean Coast Man 46:507–526

Hjort J (1914) Fluctuations in the great Wsheries of northern Eu-rope. Rapp P-V Reun Cons Int Explor Mer 20:1–227

Hobson ES (1965) Diurnal-nocturnal activity of some inshoreWshes in the Gulf of California. Copeia 1965:291–302

Hoque MM, Takemura A, Matsuyama M, Matsuura S, Takano K(1999) Lunar spawning in Siganus canaliculatus. J Fish Biol55:1213–1222

Humason GL (1972) Animal tissue techniques, 3rd edn. WHFreeman and Company, San Francisco

Hunter JR, Lo NCH, Leong RJH (1985) Batch fecundity in mul-tiple spawning Wshes. In: Lasker R (ed) An egg productionmethod for estimating spawning biomass of pelagic Wsh:application to the northern anchovy, Engraulis mordax. USDep Comm, NOAA Tech Rep NMFS 36, pp 67–78

Hunter JR, Macewicz BJ, Sibert JR (1986) The spawning fre-quency of skipjack tuna, Katsuwonus pelamis, from the southPaciWc. Fish Bull US 84:895–903

Hutchings JA, Reynolds JD (2004) Marine Wsh population col-lapses: consequences for recovery and extinction risk. Bio-Science 54:297–309

Hutchings JA, Bishop TD, McGregor-Shaw CR (1999) Spawningbehaviour of Atlantic cod, Gadus morhua: evidence of matecompetition and mate choice in a broadcast spawner. Can JFish Aquat Sci 56:97–104

Kahru M, Marinone SG, Lluch-Cota SE, Parés-Sierra, MitchellBG (2004) Ocean-color variability in the Gulf of California:scales from days to ENSO. Deep Sea Res II 51:139–146

Lee YD, Park SH, Takemura A, Takano K (2002) Histologicalobservations of seasonal reproductive and lunar-relatedspawning cycles in the female honeycomb grouper Epi-nephelus merra in Okinawan waters. Fish Sci 68:872–877

Levin PS, Grimes CB (2002) Reef Wsh ecology and grouper con-servation and management. In: Sale PF (eds) Coral reef Wsh-es, dynamics and diversity in a complex ecosystem.Academic, New York, pp 377–389

Matos-Caraballo D, Posada JM, Luckhurst BE (2006) Fishery-dependent evaluation of a spawning aggregation of tigergrouper (Mycteroperca tigris) at Vieques Island, PuertoRico. Bull Mar Sci 79:1–16

McGovern JC, Wyanski DM, Pashuk O, Manooch II CS, Sedber-ry GR (1998) Changes in the sex ratio and size at maturity ofgag, Mycteroperca microlepis, from the Atlantic coast of thesoutheastern United States during 1976–1995

Moe MA Jr (1969) Biology of the red grouper, Epinephelus mo-rio (Valenciennes) from the eastern Gulf of Mexico. Fla DepNat Resour Mar Res Lab Professional Papers Ser 10:1–95

Nagahama Y (1983) The functional morphology of teleost go-nads. In: Hoar WS, Randall DJ, Donaldson EM (eds) Fishphysiology, volume IX, reproduction, part A, endocrine tis-sues and hormones. Academic, New York, pp 223–275

Nemeth RS (2005) Population characteristics of a recovering USVirgin Islands red hind spawning aggregation following pro-tection. Mar Ecol Prog Ser 286:81–97

Parrish JK (1992) Levels of diurnal predation on a school of Xat-ironherring, Harengula thrissina. Environ Biol Fish 34:257–263

Pauly D, Christensen V, Dalsgaard J, Froese R, Torres F (1998)Fishing down marine food webs. Science 279:860–863

Petersen CW, Warner RR (1998) Sperm competition in Wshes. In:Birkhead T, Moller A (eds) Sperm competition and sexualselection. Academic, London, pp 435–463

Platt T, Fuentes-Yaco C, Frank KT (2003) Spring algal bloom andlarval Wsh survival. Nature 423:398–399

Press WH, Teukolsky SA, Vetterling WT, Flannery BP (2002)Numerical recipes in C++: the art of scientiWc computing,2nd edn. Cambridge University Press, Cambridge

Ramírez RM, Rodríguez CM (1990) SpeciWc composition of thesmall scale Wshery capture at Isla Cerralvo, BCS, Mexico. InvMar CICIMAR 5:137–141

Rhodes KL, Sadovy Y (2002) Temporal and spatial trends inspawning aggregations of camouXage grouper, Epinepheluspolyphekadion, in Pohnpei, Micronesia. Environ Biol Fish63:27–39

Roberts CM, Bohnsack JA, Gell F, Hawkins JP, Goodridge R(2001) EVects of marine reserves on adjacent Wsheries. Sci-ence 294:1920–1923

Robertson DR, Warner RR (1978) Sexual patterns in the labroidWshes of the western Caribbean, II, the parrotWshes (Scari-dae). Smithson Contrib Zool 255:1–26

Sadovy Y (1996) Reproduction in reef Wshery species. In: PoluninNVC, Roberts CM (eds) Reef Wsheries. Chapman and Hall,London, pp 15–60

Sadovy Y, Colin PL (1995) Sexual development and sexuality inthe Nassau grouper. J Fish Biol 46:961–976

Sadovy Y, Domeier ML (2005). Are aggregation-Wsheries sus-tainable? Reef Wsh Wsheries as a case study. Coral Reefs24:254–262

Sadovy Y, Eklund AM (1999). Synopsis of biological data on theNassau grouper, Epinephelus striatus (Bloch, 1792), and thejewWsh, E. itajara (Lichtenstein, 1822). NOAA Tech RepNMFS 146:1–65

Sadovy Y, Rosario A, Román A (1994a) Reproduction in anaggregating grouper, the red hind, Epinephelus guttatus.Environ Biol Fish 41:269–286

123

Mar Biol (2007) 151:1849–1861 1861

Sadovy Y, Colin PL, Domeier M (1994b) Aggregation andspawning in the tiger grouper, Mycteroperca tigris (Pisces:Serranidae). Copeia 1994:511–516

Sala E, Aburto-Oropeza O, Paredes G, Parra I, Barrera JC, Day-ton P (2002). A general model for designing networks of ma-rine reserves. Science 298:1991–1993

Sala E, Aburto-Oropeza O, Paredes G, Thompson G (2003)Spawning aggregations and reproductive behavior of reefWshes in the Gulf of California. Bull Mar Sci 72:103–121

Sala E, Aburto-Oropeza O, Reza M, Paredes G, López-LemusLG (2004) Fishing down coastal food webs in the Gulf of Cal-ifornia. Fisheries 29:19–25

Samoilys MA, Squire LC (1994) Preliminary observations on thespawning behavior of coral trout, Plectropomus leopardus(Pisces: Serranidae), on the Great Barrier Reef. Bull Mar Sci54:332–342

Shapiro DY (1987) Reproduction in groupers. In: Polovina JJ,Ralston S (eds) Tropical snappers and groupers: biology andWsheries management. Westview Press, Boulder, pp 295–327

Smith CL (1972) A spawning aggregation of Nassau grouper, Epi-nephelus striatus (Bloch). Trans Am Fish Soc 101:257–261

Taborsky M (1994) Sneakers, satellites, and helpers: parasitic andcooperative behavior in Wsh reproduction. Adv Study Behav23:1–100

Taborsky M (1998) Sperm competition in Wsh:’bourgeois’ malesand parasitic spawning. Trends Ecol Evol 13:222–227

Takemura A, Rahman MD, Nakamura, Park YJ, Takano K(2004) Lunar cycles and reproductive activity in reef Wsheswith particular attention to rabbitWshes. Fish Fisheries 5:317–328

Thomson DA, Findley LT, Kerstitch AN (2000) Reef Wshes of theSea of Cortez, the rocky-shore Wshes of the Gulf of Califor-nia, 2nd edn. The University of Texas Press, Austin

Thresher RE (1984) Reproduction in reef Wshes. TFH Publica-tions, Neptune City

van den Berghe EP, Warner RR (1989) The eVects of mating sys-tem on male mate choice in a coral reef Wsh. Behav Ecol So-ciobiol 24:409–415

Warner RR (1982) Mating systems, sex change and sexualdemography in the rainbow wrasse, Thalassoma lucasanum.Copeia 1982:653–661

Warner RR (1984) Mating behavior and hermaphroditism in cor-al reef Wshes. Am Sci 72:128–136

Warner RR (1997) Sperm allocation in coral reef Wshes, strategiesfor coping with demands on sperm competition. BioScience47:561–564

Warner RR, HoVman SG (1980a) Local population size as a deter-minant of a mating system and sexual composition in twotropical reef Wshes (Thalassoma spp.). Evolution 34:508–518

Warner RR, HoVman SG (1980b) Population density and theeconomics of territorial defense in a coral reef Wsh. Ecology61:772–780

Warner RR, Harlan RK (1982) Sperm competition as a determi-nant of mating system and sexual composition in two tropicalmarine Wshes (Thalassoma spp). Evolution 34:508–515

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

Zeitzschel B (1969) Primary productivity in the Gulf of Califor-nia. Mar Biol 3:201–207

Zeller DC (1998) Spawning aggregations: patterns of movementof the coral trout Plectropomus leopardus (Serranidae) asdetermined by ultrasonic telemetry. Mar Ecol Prog Ser162:253–263

123