UNIVERSITY OF
ILLINOIS LIBRARY
AT URBANA-CHAMPAIGN
BIOLOGY
MAR 2 6 198S
FIELDIANA
Zoology
Published by Field Museum of Natural History
New Series, No. 8
THE BLENNIOID FISHES OF BELIZE AND HONDURAS,
CENTRAL AMERICA, WITH COMMENTS ON THEIR
SYSTEMATICS, ECOLOGY, AND DISTRIBUTION
(BLENNIIDAE, CHAENOPSIDAE, LABRISOMIDAE,
TRIPTERYGIIDAE)
DAVID W. GREENFIELD
ROBERT KARL JOHNSON
BlOlOQy um
WlBUiJfiiu,^"
JUN 1 0 1982,
October 5, 1981
Publication 1324
THE BLENNIOID FISHES OF BELIZE AND HONDURAS,
CENTRAL AMERICA, WITH COMMENTS ON THEIR
SYSTEMATICS, ECOLOGY, AND DISTRIBUTION
(BLENNIIDAE, CHAENOPSIDAE, LABRISOMIDAE,
TRIPTERYGIIDAE)
iy\ a
FIELDIANA
Zoology
Published by Field Museum of Natural History
New Series, No. 8
THE BLENNIOID FISHES OF BELIZE AND HONDURAS,
CENTRAL AMERICA, WITH COMMENTS ON THEIR
SYSTEMATICS, ECOLOGY, AND DISTRIBUTION
(BLENNIIDAE, CHAENOPSIDAE, LABRISOMIDAE,
TRIPTERYGIIDAE)
DAVID W. GREENFIELD
Department of Biological Sciences
Northern Illinois University
Research Associate
Division of Fishes
Field Museum of Natural History
ROBERT KARL JOHNSON
Division of Fishes
Field Museum of Natural History
Department of Biological Sciences
Northern Illinois University
Accepted for publication October 2, 1978
October 5, 1981
Publication 1324
Library of Congress Catalog Card No.: 81-65060
ISSN 0015-0754
PRINTED IN THE UNITED STATES OF AMERICA
CONTENTS
List of Illustrations vii
List of Tables viii
Abstract 1
Acknowledgments 1
Introduction 2
Methods 6
Descriptions 6
Material Examined 8
Sampling Effort 9
Habitat Categories 9
Data Analysis 11
Species Accounts
Labrisomidae 14
Labrisotnus 15
(1) L. albigenys 15
(2) L. bucciferus 15
(3) L. gobio 19
(4) L. guppyi 20
(5) L. haitiensis 22
(6) L. kalisherae 22
(7) L. nigricinctus 23
(8) L. nuchipinnis 23
Malacoctenus 24
(9) M. aurolineatus 24
(10) M. boehlkei 25
(11) M. delalandei 27
(12) M. erdmani 27
(13) M. gilli 28
(14) M. macropus 28
(15) M. triangulates 29
Paraclinus 30
(16) P. barbatus 30
(17) P. cingulatus 32
(18) P. fasciatus 33
(19) P. infrons 33
(20) P. marmoratus 35
(21) P. nigripinnis 36
Starksia 36
(22) S. atlantica 37
(23) S. elongata 40
(24) S. hassi 41
(25) S. lepicoelia 41
(26) S. nanodes 42
(27) S. occidentalis 42
(28) S. sluiteri 43
(29) S. starch 44
Stathmonotus 44
(30) S. gymnodermis 45
vi „ CONTENTS
(31) S. hemphilli 45
(32) S. stahli tekla 46
Tripterygiidae 46
Enneanectes 46
(33) £. altivelis 47
(34) E. atrorus 49
(35) E. boehlkei 49
(36) E. jordani 50
(37) E. pectoralis 50
Chaenopsidae 51
Acanthemblemaria 51
(38) A. aspera 53
(39) A. greenfieldi 55
(40) A. maria 55
(41) A. spinosa 55
Coralliozetus 56
(42) C. cardonae 56
Emblemaria 57
(43) E. caldwelli 57
(44) E. hyltoni 58
(45) E. pandionis 61
Emblemariopsis 61
(46) E. leptocirris 61
(47) E. pricei 62
(48) E. signifera 65
Hemiemblemaria 65
(49) H. simulus 66
Lucayablennius 66
(50) L. zingaro 66
Blenniidae 66
Entomacrodus 67
(51) E. nigricans 67
Hypleurochilus 69
(52) H. aequipinnis 69
(53) H. springeri 70
Lupinoblennius 70
(54) L. dispar 70
Ophioblennius 71
(55) O. atlanticus macclurei 71
Parablennius 72
(56) P. marmoreus 72
Scartella 72
(57) S. cristata 72
Discussion 73
Comparisons Between Localities 73
Recurrent Group Analysis 77
Recurrent Groups Based on Belize and Honduras Data 77
Recurrent Groups Based on Belize Data 82
Recurrent Groups Based on Glovers Reef Data 83
Summary of Composition of Recurrent Groups 83
Analyses Based on Pooled Data 84
Zoogeographic Implications 91
Addendum 93
Literature Cited 102
LIST OF ILLUSTRATIONS
1. Location of collecting sites in Belize 4
2. Location of collecting sites in Honduras 5
3. Starksia atlantica Longley, 1934, female and male 39
4. Emblemaria hyltoni Johnson & Greenfield, 1976, female 60
5. Emblemariopsis pricei Greenfield, 1975, light and dark morphs 64
6. Relationship between six general localities in Belize and Honduras based on
number of shared species 76
7. Composition of recurrent groups based on Belize and Honduras data 78
8. Composition of recurrent groups based on Belize data only 79
9. Composition of recurrent groups based on Glovers Reef data only 80
10. Starksia starcki Gilbert, 1971, juvenile and female 96
11. Hypleurochilus aequipinnis (Giinther, 1861), male 100
Vll
LIST OF TABLES
1. List of the blennioid fishes taken in Belize and Honduras 7
2. Distribution of sampling effort by depth strata 10
3. Distribution of sampling effort by geographic locality 10
4. Distribution of sampling effort by geographic locality and habitat category 12
5. Frequency distribution of GMI values 14
6. Depth of capture records for 27 blennioid species 16
7. Meristic and morphometric data for specimens of Labrisomus albigenys 18
8. Co-occurrence of species of Labrisomus 19
9. Capture of Labrisomus species with respect to habitat categories 21
10. Co-occurrence of species of Malacoctenus 25
11. Capture of Malacoctenus species with respect to habitat categories 26
12. Co-occurrence of species of Paraclinus 30
13. Capture of species of Paraclinus, Starksia, and Stathmonotus with respect
to habitat categories 34
14. Meristic and morphometric data for specimens of Paraclinus infrons 35
15. Co-occurrence of species of Starksia 38
16. Comparison of certain meristic characters in Starksia elongata and S. nanodes 42
17. Comparison of certain meristic characters in Starksia lepicoelia and S. starcki 45
18. Co-occurrence of species of Stathmonotus 45
19. Co-occurrence of species of Enneanectes 47
20. Capture of Enneanectes species with respect to habitat categories 48
21. Depth and habitat distribution of co-occurrence records for four species
of Enneanectes 49
22. Co-occurrence of chaenopsid species 52
23. Capture of chaenopsid species with respect to habitat categories 54
24. Co-occurrence of blenniid species 67
25. Capture of blenniid species with respect to habitat categories 68
26. Similarity between six general localities in Belize and Honduras based on
numbers of shared species 74
27. Correspondence between calculated SI values and sampling effort 74
28. Recurrent groups of blennioid fishes formed at higher affinity levels 81
29. Captures of recurrent groups with respect to habitat categories and
geographic localities 85
30. Capture records for members of five recurrent groups at six general localities
in Belize and Honduras 86
31. Capture records for members of five recurrent groups tabulated for
nine habitat categories 87
32. Concordance in capture records among members of five recurrent groups 89
33. Pooled capture data for members of five recurrent groups at six general
localities in Belize and Honduras 90
34. Pooled capture data for members of five recurrent groups tabulated for
nine habitat categories 91
ABSTRACT
A total of 57 species of blennioid fishes representing four families (Blenniidae,
Chaenopsidae, Labrisomidae, Tripterygiidae) is reported from collections made
by the authors in Belize and Honduras, Central America, during the period 1970
through 1978. Twenty of these species are reported for the first time from Belize,
40 for the first time from Honduras, and 15 for the first time from the Caribbean
coast of Central America. Species accounts, presented for each species, include
available information on habitat association, depth distribution, co-occurrence
with other blennioid species, and geographic distribution. Where appropriate,
other information relevant to the systematics and biology of certain species is
included. Comparison of captures of blennioid species from different localities in
Belize and Honduras reveals an inshore to offshore gradient in species rich-
ness— many more blennioid species are found in offshore habitats than occur at
inshore (and especially mainland) habitats in areas we have sampled — but the
analysis also reveals changes in the composition of assemblages of blennioid
species along the inshore to offshore gradient. Recurrent group analysis resulted
in recognition of five principal groups of blennioid fishes, accounting for 23 of
the 57 species. Group I species are associated with offshore, shallow-water,
coral-rich habitats and especially with shallow, windward portions of fringing-
reefs and the Belize Barrier Reef. Group II species are associated with offshore,
relatively deep-water (greater than 45 ft), coral-rich habitats and especially with
fore-reef dropoff zones. Group III species are associated with offshore,
shallow-water, coral-rich habitats and particularly with those in relatively quiet
water such as patch reefs in atoll lagoons and patch-reef-like formations behind
the Belize Barrier Reef. Group IV species are associated with offshore, very
shallow water sites and especially with reefcrest intertidal pools or areas in the
immediately adjacent subtidal. Group V species are associated with more in-
shore (but not mainland) shallow-water habitats, typically with rock and cobble
or coarse coral rubble predominating. Captures of blennioid species from the
Caribbean coast of Central America documented in this paper raise serious ques-
tions regarding the validity of Briggs' (1974) recognition of an entirely insular
West Indian Zoogeographic Province and, more importantly, firmly demon-
strate the need for additional collecting in the western Caribbean.
ACKNOWLEDGMENTS
We are indebted to the governments of Belize and of the Republic of Honduras
for permission to collect fishes in those countries. We are especially indebted
to Mr. Winston Miller, Fisheries Administrator, Belize, and Lie. Humberto
Cabellero L., Director General de Recursos Naturales Renovables, Republica de
Honduras, C. A., for their aid in providing requisite permits.
2 FIELDIANA: ZOOLOGY
We were assisted in collecting specimens by numerous persons and especially
by A. Drew, M. Drew, G. Glodek, T. Greenfield, N. Hylton, F. Miller, R. Miller,
J. Russo, J. Thomerson, D. Wildrick, R. Williamson, R. Woods, and students
enrolled in the tropical studies program of the Associated Universities for Inter-
national Education. We also thank Rev. L. Dieckman, S.J., for assistance in
arranging field work in Belize; G. and M. J. Lomont for providing accommoda-
tions and assistance at Glovers Reef Village; and N. Hylton, captain of the MS
Miss Sabrina for invaluable assistance in field work in Honduras.
The following curators kindly lent or made available material for study: J. E.
Bohlke, Academy of Natural Sciences of Philadelphia (ANSP); C. R. Gilbert,
Florida State Museum, University of Florida (UF); C. R. Robins, Rosenstiel
School of Marine and Atmospheric Science, University of Miami (UMML); R.
Spieler, Milwaukee Public Museum.
The Division of Photography, Field Museum of Natural History, provided
photographs of the charts and line drawings. Mr. Z. Jastrzebski, Department of
Exhibition, Field Museum of Natural History, prepared the drawings. G.
Glodek, B. Peyton, and G. Whitmire aided in preparation of the manuscript.
Analysis of the data was largely done using the facilities of the Computation
Center of the University of Chicago. Patricia H. Johnson typed both rough and
final versions of the complete manuscript.
C R. Gilbert, R. F. Inger, and the late L. P. Woods read the manuscript and
offered valuable suggestions for its improvement.
Our field work in Belize was made possible in part through the support of the
following organizations: National Science Foundation (BMS75-08684, D. W.
Greenfield); American Philosophical Society, Philadelphia (Johnson Fund,
Grant Number 982, to D. W. Greenfield); Associated Universities for Interna-
tional Education; Council of Academic Deans, Northern Illinois University; Field
Museum of Natural History. We thank Klaus Ruetzler (Smithsonian Institution)
for arranging for field work at Carrie Bow Cay and for support from the IMSWE
Project and an Exxon grant to IMSWE. This report constitutes Contribution No.
35, Investigations of Marine Shallow Water Ecosystems Project, Smithsonian
Institution.
Our field work in Honduras was made possible in part through the support of
the following organizations: American Philosophical Society, Philadelphia
(Johnson Fund, Grant Number 1220, to R. K. Johnson); Field Museum of Natural
History; Northern Illinois University; University of Michigan at Ann Arbor;
Wrigley Fund for Marine Biological Research (through a grant to R. K. Johnson).
This paper is based in part on the results of the Miskito Coast Expedition
(1975) to Honduras and Nicaragua, jointly sponsored by Field Museum of
Natural History, Northern Illinois University, and the University of Michigan at
Ann Arbor.
We are especially indebted to our wives, Teresa A. Greenfield and Patricia H.
Johnson, whose assistance in field, laboratory, or office, made this work pos-
sible.
INTRODUCTION
This paper is an annotated checklist of the blennioid fishes from Belize and
Honduras, Central America. The following families are treated: Labrisomi-
GREENFIELD & JOHNSON: BLENNIOID FISHES 3
dae,1 Chaenopsidae,2 Tripterygiidae,2 and Blenniidae. Records are presented for
57 species: Labrisomidae (32), Chaenopsidae (13), Tripterygiidae (5), Blenniidae
(7) (table 1). Twenty of these species are recorded for the first time from Belize,
40 for the first time from Honduras, and 15 for the first time from the Caribbean
coast of Central America. Available information on habitat association, depth
distribution, co-occurrence with blennioid species, and geographic distribution
is presented for each species, and, for a number of the more poorly known
species, supplementary descriptive information is also presented.
Information on the marine fishes of the Atlantic coast of Central America
(Yucatan Peninsula, Mexico to Panama) is scant and scattered (see Caldwell,
1963; Gilbert & Kelso, 1971; Robins, 1971b, 1972). By far the most significant
work is that of Meek & Hildebrand (1923 to 1928) for Panama, now both old and
of limited use because of the great number of subsequent taxonomic studies and
because greatest attention was given to the Pacific side of Panama. Additional
reports on small collections of fishes from Panama include those of Fowler (1916,
1917, 1930), Breder (1925), Rubinoff & Rubinoff (1962), and Caldwell & Caldwell
(1964). Reports on several small collections of marine fishes from Costa Rica have
been presented by Fowler (1916), Caldwell et al. (1959), Caldwell (1963), and
Gilbert & Kelso (1971). Reports on two small collections from mainland
Nicaragua are given by Fowler (1903, 1923). Reports on collections from Cour-
town Cays and Albuquerque Cays are also presented by Fowler (1944). In addi-
tion, reports have been presented on collections from San Andres (Fowler, 1944,
1950) and Isla Providencia (Old Providence) (Schmitt & Schultz, 1940; Fowler,
1944, 1950; Burgess, 1978). Breder (1927) reported on the first collection of fishes
from Glovers Reef, Belize. The most recent report on the fishes of the Central
American coastal area is that of Birdsong & Emery (1968), which includes records
from Yucatan, Mexico, Belize, and Courtown and Albuquerque Cays off Nica-
ragua. Burgess (1978) in his unpublished master's thesis (University of Florida,
Gainesville) provides valuable information on the geographic range and vertical
distribution of some 53 species of blennioid fishes taken at Grand Cayman and
Isla Providencia. In addition to these general references, there are numerous
references to specimens from the Central American coast in taxonomic studies of
various fish groups. References to these studies appear where appropriate
throughout this paper.
The information on the blennioid fishes of Belize and Honduras presented in
this paper is the result of a number of separate expeditions. D. W. and T. A.
Greenfield, along with several collaborators, conducted a survey of the fishes of
Belize from 1970 through 1978. This work resulted in a total of nearly 200 marine
collections. The second major effort was made during the Miskito Coast Expedi-
tion in 1975. This expedition, jointly sponsored by Field Museum of Natural
History, Northern Illinois University, and the University of Michigan Museum
of Zoology, resulted in 31 marine collections from Honduras. In addition to the
material collected by us, we have also examined Belize specimens deposited at
the University of Miami Rosenstiel School of Marine and Atmospheric Science
[material collected mostly by W. A. Starck II and reported by Birdsong & Emery
(1968)]. An additional collection of fishes made in Belize by Mr. John W. Cooper
'Usage of this name follows George & Springer (1980).
2We choose to follow Rosenblatt (1959), Stephens (1963), and George & Springer (1980)
in recognizing as distinct the Chaenopsidae and the Tripterygiidae.
4 FIELDIANA: ZOOLOGY
(JWC station numbers) of the Smithsonian Institution, from Carrie Bow and
South water Cays, was also examined.
Collecting sites in Belize and Honduras are indicated in Figures 1 and 2.
Abbreviations for localities in Belize are: (AC) Ambergris Cay; (BC) Bugle Cays;
(BZC) Belize mainland, sites near Belize City; (GPL) Belize mainland, southern
lagoon at Gales Point; (BRB) Barrier Reef, 1 km south of Buttonwood Cay; (BRG)
Barrier Reef at Gallows Point; (BRT) Barrier Reef at Tom Owens Cay; (CBC)
Fig. 1. Location of collecting sites in Belize. AC, Ambergris Cay; BC, Bugle Cays; BRB,
Barrier Reef, 1 km S of Buttonwood Cay; BRG, Barrier Reef at Gallows Point; BRT, Barrier
Reef at Tom Owens Cay; CBC, Carrie Bow Cay; FC, Frenchman's Cay; GR, Glovers Reef;
LR, Lighthouse Reef; P, Placentia; SC, Snake Cays; SWC, South Water Cay; TI, Turneffe
Island.
V
1
01
CO
S"
U
0>
o
«
ui
&
>
o
U
o
3
5
H
K
u
3
>
u
u
u
C 00
8
II
EC
,3
c
CO
s
2 e
CO
£
J
cu
3
CO
£
(0
n
7
3
ii
TJ
r
U
0
^x
x is
,g
£
M
-fi
X;
5
M
1
i
1
1
EL
9
8
c
c
5
o
~
M
i?
|
J
5
tN
(J
>
uu
0
6 FIELDIANA: ZOOLOGY
Carrie Bow Cay; (FC) Frenchman's Cay; (GR) Glovers Reef; (LR) Lighthouse
Reef; (P) Belize mainland near Placentia; (SC) Snake Cays; (SWC) South Water
Cay; (TC) Tarpon Cay (just north of FC and not indicated on fig. 1); (TI) Turneffe
Island. Abbreviations for localities in Honduras are: (CC) Hog Islands (Cayos
Cochinos); (BRL) Honduras mainland, Cannon Island, Brus Lagoon; (TR) Hon-
duras mainland, main pier at Trujillo; (IR) Isla Roatan.
Habitats represented in our collections are varied, ranging from mangrove
swamps to well-developed coral atolls. The entire coastal area of Belize is bor-
dered by mangrove swamps that often connect to inland lagoons, providing a
gradation from brackish water to freshwater. Offshore, between the mainland
and the Barrier Reef, the water tends to be clearer and more saline than along the
coast; however, this area still receives a large amount of freshwater runoff from
the large tropical rivers during the rainy season, and the water is more turbid
and less saline than more offshore locations. In this zone are numerous
mangrove-covered cays, often with shores of coral rubble covered with heavy
algal growth. The bottom is generally characterized by wide areas of sand and
Thalassia, although scattered coral heads are present off the cays. Toward the
Barrier Reef the water becomes progressively clearer, more saline, and coral
development increases.
The Barrier Reef of Belize, which is the second longest in the world, varies
from 8 to 25 miles offshore and stretches southward from the Yucatan Peninsula,
Mexico, for 168 miles into the Gulf of Honduras. Oceanic conditions generally
prevail on the windward side of this reef; however, the coral development does
not equal that of the offshore atolls.
There are three atolls in Belize: Turneffe Island, Lighthouse Reef, and Glovers
Reef. Turneffe is closest to the Barrier Reef and has the most extensive land
development, including mangrove islands within its lagoon. Lighthouse and
Glovers Reefs are farther offshore and have the best coral development. These
atolls lack large islands, although small cays are present. Their lagoons contain
numerous patch reefs, and on the outside of the reef precipitous dropoffs are
common. The conditions at these atolls are definitely oceanic and insular.
The habitats sampled in Honduras also varied widely. Roatan, a high island,
has areas influenced by stream runoff that support mangrove growth. Rocky
shores drop down to meet coral rubble and scattered coral heads and patch reefs
within some of the bays. Offshore are well-developed coral reefs with precipi-
tous dropoffs.
The Hog Islands (Cayos Cochinos) are high and rocky. The rocky cliffs drop
down to the water's edge and often extend out into the water as points,
separating beaches of sand and gravel. On the leeward side of the islands algae-
covered rocks extend from shore down to meet a bottom of coral rubble, beyond
which are found scattered coral heads that farther offshore grade into well-
developed reefs with well-marked dropoffs. On the windward side of the is-
lands beaches of coral-algal sand extend into areas of coral rubble, and these into
areas of living coral in the form of fringing reefs.
METHODS
Descriptions
A number of the species discussed in this paper are known from very few
specimens and/or from very few previously reported localities. For those species
Table 1. List of the blennioid fishes from Belize and Honduras discussed in this paper.
The species number assigned to each species is used throughout this paper.
Species
New to
New to
Number of
Number of
No.
Species
Belize
Honduras
specimens
lots
Labrisomidae
Labrisomus
1
2
3
albigenys
bucciferus
gobio
—
X
X
X
7
95
48
2
18
10
4
5
guPW
haitiensis
—
X
X
270
70
29
18
6
kalisherae
—
X
1%
11
7
8
nigricinctus
nuchipinnis
X
X
41
71
13
16
Malacoctenus
9
aurolineatus
X
X
173
12
10
boehlkei
—
X
57
12
11
delalandei
—
X
22
4
12
erdmani
—
X
40
9
13
gilli
—
—
156
23
14
15
macropus
triangulatus
—
X
X
166
254
21
32
Paraclinus
16
barbatus
X
—
2
1
17
cingulatus
—
X
1
1
18
19
fasciatus
infrons
X
X
27
7
5
6
20
marmoratus
X
—
2
2
21
nigripinnis
—
X
121
14
Starksia
22
atlantica
—
X
70
22
23
24
elongata
hassi
X
X
X
4
2
3
2
25
26
lepicoelia
nanodes
—
X
103
57
24
20
27
occidentalis
—
—
53
14
28
sluiteri
X
—
6
4
29
starcki
—
X
1
1
Stathmonotui
30
31
32
gymnodermis
hemphilli
stahli tekla
X
X
X
1
1
45
1
1
13
Tripteryghdae
Enneanectes
33
altivelis
—
X
113
28
34
atrorus
—
X
39
17
35
boehlkei
—
X
190
30
36
37
jordani
pedoralis
X
X
X
9
87
4
20
Chaenopsidae
Acanthemblemaria
38
39
40
aspera
greenfieldi
maria
X
X
140
106
2
25
20
2
41
spinosa
—
X
50
17
Coralliozetus
42
cardonae
X
—
2
1
FIELDIANA: ZOOLOGY
Table 1. Continued.
Species
New to
New to
Number of
Number of
No. Species
Belize
Honduras
specimens
lots
Emblemaria
43 caldwelli
—
—
71
17
44 hyltoni
—
—
6
1
45 pandionis
—
—
25
4
Emblemariopsis
46 leptocirris
X
X
7
6
47 pricei
—
X
10
7
48 signifera
X
X
5
5
Hemiemblemaria
49 simulus
X
X
5
3
Lucayablennius
50 zingaro
—
X
78
24
Blenniidae
Entomacrodus
51 nigricans
—
X
367
21
Hypleurochilus
52 aequipinnis
X
X
18
3
53 springeri
X
X
9
7
Lupinoblennius
54 d/spar
X
X
19
4
Ophioblennius
55 atlanticus
macclurei
—
X
377
30
Parablennius
56 marmoreus
X
—
1
1
Scartella
57 cristata
X
X
33
7
Totals
20
40
3,939
for which we felt that information obtained from Belize and Honduras speci-
mens added substantially to our knowledge of the species in question, we pre-
sent limited descriptive information. The methods used to obtain meristic, mor-
phometric, and other quantitative data follow those used by the most recent
revisor of the taxon being considered. All measurements are straight-line,
point-to-point measurements taken to the nearest 0.1 mm with dial calipers or
needle-point dividers, or to the nearest 0.01 mm with an ocular micrometer on a
Wild M5 microscope. Unless otherwise noted, all lengths are the standard length
(SL) in mm.
Material Examined
In all cases material taken by us is listed first, by Field Museum of Natural
History (FMNH) catalogue number, with the total number of specimens indi-
cated in parentheses. The totals given for number of lots and number of speci-
mens examined (table 1 and under each species account) are based solely on
FMNH material. Only material collected by us is included in the ecological
analyses for each species and group of species. Representative series of virtually
all species collected by us in Honduras will be deposited at the Museum of
Zoology, University of Michigan.
GREENFIELD & JOHNSON: BLENNIOID FISHES 9
Any additional Belize material examined by us or referred to by previous
authors is listed following the list of FMNH material. Abbreviations used in
reference to additional material examined are: JWC — material obtained by John
W. Cooper of the Smithsonian Institution at Carrie Bow and South Water Cays
(material listed by original field number); UMML — material deposited at the
Rosenstiel School of Marine and Atmospheric Science, University of Miami, and
listed by UMML catalogue number.
Sampling Effort
Of more than 230 marine collections taken by us in Belize and Honduras
between 1970 and 1978, 127 collections yielded one or more species of blennioid
fish. Techniques utilized in obtaining the samples were varied, including
ichthyocide (chemfish, noxfish, pronoxfish), anaesthetic (quinaldine), seines,
handnets, and various combinations of these. Equally varied were the number of
persons involved and the time spent in obtaining each collection.
Inspection of the original field data (on file at Field Museum of Natural His-
tory) allowed division of the collections into two categories: Class 1 collections
were made utilizing broadcast collecting techniques (e.g., ichthyocide); Class 2
collections were much more biased in terms of sampling technique, e.g., the use
of quinaldine to collect a particular species in a particular location. Obviously
categorization of collections into one class or the other is partly subjective. Class
1 collections grade from those with partial to those with substantial bias in terms
of the probability of collecting all blennioid species present at a given site.
Nonetheless we felt it imperative to eliminate those stations clearly identifiable
as Class 2 collections from certain categories of data analysis. Some 23 Class 2
stations were recognized, leaving 104 Class 1 stations. Where appropriate, this
division is indicated in the following discussion.
There exist severe biases in our data with respect to depth of collection and site
of collection — some depth intervals and localities being repeatedly collected,
others represented by very few samples. These biases reflect solely the oppor-
tunities (or lack thereof) available to us over the eight-year period of this study.
In this paper depth intervals are reported in feet. This is because our depth
gauges and hence all original field data were recorded in feet. We feel that to
change all of these values to meters would result in a needless proliferation of
decimal points.
We have tabulated sampling effort vs. depth of sampling using three different
bases for recording depth of sampling (table 2). All three clearly show a bimodal
distribution of sampling effort; most of our collections were taken in less than 15
ft or more than 50 ft of water.
Equally severe is sampling bias with respect to localities sampled (table 3).
Although we have collected one or more species of blennioid fish at 18 separate
localities in Belize and Honduras, three localities, Glovers Reef (64 stations),
Carrie Bow Cay (27 stations), and the Hog Islands (11 stations), account for 80%
of the total effort (127 stations).
Habitat Categories
In our efforts to discern a pattern in the distribution of individual species and
groups of species, we have attempted to categorize stations according to
Table 2. Distribution of sampling effort (number of stations) by depth of capture. Three
different ways of tallying depth of capture are used. The shallowest and greatest categories
reflect tallies based respectively on the upper or lower recorded limits of sampling for a
given station. The "average" depth category is a tally based on adding the arithmetic mean
of the upper and lower depth limits recorded for a given sample added to the upper
(shallowest) depth limit for that sample (as recorded in the field).
Depth
Shallowest
Average
Greatest
(feet)
depth
depth
depth
0 to 4
80
60
50
5 to 9
2
19
13
10 to 19
2
5
20
20 to 29
2
2
2
30 to 39
3
3
4
40 to 49
3
0
0
50 to 59
7
10
7
60 to 110
28
28
31
Totals
127
127
127
Table 3. Distribution of sampling effort (number of stations) by locality for Belize and
Honduras.
No. of
Total no.
of stations
(Class 1
Locality
stations)
A. Honduras — Offshore Localities
Hog Islands (CC)
Isla Roatan (IR)
11 (11)
8 (8)
Totals
19 (19)
B. Belize and Honduras — Mainland Localities
Belize City (BZC)
Placentia (P)
Gales Point (GPL)
Brus Lagoon (BRL)
Trujillo (TR)
3 (3)
1 (1)
1 (1)
1 (1)
1 (1)
Totals
7 (7)
C. Belize — Cays inside Barrier Reef
Frenchman's Cay (FC)
Bugle Cays (BC)
Snake Cays (SC)
Tarpon Cay (TC)
2 (2)
1 (1)
1 (1)
1 (1)
Totals
5 (5)
D. Belize — Sites on Barrier Reef
Gallows Point (BRG)
Ambergris Cay (AC)
Buttonwood Cay (BRB)
Tom Owens Cay (BRT)
Carrie Bow Cay (CBC)
3 (3)
1 (1)
1 (0)
1 (1)
21 (21)
Totals
27 (26)
E. Belize — Atolls
Glovers Reef (GR)
Lighthouse Reef (LR)
64(42)
5 (5)
Totals
69(47)
species
taken
38
25
44
7
11
10
_9
16
12
3
6
15
34
38
39
18
39
10
GREENFIELD & JOHNSON: BLENNIOID FISHES 11
"habitat." Habitat categories, in the sense used in this paper, represent no more
than an effort on our part to classify stations on the basis of dominant substrate.
The categories recognized are listed and defined below (the number in brackets
refers to total number of stations and number of Class I stations, respectively).
The distribution of sampling effort with respect to locality and habitat category is
given in Table 4.
HI — extensive live coral: well-developed reef formations in depths exceeding 30 ft, e.g.,
forereef dropoff areas [38, 30).
H2 — extensive live coral: well-developed reef formations in depths less than 30 ft and in
areas of relatively high wave energy, e.g., shallow, fringing reefs [20, 19].
H3 — extensive live coral: well-developed reef formations in depths less than 30 ft and in
areas of relatively low wave energy, e.g., patch reef sites in atoll lagoons [28, 19].
H4 — rock: areas dominated by rocky substrates (ledges and cobble); no extensive coral
formations [4, 4].
H5 — coral rubble (coarse): areas in which dominant substrate consists of coarse coral
rubble and sand, often with extensive development of algal mats [6, 6].
H6 — coral rubble (fine) and sand: outer reef flat pools in very shallow (less than 2 ft)
water [7, 6].
H7 — grass beds and sand: areas dominated by extensive beds of marine spermatophytes
15, 5].
H8 — mainland localities: includes all mainland locations [7, 7].
H9 — other: all areas not appropriately placed on one of above; includes stations in
mangrove channels, on beaches of sand and mud, on pier pilings, and stations
representing broad mixtures of two or more habitat categories [12, 8].
It should be noted that categories H8 (mainland sites) and H9 ("other") are
catchall categories, allowing the classification of stations not appropriately
placed in categories HI through H7. Although this is artificial, it prevents prolif-
eration of additional categories, each of which would be represented by one or
two stations. Our habitat categories are comparable with similar classifications
produced by other workers (e.g., Clarke, 1977; Gilbert & Kelso, 1971). We rec-
ognize that our classification is oversimplified and expect that the number of
habitat categories recognized by us in Belize will grow as we continue to collect.
But, given the broadcast nature of ichthyocide collecting, we do not expect the
number of useful categories to be much more than double the number recog-
nized here, and we have found the scheme presented herein useful.
Data Analysis
Final tabulation of species occurrence by station, locality, depth interval, and
co-occurrence with other blennioid species as well as the recurrent group
analysis were performed using Fortran IV programs written by R. K. Johnson.
Data analyses were performed using the IBM 360/370 computer at the University
of Chicago Computation Center.
We attempted to estimate resemblance between major localities in Belize and
Honduras (tables 3, 26, viz., Hog Islands, Roatan, mainland localities, localities
at cays inside the Belize Barrier Reef, localities at the Belize Barrier Reef, atolls)
based on shared species of blennioid fishes. Peters (1968) lists a number of
similarity indices useful in summarizing the relationship (in species composi-
tion) between two areas. We looked at data for six major areas in Belize and
2
a,
CO
00
o
o>
U
o
60
SB LO U"> iH (N VD \o
z
2 tsn
Hrtin N
X I I
c
(0
as
1 N 1
1 \m\
in
fr
3
ra
u
o
,—t
vfi
'-1 1 1
1 Hinl
1^
2
X
1 1
1 |
(X
|!
01 o
in
1 1 1
<tf CN I
vo
2 ^
C
u re
8J
01
cn
0)
s-
cn -O
0) ^
EC
-* 1 1
Mil
■<*
g
in m
1 1
III
cn
■C Oi
o>
■c
? c
^ 5b
52 "ffl
c cn
m
1 «-< 1
1 m cn 1
00
O
00
0)
O co
X
1 1
1 CN |
CN
(J
'■S a
3 In
cn *
"8 J
CO
<n
CI r-l 1
| CN ■<* 1
O
M-l
X
1
1 rH
CN
O
1h
3 —
o>
£
1|
1
+, ra
O .9
lH
1-1 1-H 1
1 tfiOl
00
c
X
1
1 cn
cn
II
<U g
X
c5-
a 2
z
l!
z
HOON
Lf)N O
K
3
r-l
CN v£>
CN
0
en oi
^1
V
•5«
•£e2
cn
01
2
cn
in
0>
cn •'3.
'cn
O
1-1
in VJ
s-s
. <n
m cn
u
0
-J
T3 w
£ "9
-1 .- C
cn C «
0 0 JS
"2
'35
_g
m
>>
ID "q3
01 01
oSeS
Oi 01 cn
S C O
0>
c
II
H S
« « « *j
U co pa <
cn
O
z
J5
<co'U Q
W U-,
H
12
GREENFIELD & JOHNSON: BLENNIOID FISHES 13
Honduras, and this involved 15 pairwise computations for each index em-
ployed. We calculated values for four of the indices listed by Peters (1968):
(1) Jaccard Coefficient (termed "coefficient of community" by Peters, 1968; see Prentice
& Kain, 1976)
Si = c
(N, + N2 - C)
(2) Burt Coefficient (= Sorenson's Coefficient)
2C
Sl =
(N, + Ns)
(3) Otsuka Coefficient
a- c
VN, X N2
(4) Preston Coefficient (see Preston, 1962, p. 418)
SI = 1 - Z
where Z is determined by solution (for Z) of the following expression
N, \k , / N2 \|
N, + N2 - C J ' \N, + N2 - CJ =
In each case the following symbols apply:
C = number of species in common to two areas
N, = number of species in the area with fewer species
N2 = number of species in the area with more species
SI = calculated index (similarity index)
Each of these indices varies between a minimum value of zero (C=0) and a
maximum value of one (N, = C, Nt = N2). Values calculated for the 15 pairwise
comparisons for each of the four indices considered were highly concordant
(Kendall's Coefficient of Concordance, W4>15 = 0.993, p < < .01, see Tate &
Clelland, 1957; Downie & Heath, 1959), and we have chosen the Jaccard Coeffi-
cient for the comparisons presented in Table 26. This coefficient was used by
Hagmeier & Stults (1964) in their analysis of North American mammal pro-
vinces.
In determining the composition of groups of associated blennioid species, we
have used the recurrent group method developed by Fager (1957) and used by
Inger & Chin (1962), Fager & McGowan (1963), and Fager & Longhurst (1968),
among others. In this method a dichotomous index of relationship is calculated
for every pair of species. A number of the indices listed by Peters (1968) could be
used for this purpose, but we follow Inger & Chin (1962), Fager & McGowan
(1963), and Fager & Longhurst (1968) in using an index that calculates the
geometric mean of the proportion of joint occurrences (as in the Otsuka Coeffi-
cient) but applies a correction for sample size. The formula for this index, here-
after referred to as GMI, follows:
GMI = , J _ -4=
v/NA-NB 2VNT
NA is the number of occurrences of species A; NB is the number of occurrences of
species B; J is the number of joint occurrences; and species are assigned to letters
such that NA ^ NB. Pairwise values of GMI were calculated for all 57 species
included in this paper, but only data from Class I stations were used in the
recurrent group analysis.
14 FIELDIANA: ZOOLOGY
For the 57 blennioid species included in this paper, the largest possible GMI
value is 0.894 [(15) Malacoctenus triangulatus, NB = 31; (55) Ophioblennius atlanticus
macclurei, NA = 30; actual GMI for this pair (J =20) is 0.566]. The largest observed
GMI for species included in this paper is GMI = 0.700 [(4) Labrisomus guppyi,
NB = 29, (35) Enneanectes boehlkei, NA = 29; J = 23].
In actual determination of recurrent groups we followed closely the criteria
and methods of Fager (1957, pp. 589-591), except that species represented in
only one or two Class I collections were eliminated from the analysis.
Fager & McGowan (1963) and Fager & Longhurst (1968) arbitrarily chose the
value GMI s 0.500 as the breakpoint for deciding whether or not a given pair of
species showed "affinity." The reason for choosing this value was that ". . .it
was felt that the species should be found together in somewhat more than half
their recorded occurrences if they are to be grouped together" (Fager & McGo-
wan, 1963). Preliminary runs suggested that strict application of the 0.500 break-
point to our data would result in a loss of information; rather few (table 5) of the
1,5% species-pair GMI values were at or above the 0.500 value (due in part we
believe to biases in sampling effort). We therefore chose to perform the recurrent
group analysis at three arbitrarily chosen predetermined breakpoint values:
0.500, 0.400, and 0.300.
Recurrent group analyses were performed for data based on Glovers Reef
samples only, Belize samples only, and Belize and Honduras samples. Unless
otherwise noted, all GMI values recorded in this paper are based on all Belize
and Honduras (Class I) samples.
LABRISOMIDAE
The family Labrisomidae is represented in the tropical western Atlantic by
seven genera and 45 described species. The seven genera are as follows (num-
bers in parentheses refer to total number of western tropical Atlantic species,
number of species known from off the Central American coast, number of
species represented in our material from Belize and Honduras): Haptoclinus
(1,1,0), Labrisomus (9,9,8), Malacoctenus (8,8,7), Nemaclinus (1,1,0), Paraclinus
(8,6,6), Starksia (15,9,8), and Stathmonotus (3,3,3). All of the genera and 37 of the
45 described species occur off the Central American coast. Of these, 32 species
are represented in our collections, representing all but the deep-water mono-
typic genera Haptoclinus Bohlke & Robins, 1974, and Nemaclinus Bohlke &
Springer, 1975. The names employed for labrisomid taxa treated in this paper are
Table 5. Frequency distribution of GMI values based on all 57 species, all Class I Belize
and Honduras stations (see text for additional explanation).
GMI values
Percent
Cumulative
No. of values
(interval)
Frequency
total
percent
higher
s 0.000
1,013
63.47
63.47
583
0.001-0.100
150
9.40
72.87
433
0.101-0.200
189
11.84
84.71
244
0.201-0.300
108
6.77
91.48
136
0.301-0.400
68
4.26
95.74
68
0.401-0.500
43
2.69
98.43
25
0.501-0.600
18
1.13
99.56
7
0.601-0.700
7
0.44
100.0
0
Totals 1,5% 100.00
GREENFIELD & JOHNSON: BLENNIOID FISHES 15
in accord with Springer, 1955a, 1955b, 1959a, 1959b; Bohlke, 1960; Gilbert, 1971;
Bohlke & Robins, 1974; and Bohlke & Springer, 1975.
Labrisomus Swainson, 1893
The genus Labrisomus contains 18 species, nine in the eastern Pacific and nine
in the western Atlantic. One of the western Atlantic species, L. nuchipinnis, also
occurs in the eastern Atlantic. The Pacific members of the genus were revised by
Hubbs (1953), and the Atlantic members were revised by Springer (1959a). Two
additional species, one in the western Atlantic (Springer, 1959b) and one in the
eastern Pacific (Springer & Rosenblatt, 1965), have been recognized since
Springer's (1959a) revision.
All known western Atlantic species of Labrisomus occur off the Central Ameri-
can coast, and all except L. filamentosus Springer, 1959, are represented in our
collections from Belize and Honduras. The latter species was described from the
western Caribbean, and additional specimens have been reported from off
Nicaragua and Hispaniola (Springer, 1959b; Springer & Rosenblatt, 1965); all
specimens are known from depths in excess of 17 fathoms. Only one species of
Labrisomus represented in our collections, L. haitiensis, is represented in samples
taken at depths exceeding 20 ft (table 6).
(1) Labrisomus albigenys Beebe & Tee Van, 1928
This species has been reported from Haiti, Venezuela, the Campeche Banks
off Mexico, and Belize (Springer, 1959a; Birdsong & Emery, 1968; Mago, 1970)
and also from Antigua (C. R. Gilbert, pers. comm.). This constitutes the first
record for Honduras.
Springer's (1959a, pp. 437-438) description of L. albigenys was based on five
specimens, the holotype from Haiti and four specimens (FMNH 59875) from
Cayos Arcos, Campeche Banks, Mexico. We present meristic and morphometric
data for the seven specimens represented in our collections from Honduras
(table 7). We note the rather high degree of bilateral asymmetry in pectoral-fin
ray counts and in lateral line scale counts.
Ecological data. — Labrisomus albigenys appears to be the rarest species of Lab-
risomus in Belize and Honduras. We have collected L. albigenys only at the Hog
Islands and only on two occasions. Both collections were in shallow water (< 18
ft), in areas of good coral development (Habitat = H2) and strong surge at the
edge of the windward fringing reef on Little Hog Island. Species of Labrisomus
taken with L. albigenys include L. guppyi, L. kalisherae, and L. nigricinctus (table 8).
We have no detailed information on the single collection (UMML 9555) from
Lighthouse Reef reported by Birdsong & Emery (1968). Lighthouse Reef is an
area of good coral development.
Material examined. — A total of seven specimens from two collections. HON-
DURAS: CC, 84432 (5), 84433 (2). The Belize record (Birdsong & Emery, 1968)
was based on a single collection from Lighthouse Reef (UMML 9555). We have
not had the opportunity to examine this material.
(2) Labrisomus bucciferus (Poey, 1868)
This species is known from Bermuda, the Bahamas, Florida, the Greater and
Lesser Antilles, and Central America, including Belize, Nicaragua, and Yucatan
5 o^
2 g 2
S3 °'£
o c c
■Si'-*
*-*%
v > £
-a > a
u
in
■as &
on C -t-"
«n *«
3 . «£
73 73 (fl
C .^73
c ^S c
a* Cl, . ■£
N « a; ■"
^S $
c £ <2 fc
g O S O
a» >3 c .**
£'35 e 5
« -O a, oo
73 C H C
■go's
JU oj o "
o *i •** _
O
•S Sj j= oj
_ T3 B 73
■a o a) £
si si «> E
c c^3 -g
1/3 C 7
(5 S 01
S-1 <« ■£ ■£
(0 r- C *^
° §"=- JS
^ -a £ 3
<S Q-x;
•J3
=1
£ O
£»TiO
S " &-S
« § 1 §
' O (fl >-
-T-( 73 "*"
C S ?»
c/5 w (0
§2 52
t-c Ifl
o K
^ H | H I N | « I
\C OO ■<* t}< O O O O
in (nqo cNrooooo o o t-< *-h
OOrHr- lOOHOOKHvOr I H
B « OO OOOOrHCO OOOO r-irHOOOOrHOOfNOOOOO
4> ■ rH
4) O
r
E
9
c £
•St? moo i-hi-hoooo oooo t-hi-hoooocoonoooooo
J=o ^
V
n hn oooooo oooo tNoooooocgincNOoooo
if OO oooooo oooo oooooooooooooo
o
rHO OOrHPlOO oooo oooooooooooooo
t/5
ui « 5 S
K -Si i= ■*
C • ~ O ©
-g ^ 3 o
3!£ £c
£
K k-
_^
fS ^-v
n
i; ^
t-i
«.
•a ^ o
7|
9
5
16
£ IS I 2 I £ I S8 I3ISI8 | £ I *> | S I ^ I S I "> IS
2 «-•
i s i s i jqisi^i fe i^ i« i * |«^ |« | a |
SO OO t"x tx •»* <-i O O
All "w
•0 OO n»r<NOO iDrfOOOO rft^OOOO O O H N H H « On
© '-' H tN
— ^ OO «N vO i-> •"*« O O
r-irlOOOO OOrivOOO t-h i-h O O O O MOO
- =
ri\00000 r-crnOOOO OOOOOO OO
OO NITOOnrf t-ciDOOOO OOOOOO O O i-i ri O O OO C
OO OOOOOO nrtHNOO OOOOrivO OOOOOO
o
O O u
1
e
OO OOOOOO
c
OO 2
II
cvi^< nmooNoo ■** oo as s« so oo ooooi-h<n n Tf Tf \o ^ h
o o o
5. §|
ui -~ .«J
I
9
C
B A
- - -^
II
c S -S
*-* •*> ^-~ *-^ *-* -e> ^--
g cr> 3 in g so
2^ uj 21- ZS- 21- uj 2!*
5 "3
•&>
is z
17
00
•s
a,
en
S
o
u <B
c
CM ^.^^ „
« O
Q ~
S
ig a s a O
CO «H 00 CO rH
"C
6
sO
00 rH 00 C!
o
CO
HHrif)
X
i— I
fr
o
£
TJ
o
CO
C
^o
00HQ0C
(0
CM
rH rH rH CO
u
rH
'43
ai
"8
s
co o oo
3
CM 00 IT) rH
00
\D CO
VO CO rH 00
O
■^ CM
CO rH
r-t
CM rH
CM O
CM rn'
2
c '
u
><•£
5 o
*
>>
(TJ
JH
*2 in
X o>
V-
RJ
CM
TJ
th(bc
eshy
spine,
TJ
e 5
H -a
(0
l-c
• r-i (TJ
C (J
C
0
'5!
*
wid
er(fl
th*
-fin:
engt
leng
o — ,
in
en u
=> .2
O TJ
C «
a, o
nal, II +
ectoral (le:
ateral line
5
in
(TJ
u
60
•c
o
c
X!
(8
Interorbital
Eye diamet
Snout lengl
First dorsal
*
X,
c
01
elvic fin, 1
longest s
pper jaw
0) 3
< CQ
O O
•j3 -a
<tj ns
CD CD
< Oh hJ
U
X
Ph 3
Di k
£ ho
60 en
0) 3
i- '"
1 2
ffl 3
rx c
to C
| $
-C 0)
in ^
"o gj
f g,
60 5T
C >H
a! 5
(TJ —
-a -a
® '5
c o
%£
. C en
hJ «3 *C
* > —
+3 *o3 "s
52 23 c
rS 60+-
g O 60
TJ ^^
$ " II
£<CO
X m "5
r KK
18
GREENFIELD & JOHNSON: BLENNIOID FISHES 19
(Springer, 1959a, 1959b; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968). This
constitutes the first record from Honduras.
Ecological data. — Labrisomus bucciferus was taken in collections in which sam-
pling extended to 15 ft, but most specimens were taken in less than 4 ft. Al-
though L. bucciferus was taken in a wide variety of habitats in Belize and Hon-
duras (table 9), most (71% of 94 specimens) were taken in shallow water (< 4 ft),
high wave energy, fore-reef areas (H2) or reef crest pools (H6). Labrisomus buc-
ciferus was among the seven blennioid species to be taken in areas dominated by
sea grass beds (H7). In Belize, all specimens of L. bucciferus were taken either at
Barrier Reef sites (BRT, CBC) or at the offshore atolls (GR, LR). Additional Belize
records include Turneffe Island and one Barrier Reef site (SWC). In Honduras L.
bucciferus was taken at both the Hog Islands and Roatan. Labrisomus bucciferus
was taken with all other species of Labrisomus except L. albigenys but shared
highest affinity (GMI) values with Malacoctenus gilli and Entomacrodus nigricans
(table 8).
Material examined. — A total of 95 specimens from 18 collections. HONDURAS:
CC, 84443 (4), 84446 (1); IR, 84442 (11), 84444 (1), 84445 (1). BELIZE: BRT, 89406
(21); CBC, 89281 (2), 89353 (1), 89393 (3); GR, 39826 (1), 70989 (1), 71109 (1), 71147
(1), 77601 (2), 86040 (34), 86174 (1), 89407 (5); LR, 89409 (4). Additional Belize
material examined: CBC, JWC-7 (2), JVVC-23 (3); LR, UMML 9977; SWC, JWC-9
(2), JWC-24 (1); TI, UMML 9861.
(3) Labrisomus gobio (Valenciennes, 1836)
This species is known from the Bahamas, Greater and Lesser Antilles, and
Central America, including Belize, Nicaragua, and off Yucatan (Springer, 1959a,
Table 8. Co-occurrence of species of Labrisomus. Values in upper half of the matrix
(whole numbers) are actual number of records (number of stations) of co-occurrence
(based on all stations). Values in the lower half of the matrix are calculated GMI values
(based on Class I stations only).
Co-occurring species
Species
NT
N,
1
2 3 4
5
6
7
8
1. albigenys
2
2
—
0 0 2
0
1
1
0
2. bucciferus
17
16
0
3 5
2
1
4
7
3. gobio
10
10
0
.11 — 10
2
0
6
1
4- zuppy*
29
29
.17
.14 .49 —
4
5
11
5
5. haitiensis
18
17
0
0 .03 .09
—
1
2
1
6. kalisherae
11
11
.06
0 0 .19
0
—
2
8
7. nigricinctus
13
13
.06
.15 .39 .47
.01
.03
—
5
8. nuchipinnis
16
16
0
.31 0 .14
0
.48
.22
—
Listing of species sharing highest GMI values with species
of Labrisomus (five highest
values only).
1. albigenys
27 (.24),
37 (.20),
4 (.17),
55 (.17), 15 (.16)
2. bucciferus
13 (.60),
51 (.34),
8 (.31),
14 (.27), 18 (.21)
3. gobio
4 (.49),
15 (.48),
27 (.47),
9 (.40), 7 (.39)
4. guppyi
35 (.70),
55 (.62),
15 (.61),
37 (.57), 3 (.49)
5. haitiensis
33 (.41),
10 (.30),
38 (.23),
22 (.20), 25 (.20)
6. kalisherae
27 (.67),
57 (.53),
8 (.48),
55 (.46), 15 (.45)
7. nigricinctus
9 (.58),
37 (.57),
51 (.51),
4 (.47), 35 (.42)
8. nuchipinnis
21 (.54),
57 (.54),
13 (.49),
6 (.48), 9 (.38)
NT = total number of stations; N, = number of Class I stations.
20 FIELDIANA: ZOOLOGY
1959b; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968). This constitutes the
first record for Honduras.
Ecological data. — Labrisomus gobio was taken in stations in which sampling ex-
tended to 15 ft, but most specimens were taken in less than 4 ft. All specimens of
L. gobio were taken in shallow-water coral reef habitats (H2, H3), with 94% (of 48
specimens) from H2 habitats (table 9). In Honduras L. gobio was taken at only
one station, a rich shallow-water fringing reef on Roatan. In Belize L. gobio was
taken only at Barrier Reef sites (BRG, CBC) or at the atolls (GR, LR). Additional
Belize records are from Turneffe Island.
Labrisomus gobio has been taken with all species of Labrisomus except L. al-
bigenys and L. kalisherae. Labrisomus gobio and L. kalisherae are very similar in
appearance but differ in distribution — L. gobio has been taken at offshore, coral-
rich habitats in Belize and Honduras, L. kalisherae is known in Belize only from
cays inside the Barrier Reef and in Honduras only from the Hog Islands. Lab-
risomus gobio shared highest affinity (GMI) values with Labrisomus guppyi and
Malacoctenus triangulatus (table 8).
Material examined. — A total of 48 specimens from 10 collections. HONDURAS:
IR, 84431 (2). BELIZE: BRG, 77605 (11), 77607 (2); CBC, 89295 (6), 89321 (2), 89380
(8); GR, 77603 (7), 77604 (1); LR, 77602 (4), 86041 (5). Additional Belize material
examined: CBC, JWC-8 (1), JWC-11 (1), JWC-16 (5), JWC-19 (1), JWC-23 (3).
Additional Belize material reported by Birdsong & Emery 1968: LR, UMML 9561,
UMML 9984; TI, UMML 9484, UMML 9836.
(4) Labrisomus guppyi (Norman, 1922)
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Fernando de Noronha (off Brazil), Venezuela, Curacao, and Central America,
including Belize, Nicaragua, Panama, and the Yucatan (Springer, 1959a, 1959b;
Cervigon, 1966; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968; Mago, 1970;
Nagelkerken, 1974). This constitutes the first record for Honduras.
Ecological data. — Labrisomus guppyi is the most commonly represented species
of Labrisomus in our collections from Belize and Honduras. It has been taken in
collections in which sampling extended to 20 ft but was most commonly taken in
depths of 10 ft or less (24 of 29 collections). Most specimens of L. guppyi (86% of
270 specimens) were taken in areas of rich coral (H2, H3), with the remaining
specimens (all taken at the Hog Islands) from areas of rocky ledges (H4) or two
sites with both rock and coral represented in exceptionally broad stations (field
numbers: F75-47, F75-51) and classified as H9 habitats (table 9). In Belize all
specimens of L. guppyi were taken at the Barrier Reef (BRG, BRT, CBC) or atolls
(GR, LR). Additional Belize records include Turneffe Island and one Barrier Reef
site (SWC). Labrisomus guppyi was taken with every other species of Labrisomus
represented in our collections but was taken with L. albigenys and L. kalisherae
only at the Hog Islands and with L. nuchipinnis only at the Hog Islands and two
Barrier Reef sites (BRT, CBC). In cases of co-occurrence L. guppyi was virtually
always more numerous than any species of Labrisomus taken with it (true for 36
of 42 comparisons); four of the six remaining comparisons were ties, and in only
two cases (once with L. bucciferus, once with L. gobio) was L. guppyi less numer-
ous than a co-occurring species of Labrisomus. Labrisomus guppyi shares highest
affinity (GMI) values with Enneanectes boehlkei and Ophioblennius atlanticus
macclurei (table 8).
0\ CM vO I
as ° ' '
8
C(N rin
I I I I I I I I I I I I -8
r^ <N CD
s H
I I I I I I I I I II
■'S
?, I I I I ' I I
SS ^S I I II I I *8 I I
2
S I I
I I Tf O (N <N •«* (N
rH rH CO
8 °"S
CTs vO ON SO
3
I I I I I I
I I I I II
S «
,_ NJ OCO J»p 00O H* rtn
i
a
en
:<c
G
C
.&
c
21
22 FIELDIANA: ZOOLOGY
Material examined. — A total of 270 specimens from 29 collections. HON-
DURAS: CC, 84461 (4), 84463 (11), 84464 (13), 84465 (2), 84466 (10), 84467 (13),
84468 (7); IR, 84462 (16), 84469 (15). BELIZE: BRG, 70921 (1), 77612 + 86047 (14),
86025 (1); BRT, 86046 (12); CBC, 89286 (14), 89307 (1), 89325 (8), 89330 (4), 89337
(3), 89351 (1), 89385 (3), 89396 (55); GR, 70943 (1), 77606 (6), 77609 (3), 77611 (1),
86045 (28), 86048 (1); LR, 77608 (8), 77610 (14). Additional Belize material exam-
ined: CBC, JWC-11(3), JWC-16 (5); LR, UMML 9409, UMML 9501, UMML 9998;
TI, UMML 9849, UMML 10319.
(5) Labrisomus haitiensis Beebe & Tee Van, 1928
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Gulf of Mexico (Florida), and Central America, including Lighthouse Reef, Be-
lize, and Banco Chinchorro off Yucatan (Springer, 1959a, 1959b; Birdsong &
Emery, 1968; Bohlke & Chaplin, 1968; Smith, 1976). This constitutes the first
record for Honduras.
Ecological data. — Labrisomus haitiensis has been taken at greater depths than any
other species of Labrisomus represented in our collections (table 6). It was col-
lected in depths as deep as 50 ft, but most records (12 of 18) and specimens (52 of
70) are from depths less than 20 ft, and six of the 18 collections containing this
species were in depths less than 6 ft. We have taken L. haitiensis only in coral-
rich offshore habitats either patch reefs, shallow fringing reefs, or the top of
marked dropoffs (table 9), with the majority of both records and specimens (56%
of 70 specimens) from patch reef sites (H3) in the lagoon at Glovers Reef. In
Belize L. haitiensis was taken only at Barrier Reef sites (BRG, CBC) and the atolls
(GR, LR), with the majority of both records (12 of 14) and specimens (49 of 52)
from atoll localities. Additional Belize records include Turneffe Island and one
Barrier Reef locality (SWC). In Honduras we have taken L. haitiensis at the Hog
Islands and at Roatan. Labrisomus haitiensis has been taken with all species of
Labrisomus except L. albigenys. It was taken with L. kalisherae and L. nuchipinnis
only at one exceptionally broad station (F75-51) at the Hog Islands. Labrisomus
haitiensis shares highest affinity (GMI) values with Enneanectes altivelis and
Malacoctenus boehlkei (table 8).
Material examined. — A total of 70 specimens from 18 collections. HONDURAS:
CC, 84440 (2), 84441 (6); IR, 84438 (2), 84439 (10). BELIZE: BRG, 77810 (1); CBC,
89292 (2); GR, 70988 + 77613 (2), 71100 (2), 71112 (2), 71122 (1), 71123 (2), 71142
(3), 71144 + 86042 (3), 77615 (2), 84553 (5), 86043 (4), 86044 (18); LR, 77614 (3).
Additional Belize material examined: LR, UMML 9412 (3), UMML 9492 (2).
(6) Labrisomus kalisherae (Jordan, 1904)
This species has been reported from Florida, Tobago, Isla Fernando de
Noronha (Brazil), Venezuela, Belize, Nicaragua, Panama, Campeche Banks
(Mexico), and Cabo Rojo (Veracruz, Mexico) (Springer, 1959a; Rubinoff &
Rubinoff, 1962; Caldwell & Caldwell, 1964; Cervigon, 1966; Birdsong & Emery,
1968; Mago, 1970). This constitutes the first record for Honduras.
Ecological data. — Labrisomus kalisherae was taken in stations in which sampling
extended to 15 ft, but the majority of both records (eight of 11) and specimens
(188 of 196) were taken in less than 5 ft. In Belize L. kalisherae was taken only at
cays inside the Barrier Reef and in habitats dominated by coarse coral rubble and
GREENFIELD & JOHNSON: BLENNIOID FISHES 23
extensive algal mats (H5). In Honduras L. kalisherae was collected only at the Hog
Islands in areas of rocky ledges with or without extensive live coral (H2, H4) and
in one exceptionally broad station (F75-51) classified as H9 (table 9). Labrisomus
kalisherae was taken with all species of Labrisomus represented in our collections
save the purely insular L. gobio (see above), but all records of co-occurrence other
than with L. nuchipinnis are from localities in the Hog Islands where a number of
exceptionally broad collecting stations, including habitats (at a single site) rang-
ing from rock and algae, to coral rubble and sand, to live coral, obscure the
possibility of habitat separation among the species collected. Labrisomus kalisherae
shares highest affinity (GMI) values with Starksia occidentalis and Scartella cristata
(table 8).
Material examined. — A total of 196 specimens from 11 collections. HON-
DURAS: CC, 84447 (7), 84448 (3), 84449 (2), 84450 (4), 84451 (5), 84452 (2), 84453
(1). BELIZE: BC, 86051 (16); FC, 86050 (22); FC, 86052 (22); TC, 86049 (112).
(7) Labrisomus nigricinctus Rivero, 1936
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Curacao, Los Roques (Venezuela), and Central America, including Courtown
Cays off Nicaragua, Panama, and Banco Chinchorro off Yucatan (Springer,
1959a, 1959b; Rubinoff & Rubinoff, 1962; Birdsong & Emery, 1968; Bohlke &
Chaplin, 1968; Cervigon, 1968; Nagelkerken, 1974). This constitutes the first
record for Belize and for Honduras.
Ecological data. — Labrisomus nigricinctus was taken in collections in which sam-
pling extended to 18 ft, but most records (11 of 13) and specimens (36 of 41) are
from stations in 6 ft of water or less. Although L. nigricinctus was taken in a
variety of habitats (table 9), the majority of both records (9 of 13) and specimens
(78% of 41) are from rich shallow- water reefs in relatively high wave energy sites
(H2). In Belize L. nigricinctus was taken only at Barrier Reef sites (BRG, BRT,
CBC) and at Glovers Reef. In Honduras L. nigricinctus was taken at Roatan and at
the Hog Islands. Labrisomus nigricinctus was taken with all other species of Lab-
risomus but was taken with L. kalisherae only at the Hog Islands. Labrisomus
nigricinctus shares highest affinity (GMI) values with Malacoctenus aurolineatus
and Enneanectes pectoralis (table 8).
Material examined. — A total of 41 specimens from 13 collections. HONDURAS:
CC, 84434 (3), 84436 (1), 84437 (4); IR, 84435 (1). BELIZE: BRG, 77616 (1), 86026
(1), 86055 (1); BRT, 86054 (4); CBC, 89293 (9), 89320 (1), 89379 (1); GR, 77617 (2),
86053 (3).
(8) Labrisomus nuchipinnis (Quoy & Gaimard, 1824)
This species is known from both sides of the Atlantic: eastern Atlantic,
Madeira Islands to Annobon Island; western Atlantic, Bermuda, the Bahamas,
Florida, Greater and Lesser Antilles, Brazil, Venezuela, Curacao, Colombia,
Central America, including Belize, Honduras, Nicaragua, and Panama; and Gulf
of Mexico, including Mexico and Texas (Fowler, 1944; Springer, 1959a, 1959b;
Birdsong & Emery, 1968; Bohlke & Chaplin, 1968; Nagelkerken, 1974). This
constitutes the first record for Honduras.
Ecological data. — All of our records for L. nuchipinnis are from stations in less
than 10 ft of water, and 14 of 16 records are from less than 5 ft. Although L.
24 FIELDIANA: ZOOLOGY
nuchipinnis was taken in a variety of shallow-water habitats (table 9), the great
majority of specimens (80% of 71) came from rocky areas (H4) in the Hog Islands
or from coarse coral rubble (H5) at cays at and inside of the Barrier Reef in Belize.
One record is from pier pilings at Trujillo. Labrisomus nuchipinnis is known from
both mainland and insular localities, in Belize from cays inside the Barrier Reef
(BC, SC, TC), from Barrier Reef sites (BRT, CBC), and from Glovers Reef; in
Honduras from mainland sites (Trujillo) and from the Hog Islands. Labrisomus
nuchipinnis was taken with all species of Labrisomus except L. albigenys, but was
taken with L. haitiensis only at one exceptionally broad (F75-51) station at the
Hog Islands. Labrisomus nuchipinnis shares highest affinity (GMI) values with
Paraclinus nigripinnis and Scartella cristata.
Material examined. — A total of 71 specimens from 16 collections. HONDURAS:
CC, 84455 (4), 84456 (1), 84457 (2), 84458 (3), 84459 (3), 84460 (14); Mainland, pier
at Trujillo, 84454 (4). BELIZE: BC, 86060 (4); BRT, 86056 (1); CBC, 89301 (3), 89355
+ 89413 (23), 89390 (1); GR, 86057 (3), 86058 (1); SC, 86061 (1); TC, 86059 (3).
Additional Belize material examined: SWC, JWC-13 (1), JWC-24 (1).
Malacoctenus Gill, 1860
The genus Malacoctenus contains 18 species, one species (M. africanus Cadenat)
in the eastern Atlantic, eight species in the western Atlantic, and nine species
(four of which have two recognized subspecies) in the eastern Pacific. The latest
revision of the genus is that of Springer (1959a).
All known western Atlantic species of Malacoctenus occur off the Central
American coast, and all except M. versicolor (Poey) are represented in our collec-
tions from Belize and Honduras. Malacoctenus versicolor, known from the
Bahamas, Florida, Greater and Lesser Antilles, was reported from Belize
(UMML 9996 (1), Lighthouse Reef) by Birdsong & Emery (1968). Co-occurrence
data for species of Malacoctenus is presented in Table 10.
(9) Malacoctenus aurolineatus Smith, 1957
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Venezuela, and Central America, including Yucatan (Springer, 1959a, 1959b;
Birdsong & Emery, 1968; Bohlke & Chaplin, 1968; Cervigon, 1968). This consti-
tutes the first record for Belize and for Honduras.
Ecological data. — Malacoctenus aurolineatus has been taken in stations in which
sampling extended to 15 ft, but most records (11 of 12) and specimens (171 of
173) are from less than 10 ft, and the majority from less than 6 ft. Except for one
specimen taken in a reef-crest pool (H6) at Glovers Reef, all records and speci-
mens of M. aurolineatus (table 11) are from shallow, rich reef habitats (H2) in
Belize; shallow rocky habitats (H4) in Honduras; or, in the case of one exception-
ally broad station (F75-51) in Honduras, both rock and coral (H9). In Belize M.
aurolineatus was taken only from Barrier Reef sites (BRG, BRT, CBC) and the
atolls (GR, LR), with most specimens (27 of 32) from the Barrier Reef. In Hon-
duras M. aurolineatus was taken only at the Hog Islands. Malacoctenus aurolineatus
was taken with all species of Malacoctenus (table 10) except M. boehlkei (appar-
ently a generally deeper-living species) and M. delalandei (most specimens of
which were taken at more inshore localities). Malacoctenus aurolineatus shares
GREENFIELD & JOHNSON: BLENNIOID FISHES 25
Table 10. Co-occurrence of species of Malacoctcnus. Values in upper half of matrix
(whole numbers) are actual number of records (number of stations) of co-occurrence
(based on all stations). Values in the lower half of matrix are calculated GMI values (based
on Class I stations only).
Species NT N, 9 10 11 12 13 14 15
9. auroliiwatus
12 12 0 0 8
5
4
10
10. boehlkei
12 12 0 0 0
0
2
1
11. delalandei
4 4 0 0 0
2
2
2
12. erdmani
9 9 .62 0 0
5
5
8
13. gilli
21 21 .21 0 .11 .26
—
8
9
14. macropus
21 21 .14 .02 .11 .26
.27
—
10
15. triangulatus
32 31 .43 0 .09 .39
.26
.30
—
Listing of species sharing highest GMI values with species of Malacoctenus (five highest
values only).
9. auroliiwatus
12 (.62), 7 (.58), 51 (.53), 37 (.47), 4 (.44)
10. boehlkei
5 (.30), 20 (.26), 43 (.25), 25 (.20), 50 (.19)
11. delalaiidei
21 (.27), 8 (.25), 55 (.18), 6 (.15), 32 (.14)
12. erdmani
9 (.62), 57 (.46), 7 (.42), 49 (.41), 37 (.41)
13. gilli
2 (.60), 8 (.49), 51 (.43), 21 (.42), 6 (.29)
14. macropus
21 (.42), 6 (.35), 8 (.33), 57 (.30), 15 (.30)
15. triangulatus
4 (.61), 35 (.58), 55 (.57), 37 (.51), 27 (.49)
NT = total number of stations; N| = number of Class I stations.
highest affinity (GMI) values with Malacoctenus erdmani and Labrisomus nigri-
cinctus.
Material examined. — A total of 173 specimens from 12 collections. HON-
DURAS: CC, 84406 (29), 84407 (20), 84408 (44), 84486 (6). BELIZE: BRG, 77626
(1), 77633 + 86068 (2); BRT, 85987 (15); CBC, 89283 (28), 89378 (22); GR, 85988 (1),
85989 (2); LR, 77621 (3).
(10) Malacoctenus boehlkei Springer, 1959
This species has been reported from the Bahamas, Virgin Islands, Albuquer-
que Cays off Nicaragua, and Belize (Springer, 1959a, 1959b; Birdsong & Emery,
1968; Bohlke & Chaplin, 1968). This constitutes the first record for Honduras.
Ecological data. — Malacoctenus boehlkei is the deepest-living species of Malacoc-
tenus represented in our material. Our depth records for this species include
stations at 90 to 100 ft, and six of the 12 stations at which this species was taken
were in depths exceeding 50 ft; however, most specimens of this species (39 of
57, from five collections) were taken at depths of 15 ft or less. Two habitats,
fore-reef dropoff areas (HI) and patch reefs (H3), accounted for virtually all (98%
of 57 specimens) material of M. boehlkei (table 11). In Belize this species was taken
at the Barrier Reef (CBC) and the atolls (GR, LR) and has been reported from
Turneffe Island. One apparently anomalous record (G74-13, one specimen) is
from Frenchman's Cay, inside the Barrier Reef, in an area of coarse coral rubble
and sand (H5). In Honduras we have only one record of this species, seven
specimens taken in 50 ft (F75-48) at the Hog Islands. Examination of co-
occurrence data (table 10) suggests that M. boehlkei exhibits little habitat overlap
with other species of Malacoctenus. Malacoctenus boehlkei was taken with M. mac-
ropus on two occasions, both stations on patch reefs in the lagoon of Glovers
Reef. Malacoctenus boehlkei was taken with M. triangulatus only at the apparently
fc
ON <N . .
IN . Tf . CO
•o
S
I I I I ~ffi I I I I I I I I
o I cr> o o
*.*
S I I II I I "fc ^ I I
3
O rH O
« W
£ I I ^$5
g ^ Qj | | | l^s^g^ssfc
fl I I *-8 I I I I I I I I
__ NtO (NN *N OvO H^ H\D M^
JJ r-i K r-t IT) <N ^ Nlft wi nin
i
11
M
26
GREENFIELD & JOHNSON: BLENNIOID FISHES 27
anomalous (for M. boehlkei) station at Frenchman's Cay. Malacoctenus boehlkei
shares largest affinity (GMI) values with Labrisomus haitiensis, Paraclinus mar-
moratus (a species taken at only two stations), and Emblemaria caldivelli (table 11).
Material examined. — A total of 57 specimens from 12 collections. HONDURAS:
CC, 84405 (7). BELIZE: CBC, 89311 (1); FC, 86023 (1); GR, 71084 (1), 71128 (1),
71133 (1), 71138 (1), 85981 (22), 85982 (6), 85983 + 89404 (3), 86164 (12); LR, 77622
(1). Additional Belize material examined: TI, UMML 9865 (1).
(11) Malacoctenus delalandei (Valenciennes, 1836)
This species has been reported from Puerto Rico, Tobago, Natal (Brazil), Ven-
ezuela, Panama, Guatemala, and Belize (Springer, 1959a; Birdsong & Emery,
1968). This constitutes the first record for Honduras.
Ecological data. — Malacoctenus delalandei was taken by us on only four occa-
sions, twice in Belize and twice in Honduras. The Belize records are from cays
inside the Barrier Reef (BC, TC) in waters less than 4 ft, over coral rubble and
algae (H5). Additional Belize records are from the mainland. In Honduras, M.
delalandei was collected at Roatan in less than 5 ft of water, in an area of sand and
Thalassia adjacent to a pier (H7), and was collected on pier pilings at Trujillo in
less than 5 ft of water (H8). Malacoctenus delalandei is apparently absent in areas of
rich coral development. Malacoctenus delalandei was taken with M. gilli (IR, BC),
M. macropus (IR, TC), and M. triangulatus (IR, BC) (table 10). Malacoctenus delalan-
dei shares highest affinity (GMI) values with Paraclinus nigripinnis and Labrisomus
nuchipinnis.
Material examined. — A total of 22 specimens from four collections. HON-
DURAS: IR, 84403 (1); Mainland, pier at Trujillo, 84404 (12). BELIZE: BC, 85984
(2); TC, 85985 (7). Additional Belize material examined: Mainland, 1.5 miles
south-southwest of Belize City, UMML 9609; 2.5 miles north of Belize City at
southeast corner of Peter's Bluff, UMML 9787.
(12) Malacoctenus erdmani Smith, 1957
This species is known from the Bahamas, Greater and Lesser Antilles, and
Central America, including Lighthouse Reef off Belize, Albuquerque Cays off
Nicaragua, and Banco Chinchorro off Yucatan (Springer, 1959a, 1959b; Birdsong
& Emery, 1968; Bohlke & Chaplin, 1968). This constitutes the first record for
Honduras.
Ecological data. — All of our records for this species were from 10 ft of water or
less, and most records (seven of nine) and specimens (37 of 40) were from
stations in 6 ft of water or less. All specimens were taken in either rich reef
formations in shallow water (H2, all Belize records) or in areas of rocky ledges
(H4, Honduras records) except for one record from an exceptionally broad sta-
tion (F75-51) at the Hog Islands that included both rock formations and rich coral
formations (H9; see table 11). In Belize M. erdmani was taken only at Barrier Reef
sites (BRG, BRT, CBC) and was never taken at Glovers Reef or Lighthouse Reef.
Additional Belize material includes one specimen from Turneffe Island. Because
we have taken M. erdmani on only four occasions in Belize, the apparent absence
of this species at the offshore atolls (GR, LR) may well reflect inadequacies in
sampling effort. In Honduras M. erdmani was taken only at the Hog Islands.
28 FIELDIANA: ZOOLOGY
Malacoctenus erdmani was taken with all species of Malacoctenus except M. boehlkei
and M. delalandei. Malacoctenus erdmani shares highest affinity (GMI) values with
Malacoctenus aurolineatus and Scartella cristata.
Material examined. — A total of 40 specimens from nine collections. HON-
DURAS: CC, 84409 (1), 84410 (2), 84411 (9), 84412 (8), 84484 (2). BELIZE: BRG,
77625 (1), 89411 (1); BRT, 85990 (14); CBC, 89285 (2). Additional Belize material
examined: LR, UMML 9467 (1).
r
(13) Malacoctenus gilli (Steindachner, 1867)
This species is known from the Bahamas, Greater and Lesser Antilles, Ven-
ezuela, Curacao, Gulf of Mexico off Tampico (Mexico), and Central America,
including Belize, Honduras, Nicaragua, and Yucatan (Springer, 1959a, 1959b;
Cervigon, 1966; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968; Mago, 1970;
Nagelkerken, 1974).
Ecological data. — All specimens of Malacoctenus gilli were taken in less than 5 ft
of water. Malacoctenus gilli was taken in a wide variety of habitat categories (table
11), but two categories, H6, reef crest pools, and H7, sea grass beds, together
accounted for 69% of the 154 specimens of M. gilli represented in our collections.
Habitat categories H6 and H7 were represented by only seven stations and five
stations, respectively, of the 127 total. In Belize M. gilli was taken at cays inside
the Barrier Reef (BC, SC), at Barrier Reef sites (BRT, CBC), and at the atolls
(GR, LR). Additional records from Belize include Turneffe Island and two Barrier
Reef sites (CBC, SWC). Records from Honduras are from the Hog Islands and
from Roatan. Malacoctenus gilli has been taken with all species of Malacoctenus
represented in our collections except M. boehlkei, from which it apparently differs
substantially in both depth distribution and habitat preference (see above).
Malacoctenus gilli shares highest affinity (GMI) values with Labrisomus bucciferus
and Labrisomus nuchivinnis (table 10).
Material examined. — A total of 156 specimens from 23 collections. HON-
DURAS: CC, 84413 (5), 84416 (16), 84417 (2), 84418 (2), 84485 (20); IR, 39843 (1),
84414 (33), 84415 (16). BELIZE: BC, 85991 (4); BRT, 85996 (2); CBC, 89297 (3),
89300 (1), 89309 (1), 89354 (3); GR, 39822 (1), 71110 (9), 71124 (4), 77623 (1), 85992
(25), 85993 (1), 85994 (3); LR, 77624 (1); SC, 85995 (2). Additional Belize material
examined: CBC, JWC-12 (1), JWC-16 (1), JWC-21 (1); SWC, JWC-19 (4), JWC-20
(5), JWC-24 (2), JWC-25 (1); TI, UMML 9860.
(14) Malacoctenus macropus (Poey, 1868)
This species is known from Bermuda, the Bahamas, Florida, Greater and
Lesser Antilles, Venezuela, and Central America, including Belize, Nicaragua,
Panama, and Yucatan (Springer, 1959a; Caldwell & Caldwell, 1964; Birdsong &
Emery, 1968; Bohlke & Chaplin, 1968; Cervigon, 1968). This constitutes the first
record for Honduras.
Ecological data. — Malacoctenus macropus was taken in collections in which sam-
pling extended to 15 ft, but most records (17 of 21) and specimens (142 of 166) are
from less than 6 ft. Malacoctenus macropus was taken in a wide variety of
shallow- water habitats (table 11). The geographic extent of localities at which
this species was represented was also broad, excluding only mainland sites. In
GREENFIELD & JOHNSON: BLENNIOID FISHES 29
Belize this species was taken at cays inside of the Barrier Reef (SC, TC), at Barrier
Reef sites (BRT; with additional records from CBC and SWC), and the atolls (GR,
LR, with an additional record from TI). In Honduras M. macropus was taken at
the Hog Islands and at Roatan. Malacoctenus macropus was taken with every other
species of Malacoctenus represented in our collections. Malacoctenus macropus
shares highest affinity (GMI) values with Paraclinus nigripinnis and Labrisomus
kalisherae (table 10).
Material examined. — A total of 166 specimens from 21 collections. HON-
DURAS: CC, 84420 (28), 84421 (1), 84422 (5), 84423 (14), 84424 (49), 84428 (15); IR,
84419 (2), 84425 (2), 84426 (2), 84427 (6), 84429 (1), 84430 (3). BELIZE: BRT, 86002
(4); GR, 71129 + 86001 (3), 77618 (1), 77619 (2), 77620 (1), 86004 (3), 86006 (1); SC,
86003 (8); TC, 86005 (16). Additional Belize material examined: CBC, JWC-3 (2),
JWC-12 (25); LR, UMML 9986; SWC, JWC-9 (4), JWC-13 (8), JWC-20 (2), JWC-25
(1); TI, UMML 9800.
(15) Malacoctenus triangulatus Springer, 1959
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Fernando de Noronha (Brazil), Venezuela, Curacao, Colombia, Central America,
including Belize, Nicaragua, Panama, and Yucatan, and Gulf of Mexico at Cabo
Roja, Veracruz (Springer, 1959a, 1959b; Birdsong & Emery, 1968; Bohlke &
Chaplin, 1968; Nagelkerken, 1974; Springer & Gomon, 1975). This constitutes
the first record for Honduras.
Ecological data. — Malacoctenus triangulatus is represented in our collections at
more stations than any other blennioid species. It is represented by more speci-
mens than any other species except Labrisomus guppyi, Entomacrodus nigricans,
and Ophioblennius atlanticus macclurei. Malacoctenus triangulatus was taken at one
station (F75-15) in which sampling was conducted between 45 and 60 ft. All
other records are from less than 20 ft, and most records (27 of 32) and specimens
(211 of 254) are from stations in less than 10 ft of water. Malacoctenus triangulatus
was taken in a wide variety of habitats (table 11), but the majority of records (25
of 32) and specimens (95% of 254) are from rich shallow-water fringing reefs or
patch reefs (H2, H3), areas of rocky ledges (H4), or from two exceptionally broad
Hog Island stations (F75-47, F75-51) with both rock and coral formations
present (H9). The geographic extent of localities at which M. triangulatus was
taken is equally broad, excluding only the mainland stations. In Belize M. trian-
gulatus was taken at cays behind the Barrier Reef (BC, FC, SC), Barrier Reef sites
(BRG, CBC, with additional records from SWC), and the atolls (GR, LR, with
additional records from Turneffe Island). Very few records (three of 18) and
specimens (four of 85) are from cays inside of the Barrier Reef. In Honduras M.
triangulatus was taken at Roatan and the Hog Islands. Malacoctenus triangulatus
was taken with all other species of Malacoctenus represented in our collections.
Malacoctenus triangulatus shares highest affinity (GMI) values with Labrisomus
guppyi and Enneanectes boehlkei.
Material examined. — A total of 254 specimens from 32 collections. HON-
DURAS: CC, 84391 (18), 84392 (37), 84393 (30), 84394 (10), 84395 (1), 84397 (20),
84398 (1), 84399 (13), 84487 (3); IR, 84390 (18), 84396 (2), 84400 (11), 84401 (3),
84402 (2). BELIZE: BC, 85999 (2); BRB, 86000 (1); BRG, 70922 (1), 86065 (15); CBC,
89299 (16), 89319 (9), 89331 (1), 89338 (2), 89366 (4), 89392 (1); FC, 85998 (1); GR,
30 FIELDIANA: ZOOLOGY
85997 (2), 86062 (16), 86063 (1), 86064 (1); LR, 86066 (1), 86067 (10); SC, 86024 (1).
Additional Belize material examined: LR, UMML 9268, UMML 9543; SWC,
JWC-24 (2); TI, UMML 9472, UMML 9824, UMML 10314.
Paraclinus Macquard, 1889
The genus Paraclinus contains 19 species, eight in the western Atlantic and 11
in the eastern Pacific. The latest revision of eastern Pacific species is that of
Rosenblatt & Parr (1969), and the latest revision of western Atlantic species, that
of Springer (1955a). Two additional western Atlantic species were described by
Bohlke (1960).
Six of the eight known western Atlantic species of Paraclinus occur off the
Central American coast. Only P. grandicomis (Rosen) [known from the Bahamas,
Florida, and south through the Lesser Antilles (Bohlke & Chaplin, 1968)] and P.
naeorhegmis Bohlke (known only from the Bahamas) are not represented in our
collections from Belize and Honduras. Co-occurrence data for species of Parac-
linus is presented in Table 12.
(16) Paraclinus barbatus Springer, 1955
This species has been reported from the Bahamas, Virgin Islands, and Santa
Marta, Colombia (Springer, 1955a; Bohlke & Chaplin, 1968; Palacio, 1974). The
specimens listed below are the first documented record for Belize and represent
a significant range extension for the species.
Paraclinus barbatus was described by Springer (1955a) from a single (28.3 mm
SL) specimen collected at St. Thomas, Virgin Islands, in 1919. Bohlke & Chaplin
(1968) record a second specimen taken in the Bahamas in 1966 and present an
illustration of the Bahaman specimen. A third specimen (UMML 30301) is re-
ported by Palacio (1974) from Colombia. The two specimens recorded below
from Belize thus represent the fourth and fifth known specimens of P. barbatus.
Table 12. Co-occurrence of species of Paraclinus. Values in upper half of matrix (whole
numbers) are actual number of records (number of stations) of co-occurrence (based on all
stations). Values in the lower half of the matrix are calculated GMI values (based on Class I
stations only).
Co-occurring species
Species NT N, 16 17 18 19 20 21
16. barbatus 1 1— 0 0000
17. cingulatus 1 10— 0001
18. fasciatus 550 0—003
19. infrons 660 0 0—00
20. marmoratus 220 0 00—0
21. nigripinnis 14 14 0 .13 .22 0 0 —
Listing of species sharing highest GMI values with species of Paraclinus (five highest
values only).
16. barbatus 10 (.14), 34 (.12), 26 (.11), 50 (.10), —
17. cingulatus 31 (.50), 49 (.29), 47 (.25), 53 (.22), 57 (.19)
18. fasciatus 21 (.22), 2 (.21), 14 (.18), 13 (.18), —
19. infrons 43 (.41), 50 (.40), 26 (.34), 48 (.20), 34 (.18)
20. marmoratus 10 (.26), 14 (.20), 32 (.06), 39 (.05), 2 (.05)
21. nigripinnis 8 (.54), 27 (.44), 6 (.43), 14 (.42), 13 (.42)
NT = total number of stations; Nr = number of Class I stations.
GREENFIELD & JOHNSON: BLENNIOID FISHES 31
These two specimens [FMNH 71085, original field number G 70-127, 2 (39.1 to
46.0)] were collected at Glovers Reef in December, 1970, at a depth of 70 to 100 ft.
Five additional specimens were taken at Glovers Reef in June, 1971, at depths of
60 to 100 ft (original field number G 71-25), but the preservative used failed, and
the material was lost. In the following paragraphs we present a description of
the two Belize specimens.
Meristic characters. — In the following listing values for the two Belize speci-
mens (39.1, 46.0), Palacio's (1974) (31 mm) specimen, and the holotype (28.3
mm, Springer, 1955a) are given in that order. Dorsal-fin elements: XXIX, 1;
XXIX, 1; XXVIII, 1; XXVIII, 1. Anal-fin elements: II, 19; II, 20; II, 19; II, 19.
Pectoral-fin rays: 13 (all). Pelvic-fin elements: I, 3 (all). Branchiostegal rays: 6
(all). Caudal-fin rays: 13, 13, 12, 12. Lateral line scales: — , — , 37, 37 (the number
of lateral line scales could not be determined in the Belize specimens).
Morphometric characters. — The following morphometry characters, measured
according to the methods of Springer (1955a), are listed as thousandths of the SL
and are based on the 46.0-mm Belize specimen and the 28.3-mm holotype
[values in parentheses, data from Springer (1955a)]. Unfortunately, the 39.1-mm
Belize specimen was deformed during preservation, and values of morphomet-
ric characters are not presented for this specimen. Tip of snout to tip of opercular
spine, 274 (297). Tip of snout to anterior bony margin of orbit, 76 (78). Horizontal
bony diameter of orbit, 83 (88). Interorbital width, 48 (49). Tip of snout to
greatest expansion of maxillary, 130 (141). Tip of snout to dorsal-fin origin, 207
(212). Tip of lower jaw to anal-fin origin, 513 (495). Preopercular length, 54 (67).
Distance from pelvic- fin insertion to anal-fin origin, 309 (283). Length of first
dorsal-fin spine, 85 (92). Length of second dorsal-fin spine, 91 (106). Length of
third dorsal-fin spine, 87 (106). Length of fourth dorsal-fin spine, 87 (106).
Length of 22nd dorsal-fin spine, 143 (148). Length of penultimate dorsal-fin
spine, 120 (117). Length of last dorsal-fin spine, 98 (106). Distance between base
of third and base of fourth dorsal-fin spines, 70 (78). Length of dorsal-fin base,
837 (777). Length of anal-fin base, 483 (495). Length of first anal-fin spine, 70
(71). Length of second anal-fin spine, 89 (106). Longest pectoral-fin ray, 276
(247). Length of longest pelvic-fin ray, 243 (247). Least depth of caudal peduncle,
98 (88). Greatest depth of body, 302 (276).
Description. — The following abbreviated description is based on the two Belize
specimens. A fleshy barbel extending ventrally from tips of lower jaw, length of
barbel slightly exceeding horizontal diameter of eye. Lips thick and fleshy. Nasal
cirrus simple and filamentous. Supraorbital and nuchal cirri paddle-shaped with
shallow crescent-shaped lobes on distal margin. A strongly developed but sim-
ple opercular spine extending posterior to a vertical through base of third
dorsal-fin spine. Lateral line scales apparently arranged as figured and described
by Springer (1955a) for holotype, but an accurate lateral line scale count cannot
be obtained for the two Belize specimens due to damage to the specimens.
Color in alcohol. — Body uniformly tan. Head with dark brown blotches of pig-
ment on snout, over ventral and posterior infraorbital bones, on cheek, and over
preopercle. Barbel and nasal cirri lacking pigment, but a dark blotch of pigment
present distally on supraorbital and nuchal cirri. Three dark, irregular brown
bars present on each pectoral and pelvic fin, and four dark brown bars present
32 FIELDIANA: ZOOLOGY
on caudal fin (the middle two bars set closely together). Dorsal fin lacking pig-
ment except for three very prominent dark ocelli, each subequal in size to eye
diameter. Dorsal ocelli appearing over 16th (15th) to 18th, 22nd to 24th, and 28th
to 29th dorsal-fin spines. Anal fin with dense dark brown pigmentation over
rays and membranes except for distal ca. 20% of length of anal-fin rays, which
are clear.
Color in life. — The first description of coloration in freshly captured specimens
follows. Body burnt-orange red, a white blotch running posteroventrally from
pectoral axil to ventral surface. Head darker, grading into purple-red anteriorly.
White blotch on posteroventral edge of opercle, running across branchiostegal
membrane onto ventral surface of body. A purple-red bar running from ventral
edge of orbit to isthmus, with areas anterior and posterior to bar white. Interor-
bital, snout, upper and lower lips purple-red, appearing as a bar running an-
teroventrally from orbit. Iris of eye red-orange with several black lines radiating
from black pupil. Pectoral and pelvic fins banded with alternating black and
yellow bars. Dorsal fin colored as body on basal two-thirds, distal one-third
white, with occasional scattered white blotches on dorsal membranes. Posterior
half of dorsal with three black ocelli, surrounded by concentric yellow and dark
red rings. Basal one-half of anal fin black, distal one-half white, ends of anal-fin
rays light brown to black. Basal portion of caudal fin white, followed by a
narrow, vertical band, posterior three-fourths of caudal fin with three indistinct
black vertical bars (lighter than first band), distal portion of caudal fin white.
Ecological data. — Paraclinus barbatus has been taken by us on only two occa-
sions, only at Glovers Reef, in depths exceeding 60 ft, in areas of rich coral
development in the reef-front dropoff zone (HI). This species was not taken
with any other species of Paraclinus.
Material examined. — Two specimens from one collection. BELIZE: GR, 71085
(2).
(17) Paraclinus cingulatus (Evermann & Marsh, 1900)
This species has been reported from the Tortugas, the Bahamas, Cuba, and
Puerto Rico (Springer, 1955a; Bohlke & Chaplin, 1968). This constitutes the first
record for Honduras and a significant range extension for the species.
The single (17.0 mm) specimen of P. cingulatus represented in our material was
taken 21 May 1975 (original field number FMNH 75-51) in less than 10 ft of water
on the west side of Big Hog Island, Honduras. The specimen was collected in an
area of rock, coral rubble, and sand inshore, with small but rich reef formations
further offshore. Twenty-three specimens of P. nigripinnis were taken at the
same station. The specimen agrees very well with the description of P. cingulatus
provided by Springer (1955a, p 437). The following counts are based on the
Honduras specimen: dorsal-fin elements, XXVII, 0. Anal-fin elements, II, 15.
Pectoral-fin rays, 12. Pelvic-fin elements, I, 2. Caudal-fin rays, 13. Branchio-
stegal rays, 6. Lateral line scales, 30.
Color in life. — Taken from a 35-mm Kodachrome. Body orange, crossed by four
black bars. First bar runs from anus onto first three spines posterior to notch in
the dorsal fin, where it becomes reddish brown; anterior to this bar, body white
below midline and orange above. Second bar extends from first four elements of
GREENFIELD & JOHNSON: BLENNIOID FISHES 33
anal fin up onto basal half of dorsal fin, retaining its black color. Third bar runs
from sixth through eighth anal elements up onto dorsal fin and nearly to its
margin; here it merges with uppermost portion of second bar. Fourth bar well
separated from and less intense in color than other three, extending from last
five elements of anal fin up onto last seven elements of dorsal fin, where it
becomes reddish brown in color. Spines of first portion of dorsal fin orange,
crossed by two or three black bands; second portion of dorsal fin orange, mar-
gined with white, and crossed by bars of the body. First half of anal fin black,
except for small orange patch at base of elements between second and third body
bars; posterior half orange, grading posteriorly into brown and then black; entire
margin of anal fin clear and expanded posteriorly to include distal halves of last
four elements and membranes between them. Orange color of body ends ab-
ruptly at caudal-fin peduncle, with white fin base bordered posteriorly by dark
brown line; fin rays also white, crossed with five or six brown bars and mem-
branes between them clear. Pectoral fins white with scattered brown bars on
rays similar to those on caudal fin. Pelvic fin black basally and white distally.
Head mottled with black and rose, prominent black bar running from lower
margin of eye posteriorly along upper jaw to behind maxillary; lips, maxillary,
and iris rose; opercular membranes orange.
Material examined. — One specimen from one collection. HONDURAS: CC,
84480 (1).
(18) Paraclinus fasciatus (Steindachner, 1876)
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Venezuela, Colombia, and Central America, including Belize, Guatemala, and
Panama (Springer, 1955a; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968).
This constitutes the first record for Honduras.
Ecological data. — We have taken P. fasciatus on only five occasions, always in
less than 4 ft of water. All five stations were in or adjacent to sea grass beds, with
22 of the 27 total specimens (table 13) from two very shallow stations in extensive
sea grass beds at Port Royal, Roatan. In no case was P. fasciatus taken in or near
extensive live coral. The Belize collections containing this species were both
mainland sites (Placentia and near Belize City). In Honduras P. fasciatus was
taken at Roatan and the Hog Islands. In Belize P. fasciatus was taken with no
other blennioid species. In Honduras P. fasciatus was taken with Labrisomus
bucciferus, Malacoctenus gilli, M. macropus, and Paraclinus nigripinnis in all three
collections in which P. fasciatus was represented.
Material examined. — A total of 27 specimens from five collections. HON-
DURAS: CC, 84483 (1); IR, 84481 (5), 84482 (17). BELIZE: P, 86030 (3); mainland,
beach at St. John's College near Belize City, 86029 (1). Additional Belize material
examined: mainland, 2.4 km south -south west of Belize City, UMML 9607.
(19) Paraclinus infrons Bohlke, 1960
This species has previously been reported only from the Bahamas (Bohlke,
1960; Bohlke & Chaplin, 1968). This constitutes the first record for Belize and a
significant range extension for the species. Paraclinus infrons was described from
four (12.6 to 18.8) specimens, all from or from near Grand Bahama Island.
Bohlke & Chaplin (1968) mention but do not list additional material from Grand
8
I I I I I I I I I I I I
t» nh mm
I I I I I I I I I I
a hsns i i ii ii i i Ns
g | , ^ | , | , | | «3«ig
a II l l 38 rt3 WS l l mPJ
CM I I NNOOninNrHNIflOO'Ji't
j I I O Tt r-< rH CM CM
9) P^
£ II | | ^8 38 SK ^S | |
ZinN -*rH no 'S'cq ot^ ^cn ro in
N HN NN NO NIT) Hid Ht
r-t i-H
II
g.
I
5
s
•■c
c
•S
8
6
|
j3
1
c
1
vi
vi
vi
vj
a
uS
vo
K
CM
cm
CM
;3 .3
34
GREENFIELD & JOHNSON: BLENNIOID FISHES
35
Bahama Bank. Our material of P. infrons, seven (13.4 to 21.4) specimens from six
collections, was collected either at Glovers Reef (three collections, four speci-
mens) between 50 and 90 ft, near the top of the dropoff for each site, or the
Barrier Reef at Carrie Bow Cay (three collections, three specimens) between 70
and 80 ft (in each case) in the dropoff zone. All six sites were areas of rich coral
development (HI). Paraclinus infrons was taken with no other species of Parac-
linus (table 12) and, with Paraclinus barbatus, is the deepest-living species of
Paraclinus represented in our collections. Paraclinus infrons shares highest affinity
(GNU) values with Emblemaria caldwelli and Lucayablennius zingaro. We present
(table 14) meristic and morphometric data for the Belize material.
Material examined. — A total of seven specimens from six collections. BELIZE:
CBC, 89332 (1), 89359 (1), 89414 (1); GR, 71116 (1), 86031 (1), 86032 (2).
Table 14. Meristic and morphometric data for Paraclinus infrons from Belize.
Character Carrie Bow Cay Glovers Reef Bahamas*
Number of specimens 3 4 4
Range in SL (mm)
13.4-21.0
16.0-21.4
12.6-18.8
Spinous dorsal
26-27
27-28
26-27
Soft dorsal
1 (all)
Kail)
Kail)
Anal, H +
18 (all)
18 (all)
17-18
Pectoral
12 (all)
12 (all)
12 (all)
Pelvic
1,3 (all)
1,3 (all)
1,3 (all)
Caudal
13 (all)
13 (all)
12-13
Body depth at dorsal origin
224-233
219-221
213-231
Body depth at anal origin
216-243
231-243
214-227
Caudal peduncle depth
104-114
94-103
98-110
First dorsal-fin spine
length
210-224
187**
178-194
Third dorsal-fin spine
length
67-71
65**
—
Pectoral-fin length
306-310
257-300
284-294
Pelvic-fin length
305-313
300-319
317-339
Head length
376-3%
336-388
372-405
Horizontal eye diameter
104-105
85-94
85-95
Upper jaw length
157-162
150-156
154-167
Snout length
105-112
103-112
110-119
All measurements expressed as thousandths of the SL. Morphometric data based on two
(89414, 13.4; 89359, 21.0) specimens from Carrie Bow Cay and three (71116, 19.0; 86031,
21.4; 86032, 16.0) specimens from Glovers Reef.
*Data from Bohlke (1960).
**Based on one specimen (86031, 21.4).
(20) Paraclinus marmoratus (Steindachner, 1876)
This species has been reported from the Bahamas, Florida, Cuba, and Ven-
ezuela (Springer, 1955a; Bohlke & Chaplin, 1968, Cervigon, 1968). This consti-
tutes the first record for Belize and a significant range extension for the species.
Springer (1955a) based his description of P. marmoratus on 157 (22.5 to 63)
specimens, all but one from Florida. Bohlke & Chaplin (1968) report three speci-
mens from the Bahamas, two of which were said to have been captured on small
patch reefs in 17 to 19 ft of water. The two Belize specimens were collected on
36 FIELDIANA: ZOOLOGY
small patch reefs in the lagoon near Long Cay, Glovers Reef, in less than 15 ft of
water (H3). No other species of Paraclinus was represented in these two collec-
tions. In general the Belize specimens agree well with Springer's (1955a) descrip-
tion of P. marmoratus. We present values for meristic characters for the two Belize
specimens (FMNH 86027, 29.1 mm; FMNH 86028, 24.8 mm) listed in that order:
dorsal-fin elements, XXVIII, 1; XXVIII, 1; anal-fin elements, II, 16 (correct as
given but a noticeable gap between the bases of the 14th and 15th anal-fin rays
suggests that the abnormally low count is due to injury or teratology); II, 19;
pectoral-fin rays, 13 (both); pelvic-fin elements, I, 3 (both); caudal-fin rays, 13
(both); lateral line scales, ? 35, ? 34.
Material examined. — Two specimens from two collections. BELIZE: GR, 86027
(1), 86028 (1).
(21) Paraclinus nigripinnis (Steindachner, 1867)
This species is known from Bermuda, the Bahamas, Florida, Greater and
Lesser Antilles, Brazil, Venezuela, and Central America, including Belize, Costa
Rica, and Nicaragua (Springer, 1955a; Caldwell, 1963; Birdsong & Emery, 1968;
Bohlke & Chaplin, 1968). This constitutes the first record for Honduras.
Ecological data. — All stations at which this species was taken were in less than
10 ft of water, and the majority of both records (10 of 14) and specimens (70 of
121) are from less than 4 ft. Paraclinus nigripinnis was taken in a variety of habitats
(table 13), but the majority of both records (8 of 14) and specimens (70% of 121)
were from shallow rocky sites at the Hog Islands (H4), coarse coral rubble and
algae at cays inside the Barrier Reef (H5), or very shallow intertidal (or barely
subtidal) reef crest pools at the Hog Islands or Glovers Reef (H6). In Belize this
species was taken at cays inside the Barrier Reef (BC, SC, TC) and at the atolls
(GR, LR). Additional records for Belize include one mainland site near Belize
City, one Barrier Reef site (CBC), and Turneffe Island. In Honduras P. nigripinnis
was taken at Roatan and the Hog Islands. Paraclinus nigripinnis was taken with P.
cingulatus (only at the Hog Islands) and P. fasciatus (only at the Hog Islands and
Roatan). Paraclinus nigripinnis shares highest affinity (GMI) values with Lab-
risomus nuchipinnis and Starksia occidentalis (table 12).
Material examined. — A total of 121 specimens from 14 collections. HON-
DURAS: CC, 84472 (23), 84474 (16), 84477 (6), 84478 (4), 84479 (2); IR, 84473 (1),
84475 (1), 84476 (3). BELIZE: BC, 86038 (6); GR, 86034 (7), 86035 (23); LR, 86033
(1); SC, 86037 (6); TC, 86036 (22). Additional Belize material examined: CBC,
JWC-23 (1); LR, UMML 9980; TI, UMML 9862; mainland, 2.4 km south-
southwest of Belize City, UMML 9606.
Starksia Jordan & Evermann, 1896
The genus Starksia contains 24 described species, 15 from the western Atlantic
and nine from the eastern Pacific. The eastern Pacific species were revised by
Rosenblatt & Taylor (1971). Reviews of the Atlantic species may be found in
Bohlke & Springer (1961), Gilbert (1965, 1971), and Greenfield (1978).
Eight of the 15 Atlantic species of Starksia are represented in our material from
Belize and Honduras. Six Atlantic species do not occur or are not known to occur
in the western Caribbean: S. brasiliensis (south Brasil), S. culebra (Haiti, Puerto
GREENFIELD & JOHNSON: BLENNIOID FISHES 37
Rico, and the lesser Antilles south to and including St. Vincent), S. fasciata
(Bahamas, Cuba, Antigua, Dominica), S. guttata (Tobago Cays and Grenadines
south to Trinidad and west to Curacao), S. ocellata (both coasts of Florida and
north to North Carolina, questionably from the Bahamas), S. variabilis (known
only from Santa Marta, Colombia). Starksia y-lineata is known from Grand
Cayman Island and Courtown Cays off Nicaragua. Therefore, nine of the 15
described western Atlantic species of Starksia occur off the Central American
coast. Co-occurrence data for species of Starksia is presented in Table 15.
(22) Starksia atlantica Longley, 1934
This species has been reported from the Bahamas, Haiti, Antigua, Venezuela,
Curacao, Old Providence Island, Lighthouse Reef and Turneffe Island (Belize),
Courtown Cays off Nicaragua, and Yucatan (Bohlke & Springer, 1961; Birdsong
6 Emery, 1968; Bohlke & Chaplin, 1968; Mago, 1970; Gilbert, 1971; Nagelkerken,
1974). This constitutes the first record for Honduras.
Our material agrees well with the description of this species provided by
Bohlke & Springer (1961). The following values for meristic characters are based
on 10 specimens, all from Belize. Dorsal-fin elements: XVIII, 8 (1); XIX, 8 (7); XX,
7 (2). Anal-fin elements: II, 15 (3) or 16 (7). Pectoral-fin rays: 13 (1), 14 (7), 15 (2).
Caudal-fin rays: 12 (1) or 13 (9).
The only consistent difference detected between our material from Belize and
Honduras and the description of Bahaman specimens provided by Bohlke &
Springer (1961) lies in the major pattern of pigmentation on the body. In the
Bahaman specimens (according to Bohlke & Springer, 1961) the pattern typically
consists of a more or less reticulate pattern of fine lightly pigmented lines that
break up the darker background into blocks of irregular size and shape, usually
arranged in about three horizontal tiers (illustrated in Bohlke & Chaplin, 1968, p.
526). In our material from the western Caribbean the pattern of body pigmenta-
tion in female specimens is usually much more regular (fig. 3A), consisting of 10
vertical bars of dark pigmentation separated by narrow lightly pigmented lines.
Only the last two (posteriormost) bars are sufficiently irregular in outline as to
correspond to the description of the Bahaman specimens. The first eight bars in
virtually all female western Caribbean specimens are quite regular in outline and
unbroken by horizontal white lines. All specimens, male and female, from the
western Caribbean have a prominent dark spot enclosing the bases of the last
dorsal-fin rays (fig. 3A, B). The presence of a dark spot enclosing the bases of the
last several anal-fin rays is apparently a sexually dimorphic character, present in
25 of 28 female specimens examined for it, present in only one of 28 male
specimens examined for it (chi-square = 41.35, p < .001). Sexual dimorphism in
color pattern (in specimens in alcohol) is also evident in the lesser development
of dark pigmentation on the body of males, the consequently lower contrast
between the wide dark bars and the narrow light bars, and in the lesser vertical
extent of the narrow light bars with a consequently incomplete definition of the
barred color pattern ventrally (fig. 3B).
Color in life (male). — Taken from 35-mm Kodachrome. Head reddish brown
with dark brown blotch on cheek posterioventrad to eye; upper and lower jaws,
snout, interorbital, and ventral surface of head red; pupil black, surrounded by
red iris with gold reflections. Body dark green, crossed by inconspicuous narrow
<Ti O rH O rH rH rH O
S 5
tN
_ !
•9 s
£'■2
3 2
C <f>
43 <«
~ to
b re
00
CO O O O O rH I o
§G
<N
1
>- c
13
r- ^
O <N O t-h rH 1 OP1
"3.3
2 5
to a
a* 2
53U
i* o>
in
'0
O"
a,
in
CM
vO tN rH Ov 1 OOrj
I
o
5 «
-9 -2
oc
c
1
el
1
3
u
O
1
o b
in
ID -h rH I (N O O O
1 co
>
to
01
J: w
O
U
$
s
X to
0*
■£ 1
>
s
e °
cs
o o 1 ^ m o o o
1 o o
£»
•2
o «
5?
■*- X.
-c
K
1 S3
5
V)
o> o
■—
Cu"-
3
rH 1 o ts in oo o os
o
ac
I O rH rH (Nj
9J
^^ ^.^ ^_^ ^_^ ^^ ^^
2*
C 01
6
Tfin^ostNmoocN
rtnotNni'NH
.a 3
01
3 >
a O
a.
intNtnomininoo
5ft
5
mcocomcMincnco
^vS'^'o^inCo^'co'
COriOMfDH/rtH
s
1 (NONMOHO
I O r* rH CS
1
(0
^^•^^^•^- — - — • w w
'"5 *"°
01
HoomtnoMONS
3
r}<COCMTt<rHCOCOCM
V. -»2
o c
>
0? 00 in CM" f^" 6>T CM "*
z~
MnNNO^Tfri
—
2
nHonn^^H
</i O
01 _
tN CM (N rH
o
TfNvO>OnNNO
tN tN (N m fO rH
•a 7,
—
s»
o>
oosinovinavmm
f,
LfitNOtn^^HiM
O n
o> y
c £
1 3
is
z
NtOM'tO^^ri
CM CM CM rH
2
aC
c
•c
m
en
a>
6
inos*f)HHino\oo
tn cn rt tn en rH tj<
in'oo'os'cn'Ss'f^F^ON
mC)rlH/Tji\0't(S
ONvoo^t'Ovotntn
co m in en m cm
£^
" c
3 ©
*'-2 •
9
•8
•52
ililllil
"S
Ol
a
en
—
o
•52
lisllll!
a;
a,
ajiillis
OJD
C
a| sg.§ sis
< 3>>
tn
cs>i)'C~S;0«w
•J3
«!u>*;--sco«w
H^"5
CMCMCMCMCMCMCMCM
t»
cscn^invor^ooos
CMCMCs|CM(NCM(NtN
o o
2
u
o
o>
1
3
C
II
Z
in
C
o
•J3
«
in
"o
o>
1
3
C
38
to
<3\
00
60
c
o
-J
a
t/5
<
(J
39
40 FIELDIANA: ZOOLOGY
white bars. Spines and rays of dorsal fin dark green, membranes clear to dusky.
Caudal, pelvic, and pectoral fins clear.
Color in life (female). — Taken from 35-mm Kodachrome. Head dark brown
dorsally, light brown on sides; anterior portion of upper and lower jaws dark
brown, remainder light brown; pupil black, surrounded by red iris with gold
reflections. Body dark brown, crossed by conspicuous narrow white bars.
Spines and rays of dorsal fin brown, membranes clear with small red spots along
base of posterior portion of spinous dorsal fin and soft dorsal fin. Caudal fin
dusky. Pectoral and pelvic fins clear.
Ecological data. — Starksia atlantica has been taken in collections where sampling
extended to 80 ft, but most of our records (17 of 22) for this species are from
stations in less than 15 ft, and a majority of both records (13 of 22) and specimens
(39 of 70) are from stations in less than 6 ft. Starksia atlantica was taken only in
areas of rich coral formations (table 13), and most specimens (96% of 70) were
taken in outer reef and reef flat sites (H2) at Glovers Reef, Lighthouse Reef, or
the Barrier Reef or from patch reefs (H3) in the lagoon at Glovers Reef. In Belize
S. atlantica has been taken at Barrier Reef sites (BRG, CBC) and at the atolls (GR,
LR, with additional records from TI). In Honduras S. atlantica was taken only in
one deep (45 to 60 ft) station at Roatan. Starksia atlantica was taken with S.
elongata, S. lepicoelia, S. nanodes, and S. sluiteri (table 15). Starksia atlantica shares
highest affinity (GMI) values with Acanthemblemaria greenfieldi and Enneanectes
boehlkei.
Material examined. — A total of 70 specimens from 22 collections. HONDURAS:
IR, 84534 (1). BELIZE: BRG, 89405 (1); CBC, 89298 (2), 89305 (1), 89322 (8), 89334
(1), 89342 (1), 89346 (7), 89362 (6), 89370 (1), 89374 (3), 89387 (7), 89400 (1); GR,
71101 (1), 71130 (4), 77627 (14), 77628 (1), 77629 (1), 77631 (1), 86012 (6), 86013 (1);
LR, 77630 (1). Additional Belize material examined: CBC, JWC-16 (2); LR, UMML
9497 (1); TI, UMML 9558 (7), UMML 9854 (6), UMML 10300 (1).
(23) Starksia elongata Gilbert, 1971
This species has previously been reported only from the southern Bahamas
(Gilbert, 1971). This constitutes the first record for Belize and for Honduras and a
significant range extension for the species.
Starksia elongata was described from eight specimens, five males, two females,
and one unsexed individual, collected in the southern Bahamas on small- to
medium-sized coral formations in 5 to 15 ft of water. We have collected S.
elongata on three occasions, twice at the Hog Islands, once at Carrie Bow Cay.
Two female specimens (FMNH 84575, 18.2, 19.1) were taken in 2 to 18 ft of water
in surge channels in the windward edge of the reef-top in an area of good coral
development on Little Hog Island (H2). One male specimen (FMNH 84576, 23.0)
was taken in 50 ft of water on the leeward side of Northwest Cay (Hog Islands)
in an area of good coral development (HI). One (female) specimen (FMNH
89368, 20.5) was taken in 0 to 4 ft of water in a patch reef formation near the
south end of Carrie Bow Cay (H3). Starksia elongata was taken with five other
species of Starksia (table 15).
All four specimens agree very well with Gilbert's (1971) description of S.
elongata. We present only values for meristic characters here (based on all four
GREENFIELD & JOHNSON: BLENNIOID FISHES 41
specimens in the order listed above). Dorsal fin: spines, 20, 21, 21, 20; segmented
rays, 8 (all); total elements, 28, 29, 29, 28. Anal fin: spines, 2 (all); segmented
rays, 17, 18, 17, 17. Pectoral-fin rays: 14 (all). Total elements, dorsal + anal +
pectoral: 61, 63, 62, 61. Pelvic-fin segmented rays, 2 (all). Caudal-fin segmented
rays, 13 (all). Arched lateral line scales, 18, ? 18, ? 18.
Material examined. — Four specimens from three collections. HONDURAS: CC,
84575 (2), 84576 (1). BELIZE: CBC, 89368 (1).
(24) Starksia hassi Klausewitz, 1958
This species has previously been reported from the Bahamas, Puerto Rico,
Virgin Islands, Antigua, Guadeloupe, Bonaire, Venezuela, and Panama (Cervi-
gon, 1968; Gilbert, 1971). This constitutes the first record for Belize.
Ecological data. — This species was taken only at two stations near the top of the
dropoff zone at Glovers Reef, both at depths in excess of 60 ft, both in areas of
rich coral development (HI). Starksia hassi was taken with S. lepicoelia and S.
nanodes (table 15).
Material examined. — Two specimens from two collections. BELIZE: GR, 86014
(1), 86015 (1).
(25) Starksia lepicoelia Bohlke & Springer, 1961
This species has been reported from the Bahamas, Virgin Islands, Antigua,
Old Providence Island, Grand Cayman Island, Turneffe Island and Lighthouse
Reef (Belize), Honduras, Albuquerque Cays off Nicaragua, Alacran Reef, and
Banco Chinch orro off Yucatan (Gilbert, 1971).
Ecological data. — Starksia lepicoelia was taken at more stations and is rep-
resented by more specimens than any other species of Starksia in our collections.
This may well be an artifact of sampling effort, because the overwhelming
majority of both records (17 of 24) and specimens (74 of 103) are from Glovers
Reef (see table 3). Starksia lepicoelia was taken by us in less than 4 ft of water [IR,
F75-14, large but isolated coral formation in Caribe Pt. Bite Bay (H3)], but the
majority of both records (13 of 24) and specimens (66 of 103) are from depths in
excess of 40 ft (table 6). All specimens of S. lepicoelia are from areas of rich coral
development (table 13), and the clear majority of both records (15 of 24) and
specimens (76% of 103) are from reef-front dropoff zones (HI or, in the one case
in which one specimen of S. lepicoelia was taken inquiline in a sponge, recorded
as H9). Starksia lepicoelia was taken at a total of five sites, three in Belize (CBC,
GR, LR) and two in Honduras (CC, IR). Additional records from Belize include
Turneffe Island. Starksia lepicoelia was taken with all species of Starksia except S.
starcki. Starksia lepicoelia shares highest affinity (GMI) values with Enneanectes
atrorus and E. altivelis.
Material examined. — A total of 103 specimens from 24 collections. HON-
DURAS: CC, 84537 (2), 84538 (2); IR, 84535 (2), 84536 (8). BELIZE: CBC, 89296
(6), 89341 (4); GR, 70942 (1), 70970 (5), 70977 (1), 71080 (1), 71090 (1), 71106 (8),
71117 (10), 71134 (2), 71139 (1), 71143 (8), 71151 (1), 77632 (4), 77703 (5), 82589 (1),
84552 (1), 86019 (23), 86020 (1); LR, 77702 (5). Additional Belize material exam-
ined: LR, UMML 9411 (1), UMML 9474 (14); TI, UMML 9546 (4), UMML 9837 (4).
42 FIELDIANA: ZOOLOGY
(26) Starksia nanodes Bohlke & Springer, 1961
This species has been reported from the Bahamas, Virgin Islands, Haiti, Old
Providence Island, Grand Cayman Island, Colombia, Turneffe Island (Belize),
and Albuquerque Cays off Nicaragua (Birdsong & Emery, 1968; Gilbert, 1971;
Palacio, 1974). This constitutes the first record for Honduras.
Among the western Atlantic species of Starksia, only S. elongata and S. nanodes
have a pair of hypural-shaped dark markings at the base of the caudal fin. These
two species may be distinguished by characters provided in the key to Starksia by
Gilbert (1971). They may also be distinguished, without overlap in character
values, in the sum of total dorsal + anal + pectoral elements (table 16).
Table 16. Comparison of counts of total dorsal + anal + pectoral elements (spinous +
soft) for Starksia elongata and S. nanodes.
Species 56 57 58 59 60 61 62 63 64 N
Starksia nanodes,
Belize and Honduras 2 4 9 1 — — — — — 16
Starksia elongata,
Belize and Honduras — — — — — 2 1 1 — 4
Starksia elongata,
Bahamas* ______ 3 4 1 8
*Data from Gilbert 1971, p. 198.
Ecological data. — All but four records for this species are from depths in excess
of 45 ft, and confirmed depth records extend to 90 to 100 ft (near the maximum
depth of sampling in our data). All records are from areas of rich coral develop-
ment (table 13), with the majority of both records (16 of 20) and specimens (79%
of 57) from reef-front dropoff zones (HI). Starksia nanodes was taken at only four
sites, two in Belize (CBC, GR), two in Honduras (CC, IR). The majority of
records (12 of 20) are from Glovers Reef, but the majority of specimens (30 of 57)
are from the Barrier Reef near Carrie Bow Cay. Starksia nanodes was taken with all
species of Starksia except S. sluiteri. Starksia nanodes shares highest affinity (GMI)
values with Lucayablennius zingaro and Enneanectes atrorus (table 15).
Material examined. — A total of 57 specimens from 20 collections. HONDURAS:
CC, 84540 (4); IR, 84539 (2). BELIZE: CBC, 89304 (3), 89336 (5), 89347 (7), 89356
(6), 89361 (1), 89398 (8); GR, 77634 (1), 77635 (1), 77636 (1), 77637 (1), 77638 (3),
77639 (2), 77640 (2), 77808 + 86011 (2), 77809 (3), 86016 (1), 86017 (2), 86018 (2).
Additional Belize material examined: TI, UMML 9576 (1).
(27) Starksia occidentalis Greenfield, 1978
This species ranges from the east side of the Yucatan Peninsula, Mexico, south
along the coast of Central America to Panama (Greenfield, 1978). Records of S.
ocellata from Old Providence Island (Fowler, 1950) and Panama (Gilbert, 1971)
refer to this species.
Ecological data. — Starksia occidentalis has been taken in stations in which sam-
pling extended to 50 ft, but most records (11 of 14) and specimens (40 of 53) are
from less than 10 ft of water. Starksia occidentalis was taken in a variety of habitats
(table 13), mainly coral and/or rock in Honduras (H2, H4, H9) and coral or coarse
coral rubble (H2, H5) in Belize. In Belize S. occidentalis was taken at cays inside
GREENFIELD & JOHNSON: BLENNIOID FISHES 43
the Barrier Reef (BC, FC, SC, TC) and at the Barrier Reef (BRG). Along with
Labrisomus kalisherae and Ophioblennius atlanticus macclurei, S. occidentalis was the
only species of blennioid represented in our collections to be taken at all four
sites at cays behind the Barrier Reef (species taken at three of the four sites
include: Labrisomus nuchipinnis, Malacoctenus triangulatus, Paraclinus nigripinnis).
In Honduras S. occidentalis was taken at Roatan and the Hog Islands. Starksia
occidentalis was taken with five of the other seven species of Starksia represented
in our collections (table 15). It was taken with S. elongata, S. lepicoelia, S. nanodes,
and S. starcki only at the Hog Islands (and in all but one case was taken with all at
the same station, F75-48). Starksia occidentalis was taken with S. sluiteri only at the
one Belize Barrier Reef collection (BRG) in which S. occidentalis was represented.
Starksia occidentalis shares highest affinity (GMI) values with Labrisomus kalisherae
and Malacoctenus triangulatus.
Material examined. — A total of 53 specimens from 14 collections. HONDURAS:
CC, 84377 (5), 84378 (6), 84379 (2), 84380 (1), 84381 (6), 84383 (6), 84384 (9), 84581
(6); IR, 84382 (5). BELIZE: BC, 84386 (1); BRG, 84385 (1); FC, 84387 (1); SC, 84389
(1); TC, 84388 (3).
(28) Starksia sluiteri (Metzelaar, 1919)
This species has been recorded from Antigua, Dominica, Bonaire, Los Roques
and Curacao (off Venezuela), Old Providence Island, Venezuela, Nicaragua, and
Yucatan (Bohlke & Springer, 1961; Birdsong & Emery, 1968; Gilbert, 1971). This
constitutes the first record for Belize.
Color in life. — Taken from 35-mm Kodachrome of specimens from Carrie Bow
Cay. Head: yellowish brown dorsally; preopercle with large bluish black irides-
cent blotch bordered posteriorly with salmon-red; opercle brown with white line
bordering preopercle on ventral half; anterior portion of upper lip with red bar,
remainder of upper jaw and entire lower jaw and ventral surface of head yellow;
white line running from ventral surface of eye along dorsal margin of upper jaw,
under preopercular blotch and joining white line of opercle; eye with black pupil
surrounded by narrow red ring on iris, remainder of iris black with five yellow
lines radiating out from red ring in spokelike fashion to outer margin of eye;
salmon-red line on circumorbital bones beginning under center of eye and ex-
tending dorsally along posterior margin of eye to line opposite top of orbit. Body
with series of seven black blotches along side just above midline, smaller blotch
above and below major row, lower row under and touching middle series, upper
row set between middle series and touching dorsal fin, posteriormost blotches
forming hypural blotch; edges of scales outlined with brownish red, centers tan
above midline and bluish grey below midline; belly grey. Dorsal fin with pair of
salmon-red dots above each black blotch of body that extends onto fin, fin spines
dusky, membrane in anterior half of fin clear, posterior half of spinous dorsal fin
and soft dorsal fin dusky, rays of soft dorsal fin black. Anal fin dusky with
salmon-red margin. Caudal fin clear, crossed by numerous fine reddish brown
lines, distal margin salmon-red. Pectoral fins clear with salmon-red ventral and
posterior margins. Pelvic fins white.
Ecological data. — Starksia sluiteri is represented in our collections by six speci-
mens taken at four stations, all in Belize, three at Barrier Reef sites (BRG, CBC),
one at Glovers Reef. All stations positive for this species were in shallow water
44 FIELDIANA: ZOOLOGY
(all less than 15 ft, all but one less than 5 ft). All stations were typified by richly
developed coral formations in reef crest and fore-reef areas (H2). Starksia
sluiteri was taken with three other species of Starksia (table 15) and shares highest
affinity (GMI) values with Labrisomus gobio and Malacoctenus aurolineatus.
Material examined. — Six specimens from four collections. BELIZE: BRG, 86021
(1); CBC, 89294 (1), 89394 (3); GR, 86022 (1).
(29) Starksia starcki Gilbert, 1971
This species has previously been reported only from the type locality, Looe
Key in the lower Florida Keys (Gilbert, 1971). This constitutes the first record for
Honduras and a significant range extension for the species.
Gilbert's (1971) description of S. starcki was based on six (20.3 to 27.3) speci-
mens from two collections. All specimens were taken in surge channels, less
than 25 ft in depth. Our single specimen of S. starcki [FMNH 84574, 1 (16.3),
original field number FMNH 75-48] was taken 20 May 1975 at Northwest Cay,
Hog Islands, Honduras, in about 50 ft of water in an area of good coral develop-
ment (HI). Three additional species of Starksia were represented in this collec-
tion (table 15).
Our identification of this specimen of S. starcki is based on comparison of the
Honduras specimen with both Gilbert's (1971) description of S. starcki as well as
with three paratypes of S. starcki examined by us.
An abbreviated description of the Honduras specimen follows. Meristic char-
acters: dorsal-fin elements, XX, 9. Anal-fin elements, II, 18. Pectoral-fin rays,
13/13. Pelvic-fin elements, I, 2. Segmented caudal-fin rays, 13. Neither lateral
line scale number or arrangement could be determined in the Honduras speci-
men, due to its small size. Morphometric characters: listed as thousandths of the
SL, taken according to the methods of Gilbert (1971). Head length, 344. Snout
length, 61. Upper jaw length, 135. Horizontal eye diameter, 92. Greatest body
depth, 202. Length of first dorsal-fin spine, 86. Length of longest pectoral ray,
301. Length of longest pelvic-fin ray, 245. Belly fully scaled. Pores from infraorbi-
tal ossifications single. Nasal, supraorbital, and nuchal cirri simple, filamentous,
and subequal in length. Neither vomerine or palatine teeth could be seen. Ap-
pressed pectoral-fin reaching a vertical through base of seventh segmented
anal-fin ray. Appressed pelvic fin reaching past anal-fin origin. Major color
pattern consisting of 10 irregular, broken, widely spaced chocolate bars on body
which contrast strongly with a light background. Distribution of head and body
pigmentation very similar to that described by Gilbert (1971) for the type series.
Gilbert (1971, p. 203) notes that soft dorsal- and anal-fin ray counts are higher
in S. starcki than in S. lepicoelia (the only other western Atlantic species with a
fully scaled belly), but that overlap occurs. We note that total dorsal- plus anal-
fin ray counts will separate the two species without overlap (table 17).
Material examined. — One specimen from one collection. HONDURAS: CC,
FMNH 84574 (1). We have also examined three paratypes of S. starcki: ANSP
109800, 1 (26.3); UF 16188, 1 (20.3); UF 17279, 1 (20.5).
Stathmonotus Bean, 1885
The genus Stathmonotus contains six described species, three in the western
Atlantic and three in the eastern Pacific. The latest revision of Stathmonotus is
GREENFIELD & JOHNSON: BLENNIOID FISHES 45
Table 17. Comparison of total dorsal-fin plus anal-fin counts (spinous + soft) for Starksia
lepicoelia and S. starcki.
Species 43 44 45 46 47 48 49 50 51 N
Starksia lepicoelia,
Belize and Honduras 1 2 19 6 6 — — — — 34
Starksia starcki, Honduras — — — — — — 1 — — 1
Starksia starcki, Florida* — — — — — — 2 3 1 6
*Data from Gilbert 1971, p. 202.
Table 18. Co-occurrence of species of Stathmonotus. Values in upper half of matrix
(whole numbers) are actual number of records (number of stations) of co-occurrence
(based on all stations). Values in lower half of matrix are calculated GMI values (based on
Class I stations only).
Co-occurring species
Species NT N, 30 31 32
30. gymnodermis 11 — 0 0
31. hemphilli 110 — 1
32. stahli tekla 13 13 0 .14
Listing of species sharing highest GMI values with species of Stathmonotus (five highest
values only).
30. gymnodermis 9 (.14), 7 (.14), 21 (.13), 2 (.12), 8 (.12)
31. hemphilli 17 (.50), 49 (.29), 47 (.25), 53 (.22), 57 (.19)
32. stahli tekla 35 (.37), 37 (.32), 51 (.32), 55 (.31), 27 (.31)
NT = total number of stations; Ni = number of Class I stations.
that of Springer (1955b). All three western Atlantic species of Stathmonotus occur
in our material from Belize and Honduras. Co-occurrence data for species of
Stathmonotus is presented in Table 18.
(30) Stathmonotus gymnodermis Springer, 1955
This species has been reported from the Bahamas, Puerto Rico, Virgin Islands,
and Venezuela (Springer, 1955b; Cervigon, 1966; Bohlke & Chaplin, 1968; Mago,
1970). This constitutes the first record for Belize and a significant range extension
for the species.
Stathmonotus gymnodermis is represented in our material by one specimen (13.9
mm SL) taken at Glovers Reef in a reefcrest pool with a coral rubble and algae
bottom in less than 2 ft of water (H6). Counts for this specimen follow: dorsal-fin
elements, XLII, 0; anal-fin elements; II, 23; pectoral-fin rays, 8; segmented
caudal-fin rays, 11. In both counts and other characters this specimen agrees
well with Springer's (1955b) description of S. gymnodermis.
Material examined. — One specimen from one collection. BELIZE: GR, 86010 (1).
(31) Stathmonotus hemphilli Bean, 1885
This species has been reported from the Bahamas, Florida, St. Croix, Haiti,
and Antigua (Springer, 1955b; Bohlke & Chaplin, 1968). This constitutes the first
record for Honduras and a significant range extension for the species.
Stathmonotus hemphilli is represented in our material by one specimen (28.9
mm SL) taken at Big Hog Island from a massive living coral head in less than 6 ft
46 FIELDIANA: ZOOLOGY
of water. Counts for this specimen are as follows: dorsal-fin elements, L, 0;
anal-fin elements, II, 28; pectoral-fin rays, 4; segmented caudal-fin rays, 11. In
both counts and other characters this specimen agrees well with the diagnosis of
this species provided by Springer (1955b, p. 68).
Material examined. — One specimen from one collection. HONDURAS: CC,
84360 (1).
(32) Stathmonotus stahli tekla Nichols, 1910
Springer (1955b) recognized two subspecies of S. stahli, the nominate sub-
species S. s. stahli (Evermann & Marsh, 1899), reported from Bonaire, Puerto
Rico, Virgin Islands, Martinique, and Venezuela (Springer, 1955b; Cervigon,
1966; Mago, 1970), and S. s. tekla Nichols, reported from the Tortugas, Bahamas,
Cuba, and Old Providence Island (Springer, 1955b; Bohlke & Chaplin, 1968).
Birdsong & Emery (1968) reported S. stahli (without assignment of their material
to subspecies) from Belize and Yucatan. We continue to follow Springer (1955b)
and Bohlke & Chaplin (1968) in recording the form taken by us in Belize and
Honduras as S. s. tekla (13 of 13 specimens counted had 11 segmented caudal-fin
rays). This constitutes the first record for Honduras.
Ecological data. — Stathmonotus stahli tekla was taken in collections in which
sampling extended to 30 ft, but the large majority of both records (11 of 13) and
specimens (43 of 45) are from stations in less than 10 ft of water. Although S. s.
tekla was taken in a variety of habitats (table 13), the clear majority of both
records (9 of 13) and specimens (77% of 45) are from rich shallow-water coral
formations (H2, H3), shallow areas typified by rocky ledges (H4), or both coral
and rock reef formations (H9, in the case of F75-51, an exceptionally broad
station). In Belize S. s. tekla was taken at cays inside of the Barrier Reef (BC, TC),
Barrier Reef sites (CBC with an additional record from SWC), and the atolls (GR,
LR with an additional record from TI). In Honduras S. s. tekla was taken at the
Hog Islands and at Roatan. Stathmonotus stahli tekla was taken with S. hemphilli at
the one station successful for the latter. Stathmonotus stahli tekla shares highest
affinity (GMI) values with Enneanectes boehlkei, E. pectoralis, and Entomacrodus
nigricans (table 18).
Material examined. — A total of 45 specimens from 13 collections. HONDURAS:
CC, 84362 (1), 84363 (9), 84364 (5); IR, 84361 (2). BELIZE: BC, 86008 (1); CBC,
89291 (7), 89317 (2), 89341 (7), 89402 (1), 89415 (3); GR, 71125 (1), 86007 (1); FC,
86009 (5). Additional Belize material examined: CBC, JWC-19 (2); LR, UMML
9987; SWC, JWC-16 (1); TI, UMML 9856.
TRIPTERYGIIDAE
Enneanectes Jordan & Evermann, 1895
The family Tripterygiidae is represented in the western Atlantic by five species
belonging to the New World genus Enneanectes. The western Atlantic species of
Enneanectes were revised by Rosenblatt (1960). The status of eastern Pacific
species remains unclear, pending publication of Rosenblatt's revision of the
family (Rosenblatt, 1959), but none of the species of Enneanectes is amphi-
American. All five western Atlantic species of Enneanectes are represented in our
collections from Belize and Honduras. Co-occurrence data for species of En-
neanectes is presented in Table 19.
0
5
0
0
3
16
3
GREENFIELD & JOHNSON: BLENNIOID FISHES 47
Table 19. Co-occurrence of species of Enneanectes. Values in upper half of matrix (whole
numbers) are actual number of records (number of stations) of co-occurrence (based on all
stations). Values in lower half of matrix are calculated GMI values (based on Class I
stations only).
Co-occurring species
Species NT N, 33 34 35 36 37
33. altivelis 28 28 — 8 10
34. atrorus 17 16 .28 — 0
35. boehlkei 29 29 .26 0 —
36. jordani 4 3 0 0 .23
37. pectoralis 20 20 .12 0 .57 .28
Listing of species sharing highest GMI values with species of Enneanectes (five highest
values only).
33. altivelis 5 (.41), 25 (.39), 26 (.37), 22 (.31), 34 (.28)
34. atrorus 50 (.56), 26 (.45), 25 (.43), 43 (.34), 33 (.28)
35. boehlkei 4 (.70), 15 (.58), 37 (.57), 55 (.55), 22 (.50)
36. jordani 23 (.38), 37 (.28), 35 (.23), 55 (.22), 15 (.22)
37. pectoralis 55 (.64), 35 (.57), 4 (.57), 7 (.57), 15 (.51)
NT = total number of stations; N, = number of Class I stations.
(33) Enneanectes altivelis Rosenblatt, 1960
This species has been reported from the Bahamas and Central America, in-
cluding Belize, Nicaragua, and Yucatan (Rosenblatt, 1960; Bohlke & Chaplin,
1968; Birdsong & Emery, 1968). This constitutes the first record for Honduras.
Ecological data. — Enneanectes altivelis has the broadest bathymetric distribution
of any species of Enneanectes represented in our collections. It has been taken in
less than 4 ft and as deep as 100 to 110 ft (the maximum depth of our sampling
efforts). If capture records for E. altivelis are grouped into three depth categories
(0-4, 5-59, ^ 60 ft), the actual distribution of capture records (table 6) does not
differ from what would be expected on the basis of distribution of sampling
effort (table 2) (\i = 270, p > .20, Yates Correction applied). There is therefore
in our data no evidence that E. altivelis exhibits a particular depth preference
over the depth range covered by our samples. Virtually all records and speci-
mens of E. altivelis came from areas typified by rich coral formations (table
20 — the two records cited as H9 are from F75-47 and F75-51 in the Hog Islands,
two exceptionally broad stations that included both rock and coral reef
formations in the area sampled). The majority of both records (20 of 28) and
specimens (81% of 113) came from reef-front dropoff zones (HI) or patch reefs
(H3). In Belize E. altivelis was taken at Barrier Reef sites (BRG, CBC) and the
atolls (GR, LR, with an additional record from Turneffe Island). In Honduras E.
altivelis was taken at Roatan and the Hog Islands. Enneanectes altivelis was taken
with three other species of Enneanectes: in deep water with E. atrorus and in
shallow water with E. boehlkei and E. pectoralis (table 21). Enneanectes altivelis
shares highest affinity (GMI) values with Labrisomus haitiensis and Starksia
lepicoelia (table 19).
Material examined. — A total of 113 specimens from 28 collections. HON-
DURAS: CC, 84519 (3), 84520 (1), 84521 (6), 84522 + 84523 (7), 84524 (1); IR,
84518 (12). BELIZE: BRG, 89403 (1); CBC, 89349 (7), 89357 (4), 89399 (1); GR,
70966 (3), 86131 (1), 86132 (2), 86133 (8), 86134 + 86135 (3), 86136 + 86137 (3),
s
4
ffv NN rHfl flrH
I I
I I
I I I I I I I I I I
° s
-»-» LI
R3
(8
£1111111111
£ 1
Oh 3
M (0
* i
II
§ g
5 £
cfl O
^ Oh
on fcj
T3 Oh
u C
0) en
S Jr,
5 -Q
8 c
O a)
go
| o
Q ^
• >-
S^-
W *
>-) 01
H 1
C
o
60
I I I I I I I I
0> v£> r- 1 H n s
5
CO On 1-h CN
¥ iHT|i HO\ I
00 CO KO> 0> CO t*< On ON
Nrl HM NOO CN 00
01
Oh
CD
5 *
~ -C Q
•a s
3
48
GREENFIELD & JOHNSON: BLENNIOID FISHES 49
Table 21. Number of stations at which Enneanectes altivelis was captured with other
species of Enneanectes categorized by depth interval and habitat type.
A. Depth interval (feet)*
Co-occurring species
of Enneanectes
0 to 4
5 to 39
^40
N
34. atrorus
35. boehlkei
36. pectoralis
0
8
5
0
2
0
8
0
0
8
10
5
HI
B.
H2
Habitat type
H3
H4
H9
N
34. atrorus
35. boehlkei
36. pectoralis
8
0
0
0
4
3
0
3
0
0
1
1
0
2
1
8
10
5
* Stations assigned to depth interval on the basis of shallowest sampling depth.
86138 + 86139 (12), 86140 (2), 86141 (3), 86142 (5), 86143 (5), 86144 (3), 86145 (2),
86146 (1), 86147 (2), 86148 (8), 86149 (5), 86150 (2). Additional Belize material
examined: LR, UMML 9490 (3), UMML 9566 (4); TI, UMML 10306 (2).
(34) Enneanectes atrorus Rosenblatt, 1960
This species has been reported from the Bahamas and from Turneffe Island
(Belize) (Rosenblatt, 1960; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968).
This constitutes the first record for Honduras.
Ecological data. — Enneanectes atrorus has been taken only in relatively deep
water, always at depths in excess of 30 ft, with most records (14 of 17) and
specimens (31 of 39) from depths of 60 ft or more. We have taken this species in
depths as great as 90 to 100 ft. All records are from reef-front dropoff zones (HI),
including a record of one specimen (FMNH 82590) taken inquiline in a sponge
(recorded as H9 in table 20). In Belize this species was taken only at the atolls
(GR, LR, with an additional record at Turneffe Island). Whether it is truly absent
from the Barrier Reef or not is unknown, but it was not taken in any of the six
deep (> 40 ft) stations in the reef-front dropoff area at CBC. In Honduras we
took this species only at the one deep (45 to 60 ft) station at Roatan in the dropoff
zone outside the entrance to Dixon Cove. Enneanectes atrorus was taken only with
E. altivelis among the species of Enneanectes and shares highest affinity (GMI)
values with Lucayablennius zingaro and Starksia nanodes (table 19).
Material examined. — A total of 39 specimens from 17 collections. HONDURAS:
IR, 84516 (4). BELIZE: GR, 70930 (1), 71078 (1), 71097 + 86116 (3), 82590 (1), 86115
(2), 86117 + 86118 (6), 86119 (5), 86120 (4), 86121 (1), 86123 (1), 86124 (1), 86125
(3), 86126 (2), 86128 (1), 89410 (1); LR, 86127 (2). Additional Belize material
examined: TI, UMML 9852 (2).
(35) Enneanectes boehlkei Rosenblatt, 1960
This species has been reported from the Bahamas, Florida, Puerto Rico, Virgin
Islands, Venezuela, and Central America, including Belize, Costa Rica, Nica-
ragua, and Yucatan (Rosenblatt, 1960; Caldwell, 1963; Cervigon, 1966; Birdsong
& Emery, 1968; Bohlke & Chaplin, 1968; Mago, 1970). This constitutes the first
record for Honduras.
50 FIELDIANA: ZOOLOGY
Ecological data. — Enneanectes boehlkei has been taken in stations in which sam-
pling extended to 30 ft, but most records (25 of 29) and specimens (166 of 183) are
from stations in less than 15 ft. [Note that other than general locality (CC) we
have no data for the series of seven specimens catalogued as FMNH 84515, and
these specimens are excluded from all analyses.] The majority of records (23 of
29) and specimens (70% of 183) are from shallow richly developed coral forma-
tions (H2, H3; table 20). All but one of the remaining records and specimens are
from areas of rocky ledges (H4) or two sites with both rock and coral represented
in exceptionally broad Hog Island stations (F75-47, F75-51, H9 in table 20). In
Belize E. boehlkei was taken at Barrier Reef sites (BRG, BRT, CBC, with an addi-
tional record from SWC) and at the atolls (GR, LR, with an additional record
from TI). In Honduras E. boehlkei was taken at Roatan and the Hog Islands.
Enneanectes boehlkei was taken with the three other species of Enneanectes occur-
ring in shallow water but was not taken with the deep-water E. atrorus. En-
neanectes boehlkei shares highest affinity (GMI) values with Labrisomus guppyi and
Malacoctenus triangulatus (table 19).
Material examined. — A total of 190 specimens from 30 collections. HON-
DURAS: CC, 84506 (16), 84509 (7), 84510 (4), 84511 (5), 84512 (4), 84513 (15),
84514 (13), 84515 (7); IR, 84504 (1), 84505 (4). BELIZE: BRG, 86114 (2); BRT, 86105
(2); CBC, 89288 (1), 89290 (12), 89306 (2), 89321 (17), 89343 (8), 89348 (4), 89369
(4), 89371 (1), 89375 (4), 89382 (9); GR, 70931 (1), 86106 (2), 86109 (1), 86110 (2),
86111 (24), 86113 (1); LR, 86107 + 86108 (7), 86112 (10). Additional Belize material
examined: LR, UMML 9480, UMML 9548; SWC, JWC-24 (3); TI, UMML 9826.
(36) Enneanectes jordani (Evermann & Marsh, 1900)
This species has been reported from the Bahamas, Puerto Rico, and Venezuela
(Rosenblatt, 1960; Bohlke & Chaplin, 1968; Mago, 1970). This constitutes the first
record for Belize and Honduras and a significant range extension for the species.
The rarity of E. jordani in our collections combined with the difficulty of deter-
mining key characters in several of the specimens (particularly lateral line scale
counts, because several specimens were damaged in capture and/or preserva-
tion) aroused our suspicions as to the correctness of our identification. Re-
examination of the material confirmed that the specimens listed below best agree
with E. jordani. Bohlke & Chaplin (1968, p. 556) comment on the rarity of E.
jordani in their Bahaman collections.
Ecological data. — We have taken E. jordani on only four occasions, once in
Honduras (CC, 2 to 18 ft, H2), three times in Belize (BRB, 0 to 6 ft, H2; CBC, 0 to
4 ft, H3; GR, 3 to 15 ft, H2). Enneanectes jordani was taken with E. boehlkei and E.
pectoralis (table 19).
Material examined. — A total of nine specimens from four collections. HON-
DURAS: CC, 84517 (2). BELIZE: BRB, 86130 (1); CBC, 89363 (1); GR, 86129 (5).
(37) Enneanectes pectoralis (Fowler, 1941)
This species has been reported from the Bahamas, Florida, Virgin Islands,
Aruba, Martinique, Venezuela, and Central America, including Belize, Nica-
ragua, and Yucatan (Rosenblatt, 1960; Caldwell & Caldwell, 1964; Cervigon,
1966; Birdsong & Emery, 1968; Bohlke & Chaplin, 1968). This constitutes the first
record for Honduras.
GREENFIELD & JOHNSON: BLENNIOID FISHES 51
Ecological data. — Enneanectes pectoralis was taken in stations in which sampling
extended to 18 ft, but a majority of records (14 of 20) and specimens (65 of 87) are
from stations in less than 6 ft. Enneanectes pectoralis was taken in a number of
shallow-water habitats (table 20), but most records (14 of 20) and specimens
(89% of 87) are from richly developed shallow reef formations (H2, H3). In Belize
E. pectoralis was the only species to be taken at cays inside the Barrier Reef (BC,
TC). It was also taken at Barrier Reef sites (BRG, BRT, CBC) and at the atolls
(GR, LR, with an additional record from TI). In Honduras E. pectoralis was taken
at Roatan and the Hog Islands. Enneanectes pectoralis was taken with all three
species of Enneanectes occurring in shallow water (table 19) but was not taken
with the deep-water E. atrorus. Enneanectes pectoralis shares highest affinity
(GMI) values with Ophioblennius atlanticus macclurei and Enneanectes boehlkei.
Material examined. — A total of 87 specimens from 20 collections. HONDURAS:
CC, 84526 (1), 84527 (1), 84528 (7), 84529 (1), 84530 (1), 84531 (2); IR, 84525 (4).
BELIZE: BC, 86155 (3); BRG, 86152 (8), 86158 (5); BRT, 86157 (3); CBC, 89289 (10),
89318 (8), 89364 (5), 89376 (11), 89381 (2); GR, 86151 (1), 86154 (10); LR, 86153 (3);
TC, 86156 (1). Additional Belize material examined: TI, UMML 9571 (1).
CHAENOPSIDAE
The family Chaenopsidae is represented in the western Atlantic by 32 de-
scribed species distributed among eight genera. In the following listing of west-
ern Atlantic chaenopsid genera, the numbers in parentheses indicate the fol-
lowing information: total number of western Atlantic species, number of species
known from off the Central American coast, number of species represented in
our material from Belize and Honduras. Western Atlantic chaenopsid genera
include Acanthemblemaria (8, 6, 4), Chaenopsis (4, 1, 0), Coralliozetus (1, 1, 1),
Ekemblemaria (1, 1, 0), Emblemaria (8, 3, 3), Emblemariopsis (8, 4, 3), Hemiemblemaria
(1, 1, 1), and Lucayablennius (1, 1, 1). All western Atlantic chaenopsid genera and
18 of 32 described western Atlantic species are known from off the Central
American coast. Of these, 13 species belonging to all genera but Chaenopsis3 and
Ekemblemaria are represented in our material. The names we employ for
chaenopsid taxa treated in this paper are in accord with the recent literature:
Stephens, 1963, 1970; Robins & Randall, 1965; Robins, 1971a; Smith-Vaniz &
Palacio, 1974; Johnson & Greenfield, 1976. Co-occurrence data for chaenopsid
species are presented in Table 22.
Acanthemblemaria Metzelaar, 1919
The genus Acanthemblemaria contains 14 species, eight in the western Atlantic
and six in the eastern Pacific. The latest revision of western Atlantic species is
that of Smith-Vaniz & Palacio (1974). The latest revision of eastern Pacific species
is that of Stephens (1963). One additional eastern Pacific species has been de-
scribed subsequently (Stephens et al., 1966).
Four of the eight recognized western Atlantic species occur in our collections
from Belize and Honduras. The following species are not represented in our
collections: A. betinensis Smith-Vaniz & Palacio, 1974 (Colombia, Panama, Costa
Rica); A. chaplini Bohlke, 1957 (Florida, Bahamas); A. medusa Smith-Vaniz &
Palacio, 1974 [Lesser Antilles, Tobago, Los Roques (Venezuela)]; A. rivasi
3See addendum.
O rHOOOOfNrHOr-iOrtO
XI
6
3
C
1
>.
(J
r
01
0
»!
r
0
0
s-
•n
01
Q
£
3
c
ai
M
(8
R
U
u
E
(B
0
Nr lOrHOfNOOrHO I O rH
CO T5
'S-S
tj
a.
>
0
c
tn
cu
C
(8
0
XI
•43
rq
0
01
55
u
(0
c
01
r
c
0
R
^
A
e
U
01
in
w
e
DO
d
<
8
t^ COCMOrHOOOOCM
3
CMCMOCMOCMOO
in o o o
JPrti o o o o o o
-° ifi
c
e =
3
C >
01 _
o
1
o
U
CM O O rH O 1 O
|U
o-a
X Jj
'S 2
O O O O 1 o o
2 3
E.S
u_. (0
e u
IT) O O 1 O O O
(0 (8
»*
■C x
a5 J3
Oh «
C O
o
9
CM CM 1 O O O O
01 <s
3 X
4 fe
OS
IX 1 -<*i NO O O O
1 cm co
in O
o o o o o X?
O O CO O m
o o o o o XI
no o o o o .5
O O O CM O
00 I CO CM rH o O O I O CM O NO O
J*5 I HNH
_£" inoiNKriNrHniOMnn^
Z (N CM rH rH (Nl
X
60
C
•s
•Si a « ~ -~ c
fc § •§ § -§ -§ .2 "S SS^-
a.s « -5 g S ~ sc f .y c » &o
a,
en
8 SbS 8-8 S
■g »>S g--c .gs.S .*
ooavdrHiNm^invoNodaic) •«
COCO'*T}<-<tf-<*i'3<TtiTf<'H/i-5tf-5jiin ,J
o~ 5<" no~ fv~ cm" on o~
CO CM O CM rH CM rH
w oo m on co cm in o h inco N to
^5 in in in in >* ^ h/ h h/
! $3
3 NH/(sa,Nvom
Smvo
°^h ^
X
60
rH CO ,
•u
on m in rH m
n^HjjiHNNtnin
O SO >-.in r-f
no in Tt oo in in cm
•>* in cn co cm in rH
,- in 00 rH rH NO
5 cm cm co ■* in
N_^ ^^ s .^_, ^^ .^^
\^ ^^^-'W'O'^^
K (N 0\h/ 00O-*
cm cm co cm in co
« tv in O CM rf
H -^ co m rH ro
1 Sis
r F 3 to _
u
l.?ll|JlMS"8tlS»
« oo it sr a a * fi.ji 5. 55 55 n
52
GREENFIELD & JOHNSON: BLENNIOID FISHES 53
Stephens, 1970 (Panama, Costa Rica) (distributional records from Smith-Vaniz &
Palacio, 1974). Six of the eight known western Atlantic species of Acan-
themblemaria occur off the Central American coast.
Of the four species of Acanthemblemaria collected, two (A. maria and A. green-
fieldi) have been taken only in Belize, whereas A. aspera and A. spinosa are known
from both Belize and Honduras. Specimens of Acanthemblemaria were obtained
using both rotenone and quinaldine. In the case of A. aspera, A. greenfieldi, and
A. spinosa, a substantial proportion (.27, .41, .60, respectively) of the total indi-
viduals were taken individually by squirting a mixture of quinaldine and iso-
propanol into the worm holes in which they live, thus providing detailed infor-
mation on habitats.
(38) Acanthemblemaria aspera (Longley, 1927)
This species is known from the Bahamas, Florida, Greater Antilles, Dominica,
St. Barthelemy, and Central America, including Belize, Costa Rica, Panama,
Providencia (off Nicaragua), and Yucatan (Smith-Vaniz & Palacio, 1974). This
constitutes the first record for Honduras.
Ecological data. — Acanthemblemaria aspera was taken in both shallow and rela-
tively deep (60 to 80 ft) collections. A majority of records (16 of 25) are from
stations in less than 20 it. Smith-Vaniz & Palacio (1974, p. 215) state that records
for this species extend to 18 m (we have one confirmed record from a station in
which sampling was conducted between 60 and 80 ft), but that most specimens
have been taken in less than 9 m. We have taken A. aspera in a wide variety of
habitats (table 23), but coral was always present, varying from sparse, isolated
heads (a number of collections at Hog Islands) to rich reef development (all
Belize records except one station at Bugle Cay). Acanthemblemaria aspera occurs
more frequently on the dead bases of Montastrea heads than on the bases of
Acropora palmata or in pieces of coral rubble on the bottom (cf. A. greenfieldi, A.
spinosa). Smith-Vaniz & Palacio (1974, p. 215) give the habitat of A. aspera as
"Coral patch reefs, frequently on limestone substrates . . . ." Acanthemblemaria
aspera was taken more frequently than any other chaenopsid species in Belize
and Honduras. In Belize A. aspera was taken at cays inside the Barrier Reef (BC,
SC), at Barrier Reef sites (AC, BRB, BRT, CBC), and at the atolls (GR with
additional records from LR). In Honduras A. aspera was taken at Roatan and the
Hog Islands. Acanthemblemaria aspera was taken with every other chaenopsid
species represented in our collections (table 22) except Coralliozetus cardonae and
Emblemaria hyltoni (each represented at only one station in our collections). This
degree of co-occurrence is unique to Acanthemblemaria aspera among the
chaenopsids and probably reflects (1) the abundance of this species and (2) the
rather eurybathic distribution of this species (over the range of depths sampled
by us). Acanthemblemaria aspera shares highest affinity (GMI) values with Starksia
occidentalis and Labrisomus kalisherae.
Material examined. — A total of 140 specimens from 25 collections. HON-
DURAS: CC, 84341 (2), 84342 (1), 84343 + 84578 (10), 84344 (2), 84346 (1), 84470
(27), 84471 (1); IR, 84345 (20). BELIZE: AC, 82506 (2); BC, 86099 (1); BRB, 86101
(4); BRT, 86102 (1); CBC, 89350 (5), 89401 (16); GR, 77557 (1), 82515 (13), 82516
(1), 82517 (6), 82518 (3), 82520 (3), 82525 (5), 86100 (3), 86103 (1), 86104 (2); SC,
86098 (9). Additional Belize material examined: LR, UMML 9481 (1), UMML 9554
(1).
<J\ CM J* «H g
I I NS^8 I I
I I I I I I I I I I I I I I
I I I I I I I I I I I I I
^o
I I I I I I I I I I I I I I
8g «5 I I I I I I I I
I I
S
I I I I I I I I
8 m8 I I
CM
x
S ^3
I I I
T-H O
I I
X •
I ^3 £8 I I ^8 38
Zino o« ^<5 t^rH
CM tJ< NO rHin HN
rH t-h
t-s rH CD"* NO * 00
o a.
•n a.
v «
6
6C
I 1
W
W
54
GREENFIELD & JOHNSON: BLENNIOID FISHES 55
(39) Acanthemblemaria greenfieldi Smith-Vaniz & Palacio, 1974
This species has been reported from Providencia (off Nicaragua), Belize, Banco
Chinchorro (off Yucatan), and Jamaica (Smith-Vaniz & Palacio, 1974).
Ecological data. — All of our records for A. greenfieldi are from depths less than
15 ft. All records from Glovers Reef (10 of 20 records, 43 of 106 specimens) are
from patch reefs in the lagoon (H3). All other Belize records are from Barrier Reef
sites (AC, BRB, BRT, CBC) and richly developed shallow reef formations (H2,
H3). Smith-Vaniz & Palacio (1974, p. 211) describe the habitat of A. greenfieldi as
"silt-free patch reefs in depths usually shallower than 10 m." Acanthemblemaria
greenfieldi in Belize tends to occur in worm holes in pieces of coral rubble (mostly
Acropora) on the bottom or in large dead pieces of Montastrea and usually is not
found in holes in the dead basal portions of Acropora palmata (cf. A. spinosa).
Acanthemblemaria greenfieldi has been taken with each of the three species of
Acanthemblemaria represented in our collections. Acanthemblemaria greenfieldi
shares highest affinity (GMI) values with Starksia atlantica and Enneanectes
boehlkei (table 22). A detailed discussion of the ecology of A. greenfieldi is pre-
sented by Greenfield & Greenfield (in press).
Material examined. — A total of 106 specimens from 20 collections. BELIZE: AC,
82505 (6); BRB, 86092 (18); BRT, 86090 (3); CBC, 89284 (23), 89308 (1), 89315 (1),
89329 (6), 89340 (1), 89365 (1), 89373 (3); GR, 76348 (1), 76349 (2), 76350 (4), 77559
(4), 82514 (2), 82524 (5), 86091 (8), 86093 (10), 86094 (4), 86095 (3). Additional
Belize material examined: CBC, JWC-23 (4).
(40) Acanthemblemaria maria Bohlke, 1961
This species has been reported from the Bahamas, Grand Cayman Island,
Puerto Rico, St. Barthelemy, Dominica, St. Lucia, and Tobago (Smith-Vaniz &
Palacio, 1974). This constitutes the first record for Belize and a significant range
extension for the species. The two (22.8, 25.1) specimens from Belize agree well
with the description of A. maria provided by Smith-Vaniz & Palacio (1974).
Ecological data. — Acanthemblemaria maria is apparently rare in Belize. We have
collected only two specimens on two different occasions — both collections from
the Barrier Reef (AC, 0 to 10 ft, H2; BRT, 0 to 3 ft, H2). Both specimens were
taken from holes in pieces of coral rubble which had some algal cover. Smith-
Vaniz & Palacio (1974, p. 206) describe the habitat of A. maria as follows:
". . . inhabits limestone slopes of rocky ledges in depths to 9 m, but usually
shallower than 6 m." Both A. aspera and A. greenfieldi were taken at both stations.
Material examined. — Two specimens from two collections. BELIZE: AC, 86096
(1); BRT, 86097 (1).
(41) Acanthemblemaria spinosa Metzelaar, 1919
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Venezuela, Curacao, Albuquerque Cay, Isla San Andres, and Central America,
including Belize and Nicaragua (Nagelkerken, 1974; Smith-Vaniz & Palacio,
1974). This constitutes the first record for Honduras.
Ecological data. — Acanthemblemaria spinosa was taken at only two stations in
Honduras and at 15 stations in Belize. Only one of these 17 stations was in
56 FIELDIANA: ZOOLOGY
relatively deep water (G74-10, GR, 60 to 80 ft, HI, two specimens), the remain-
ing collections were from less than 15 ft. In Belize A. spinosa was taken only in
areas of rich coral development (H2, H3), and 70% of the total individuals (N =
50) were from patch reefs at Glovers Reef or patch-reef-like formations behind
the Barrier Reef and also just behind the reef crest at Carrie Bow Cay. The two
Honduras collections were taken at the Hog Islands in areas of rock, rubble,
sand, and isolated small coral heads (H4). Smith-Vaniz & Palacio (1974, p. 207)
describe the habitat of A. spinosa as "coral patch reefs in depths ranging from 8 to
12 m." The ecology of A. spinosa is discussed by Greenfield & Greenfield (in
press).
In Belize A. spinosa is normally not found in pieces of coral rubble but rather
appears to occur most abundantly in worm holes in the dead basal portions of
large trees of Acropora palmata, appearing to avoid holes in the living upper
portions of the coral. Photographs have been taken of several individuals living
in holes just below the lower margin of the living coral, demonstrating this
restriction to basal dead coral. The coloration of the heads of various individuals
match the background color of the algae around the worm hole, varying from
red to yellow to green or brown. Large amounts of both epiphytic and epizooic
growth in these dead areas not only render the fish more cryptic, but might also
provide greater food resources. Acanthemblemaria spinosa is usually observed
with just its head protruding from the worm hole in which it lives and is seldom
seen to leave the hole. During several hours of observation, individuals of A.
spinosa were seen to leave their holes only twice, each time darting out a distance
of 3 to 6 cm and then returning to the hole. The movement was so rapid that if
feeding occurred it was not observed. When approached (by a diver) while in its
hole, an individual of A. spinosa performs an aggressive display consisting of
opening the mouth widely and slowly swinging the head from side to side; this
display continues until the intruder is within 6 to 10 cm of the fish, at which time
it withdraws into its hole.
In Belize A. spinosa was taken only at Barrier Reef sites (BRB, CBC) and at
Glovers Reef. Acanthemblemaria spinosa was taken with A. aspera and A. greenfieldi
but not A. maria. Acanthemblemaria spinosa shares highest affinity (GMI) values
with Labrisomus guppyi and Enneanectes boehlkei (table 22).
Material examined. — A total of 50 specimens from 17 collections. HONDURAS:
CC, 84347 (5), 84348 (2). BELIZE: BRB, 86085 + 86088 (4); CBC, 89316 (2), 89328
(2), 89339 (3), 89372 (1), 89384 (1); GR, 77558 (1), 82523 (5), 86081 (1), 86082 (8),
86083 (1), 86084 (2), 86086 (4), 86087 (2), 86089 (6).
Coralliozetus Evermann & Marsh, 1899
The genus Coralliozetus contains six species, one in the western Atlantic and
five in the eastern Pacific. The most recent revision of the genus (Stephens, 1963)
was supplemented with the description of an additional species by Stephens et
al. (1966). The sole western Atlantic representative of Coralliozetus is C. cardonae.
(42) Coralliozetus cardonae Evermann & Marsh, 1899
This species has been reported from the Bahamas, Puerto Rico, Virgin Islands,
and Curacao (Stephens, 1963; Stephens et al., 1966; Bohlke & Chaplin, 1968;
Nagelkerken, 1974). This constitutes the first record for Belize and a significant
range extension for the species.
GREENFIELD & JOHNSON: BLENNIOID FISHES 57
We have taken C. cardonae on only one occasion, a reef crest station in shallow
water on the Barrier Reef (G78-7, CBC, 0 to 3 ft, H2). Our material of C. cardonae
agrees well with descriptions of this species provided by Stephens (1963) and
Stephens et al. (1966). Both specimens have two pairs of simple, short orbital
cirri (unique to C. cardonae among western Atlantic chaenopsids), and both
specimens exhibit a marked posteroventrally directed black-edged stripe ex-
tending from the posteroventral margin of the eye. Counts for the two speci-
mens (15.5, 16.0, listed in that order) are as follows: spinous dorsal, 18, 18; soft
dorsal, 10, 11; total dorsal-fin elements, 28, 29; anal fin, 11+ 18, 11+ 19; pectoral
fin, 13, 13. Counts for both dorsal and anal fins differ from modal values listed
for C. cardonae by Stephens et al. (1966, p. 436), and in the case of the 15.5-mm
specimen the counts differ without overlap. With only two specimens of C.
cardonae represented in our collections, we are not prepared at this time to
comment on the significance, if any, of these differences.
Material examined. — Two specimens from one collection. BELIZE: CBC, 89395
(2).
Emblemaria Jordan & Gilbert, 1883
The genus Emblemaria contains 13 species, eight in the western Atlantic and
five in the eastern Pacific. The most recent revision of the western Atlantic
species is that of Stephens (1970); that of the eastern Pacific species is also by
Stephens (1963). One additional western Atlantic species has been described
subsequently (Johnson & Greenfield, 1976).
Only three of the eight described species of Emblemaria are known to occur off
the Central American coast, and these three species occur in our collections from
Belize and Honduras. The following species are not represented in our collec-
tions: E. atlantica Jordan & Evermann, 1898 (Bermuda, Georgia, western Florida);
E. biocellata Stephens, 1970 (Venezuela, French Guiana, Colombia); E. culmenis
Stephens, 1970 (Venezuela); E. diphyodontis Stephens & Cervigon, 1970 (Ven-
ezuela, Colombia); E. piratula Ginsburg & Reid, 1942 (west coast of Florida)
(distributional records are from Stephens, 1970; Palacio, 1974; Shipp, 1975;
Johnson & Greenfield, 1976).
(43) Emblemaria caldwelli Stephens, 1970
This species has been reported from the Bahamas, Jamaica, and Belize
(Stephens, 1970; Johnson & Greenfield, 1976).
Color in life (male). — Taken from a 35-mm Kodachrome of FMNH 82519. Head
rust, overlaid with heavy peppering of distinct black melanophores; jaws and
interorbital darker rust than rest of head. Four yellow-gold spots along posterior
edge of preopercle. Iris yellowish red. Body cream, overlaid by heavy peppering
of distinct black melanophores, belly apparently paler due to larger and more
widely spaced melanophores; a series of cream spots (areas lacking melano-
phores) along midline of body, most numerous on anterior three-fourths of
body, numerous additional spots between midline and dorsal-fin base; reddish
brown pigment visible internally along vertebral column, haemal and neural
spines, and proximal radials of anal and dorsal fins. Caudal-fin base with red-
dish brown pigment on distal margin of hypural plate; caudal fin clear basally,
distal halves of fin membranes black. Pelvic fins cream. Pectoral fins clear, three
white spots on base. Anterior, elevated portion of dorsal fin black, first spine
58 FIELDIANA: ZOOLOGY
with four evenly spaced white bands; posterior portion as well as entire anal fin
cream with heavy peppering of melanophores, melanophores particularly con-
centrated on anterior portion.
Color in life (female). — Taken from a 35-mm Kodachrome of FMNH 86089.
Head yellowish orange, overlaid with peppering of distinct, widely spaced black
melanophores; jaws white, mottled with brown. Iris yellowish red. Belly reddish
brown, remainder of body cream, overlaid with scattering of distinct melano-
phores and yellowish orange spots; posterior half of body translucent, reddish
brown pigment clearly visible internally along vertebral column, haemal and
neural spines, and proximal radials of anal and dorsal fins. Caudal-fin base with
reddish brown pigment on distal margin of hypural plate, remainder of caudal
fin clear. Pectoral and pelvic fins with slight pink tinge. Anal fin clear except for
black pigment on elements and membranes of anterior portion. Spines and
membranes of first five elements of dorsal fin black except for cream coloration
on basal halves of membranes, remainder of fin clear with scattered pinkish bars
on elements.
Ecological data. — Emblemaria caldwelli is a deep-water species. It has not been
taken by us in less than 30 ft, and except for one specimen taken at 30 ft at
Lighthouse Reef, all records and specimens are from depths exceeding 50 ft,
with records extending to 80 to 90 ft. All records are from the reef-front dropoff
zone (HI) at the Barrier Reef (CBC) or the atolls (GR, LR). Emblemaria caldwelli
was not taken in Honduras. One specimen of £. caldwelli was collected with
quinaldine from a worm hole in dead Montastrea that was adjacent to a similar
hole inhabited by a specimen of Acanthemblemaria aspera (G76-28, GR, 50 ft). In
addition to A. aspera, E. caldwelli was taken with four other chaenopsid species
(table 22), but none of the species of Emblemaria were taken together. Emblemaria
caldwelli shares highest affinity (GMI) values with Lucayablennius zingaro and
Paraclinus infrons.
Material examined. — A total of 71 specimens from 17 collections. BELIZE: CBC,
89303 (4), 89313 (7), 89335 (7); GR, 71118 (3), 71135 (1), 71137 (3), 71140 (1), 71152
(2), 77560 (7), 77561 (4), 77563 (3), 77564 + 80414 (16), 82519 (1), 86069 (1), 86070
(2); LR, 77562 (1), 77589 (8).
(44) Emblemaria hyltoni Johnson & Greenfield, 1976 (fig. 4)
This species was described from five (20.4 to 22.9) male specimens taken off
Dixon Cove, Isla Roatan, Honduras (Johnson & Greenfield, 1976). One addi-
tional specimen, a female (FMNH 84349, 23.1 mm SL), was taken with the type
series but was not discovered until the entire collection was sorted (the males
were preserved separately in the field) some months later, too late for inclusion
in the original description. Because species of Emblemaria typically exhibit
marked sexual dimorphism, we present an abbreviated description of the only
known female specimen of E. hyltoni.
Meristic characters. — Total dorsal-fin elements 36 (XXI, 15); total anal-fin ele-
ments 24 (II, 22); pectoral-fin rays 14; pelvic- fin rays 2.
Morphometric characters. — Standard length 23.1 mm. Values expressed as
thousandths of the SL. Head length 281. Head depth 160. Head width 139.
Upper jaw length 109. Eye diameter (fleshy orbit) 70. Snout length 56. Interorbi-
GREENFIELD & JOHNSON: BLENNIOID FISHES 59
tal width (bony) 20. Predorsal length 195. Preanal length 515. Caudal peduncle
depth 75. Caudal peduncle length 78. Pectoral length 143. Pelvic length 294.
Length of first dorsal-fin spine 745. Length of second dorsal-fin spine 299.
Length of last dorsal-fin spine 70. Length of first soft-ray of dorsal fin 100.
Length of orbital cirrus 58. Length of nasal cirrus 28.
Description. — First three dorsal-fin spines in female identical in configuration
with that described for males (Johnson & Greenfield, 1976). Remainder of spin-
ous and soft dorsal-fin elements somewhat shorter in female, e.g., length of last
dorsal-fin spine 70 vs. 84 to 101 (all values for male specimens from Johnson &
Greenfield, 1976), length of first soft ray of dorsal fin 100 vs. 100 to 114.
Pores. — Mandibular 4. Preopercular 5. Posttemporal 4. Supratemporal 3. In-
fraorbital 6. Supraorbital 2. Commissural 1. Anterofrontal 1. Nasal 1. All counts
are for each side except for dorsomedial supratemporal and commissural series.
Angle between supratemporal pores 89°.
Color in alcohol. — Body of female mostly lacking pigmentation (fig. 4), melano-
phores on body limited to the following areas: epaxial area of body just beneath
anterior one-half of spinous dorsal-fin base, a few scattered melanophores on
pectoral- and pelvic-fin bases, scattered melanophores at bases of some dorsal-
and anal-fin elements, scattered melanophores on belly and adjacent to vent.
First three dorsal-fin spines pigmented as in males, remaining rays and mem-
branes of all fins unpigmented. Pigmentation best developed on snout and
especially on lips, with scattered melanophores on occiput, cheeks, and gill
covers. Isthmus and branchiostegal membranes unpigmented.
Sexual dimorphism. — Stephens (1963, p. 169) notes that the genus Emblemaria is
characterized by a high degree of sexual dimorphism. This dimorphism is said to
be most evident in the high sail-like dorsal fin of males vs. the typically low
dorsal of females. Males also usually have longer jaws, longer pelvic fins, longer
orbital cirri, and wider interorbitals than females. In £. hyltoni the most evident
sexually dimorphic character is the relative lack of pigmentation in the female vs.
the strong development of pigmentation in the males. It should be noted that the
female specimen is the largest known individual of £. hyltoni. The configuration
of the first three dorsal-fin spines is the same in both the female and male
specimens, but the remainder of the hard and soft dorsal-fin elements are
somewhat shorter in the female. There is some evidence for sexual dimorphism
in the following morphometric characters [values for the female given first fol-
lowed by the range of values for the five (20.4 to 22.9 mm) males]: upper jaw
length, 109 vs. 127 to 135; orbital cirrus length, 58 vs. 77 to 115; interorbital
width, 20 vs. 26 to 28; nasal cirrus length, 28 vs. 51 to 61; preanal distance, 515
vs. 462 to 483; length of last dorsal-fin spine, 70 vs. 84 to 101. No sexual di-
morphism is evident in length of the first dorsal-fin spine, 745 vs. 676 to 751;
length of the second dorsal-fin spine, 299 vs. 284 to 367; length of the longest
pelvic-fin ray, 294 vs. 267 to 306; or in any of the other morphometric characters
measured.
Material examined. — A total of six specimens from one collection. HON-
DURAS: IR, all specimens from FMNH 75-15 (45 to 60 ft; mistakenly reported as
30.5 m in Johnson & Greenfield, 1976) on the vertical coral face of the dropoff
(HI) just outside the entrance to Dixon Cove, CAS 33511 (1), FMNH 80412 (1),
FMNH 80413 (1), FMNH 84349 (1), UMMZ 200207 (1), USNM 214839 (1).
3
-t
00
X
z
1
1
60
GREENFIELD & JOHNSON: BLENNIOID FISHES 61
(45) Emblemaria pandionis Evermann & Marsh, 1900
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Venezuela, Colombia, Gulf of Mexico (Florida, Texas), and Central America,
including Belize (Stephens, 1970; Palacio, 1974; Smith, 1976).
Color in life. — Taken from a 35-mm Kodachrome of a mature male, FMNH
86071. Head black, with iridescent blue spots on both jaws, ventral surface of
head, and lower portions of gill membranes. Iris red. Body brownish, grading
from dark brown anteriorly to tan centrally and then to a very light tan at
caudal-fin base; scattered iridescent blue spots present along midline of body.
Prepectoral-fin area brown; an iridescent blue line running along bases of fin
rays, remainder of rays and membranes clear. Caudal-fin membranes and fin
rays clear, fin rays crossed with five or six tan bars. Anterior half of anal fin
black, posterior half clear, with slight dusky areas on distal half of fin rays and
membranes. Anterior elevated portion of dorsal fin black; posterior portion
mostly clear, tan along fin base and three or four light tan bars across each
element. Pelvic fins black.
The color pattern of females in life is as given by Stephens (1963) for color in
alcohol, with the body white and the markings brown.
Ecological data. — Emblemaria pandionis was taken only at Glovers Reef and only
in three collections, in depths ranging from 6 to 15 ft. Emblemaria pandionis has
been observed at Glovers Reef on numerous occasions. This species is found at
Glovers Reef only in a very specific habitat, viz., in empty worm or clam holes, in
broken pieces of coral, in channel areas between patch reefs, in the lagoon, and
often in channels in the reef where water enters the lagoon. Aggregations of E.
pandionis are usually in areas of strong current. This species was taken only in
Class II stations. The spectacular display of this species has been reported by
Wickler (1967).
Material examined. — A total of 25 specimens from four collections. BELIZE: GR,
39827 (1), 71136 (16), 82527 (2), 86071 (6).
Emblemariopsis Longley, 1927
The genus Emblemariopsis contains eight species, all limited to the western
Atlantic. The most recent revision of Emblemariopsis is that of Stephens, 1970.
One additional western Atlantic species has been described subsequently
(Greenfield, 1975).
Four of the eight described species of Emblemariopsis occur off the Central
American coast, of which three are represented in our collections from Belize
and Honduras. The following species are not represented in our collections: E.
bahamensis Stephens, 1961 (Bahamas, Lesser Antilles, Grand Cayman Island,
and Albuquerque Cays off Nicaragua); E. bottomei Stephens, 1961 (Venezuela);
E. diaphana Longley, 1927 (Florida); E. occidentalis Stephens, 1970 (Bahamas,
Lesser Antilles); E. randalli Cervigon, 1965 (Venezuela) (distributional records
from Birdsong & Emery, 1968; Stephens, 1970).
(46) Emblemariopsis leptocirris Stephens, 1970
This species has been reported from the Bahamas, Puerto Rico, Virgin Islands,
Antigua, and Grand Cayman Island (Stephens, 1970). This constitutes the first
record for Belize and Honduras and a significant range extension for the species.
62 FIELDIANA: ZOOLOGY
We assign the seven specimens reported here to E. leptocirris on the basis of
the following characters: orbital cirrus present, none of the anteriormost dorsal-
fin spines notably elongate, third dorsal-fin spine not notably shorter than the
first two, dorsal fin not incised between second and fifth spines, anterior dorsal
fin evenly covered with pigment to distal margin, vomerine teeth numbering
about six, in one patch. Stephens' (1970) explanation of how he determined
cephalic laterosensory pore counts is inadequate, and we have not attempted to
duplicate his methods in making pore counts for the western Caribbean speci-
mens. Values for meristic characters follow, based on five intact specimens
(84577, 14.1; 86177, 19.0; 86179, 15.0; 89417, 16.9; 89418, 17.5) with values pre-
sented in that order. Dorsal fin: total elements, 31, 33, 32, 33, 32; spines, 20, 20,
21, 20, 19; segmented rays, 11, 13, 11, 13, 13. Anal fin: spines, 2 (all); segmented
rays, 20, 22, 21, 22, 20. Pectoral-fin rays, 14, 13, 13, 13, 14.
Ecological data. — We have taken E. leptocirris on six occasions over a broad
vertical range, from three stations in less than 5 ft to three stations at depths
exceeding 40 ft (two of the three at depths exceeding 60 ft). Stephens (1970) does
not give depth of capture information for the type series. In Belize E. leptocirris
was taken only in reef -front dropoff zones (HI) or patch reefs (H3) at the Barrier
Reef (CBC) and Glovers Reef. In Honduras E. leptocirris is known only from two
specimens from one station at the Hog Islands (F75-41, 0 to 5 ft, H4). Emble-
mariopsis leptocirris was taken with E. pricei and E. signifera (table 22).
Material examined. — Seven specimens from six collections. HONDURAS: CC,
84577 (2). BELIZE: CBC, 89417 (1), 89418 (1); GR, 86177 (1), 86178 (1), 86179 (1).
(47) Emblemariopsis pricei Greenfield, 1975
This species was described from two (26.0 to 27.0) male specimens taken in
about 4 ft of water on a patch reef at Glovers Reef near Long Cay (Greenfield,
1975). Subsequent to the description of E. pricei, we have taken eight additional
specimens (17.0 to 26.1) in six collections. Five of these were collected at the type
locality at Glovers Reef, Belize, two were taken at the Barrier Reef (CBC), and
one specimen was collected at Big Hog Island, Honduras. Additional informa-
tion on meristic and morphometric characters is presented in the form of ranges
and means for measurements and frequencies for counts and includes values for
the two type specimens. The discovery of two color morphs necessitates presen-
tation of an expanded color description.
Description.— Total dorsal-fin elements: 33 (3), 34 (6), 35 (1), XX, 13 (1), XX, 14
(1), XXI, 12 (2), XXI, 13 (5), XXI, 14 (1); anal-fin elements, II, 22 (6), II, 23 (4);
pectoral-fin rays, 14 (10); vertebrae, 39 (8); head length, 211-259 (236); head
depth, 119-151 (123); head width, 115-135 (123); upper jaw length, 103-129 (115);
orbital length, 42-66 (56); snout length, 27-44 (37); interorbital width, 23-28 (26);
predorsal-fin length, 123-180 (155); preanal-fin length, 379-415 (400); caudal
peduncle depth, 70-86 (77); caudal peduncle length, 78-89 (84); pectoral-fin
length, 146-227 (188); pelvic-fin length, 88-173 (132); 1st dorsal-fin spine length,
54-85 (72); longest dorsal-fin spine, 80-118 (106); terminal dorsal-fin spine, 54-85
(65); 1st dorsal-fin soft ray, 78-105 (93).
Color in life (dark morph). — Taken from a 35-mm Kodachrome of FMNH 86181.
Head black; dorsal surface of body along dorsal-fin base black from nape pos-
GREENFIELD & JOHNSON: BLENNIOID FISHES 63
teriorly to sixth dorsal-fin spine; black gradually grades into grey from sixth
dorsal-fin spine posteriorly to fifteenth dorsal-fin spine; pectoral fins black, dis-
tal portions of lower fin rays cream; pelvic fins black basally and cream distally;
dorsal-fin spines and membranes black from first to eighth spine, posteriorly
grading into grey to first soft ray; soft dorsal-fin membranes clear, fin rays clear
with three or four regularly spaced reddish brown bands; anal-fin spines and
connecting membranes black, membranes between first five soft anal-fin rays
grey, remainder clear, fin rays clear with regularly spaced reddish brown bands;
caudal fin clear; body translucent yellowish cream, nine or ten reddish brown
pigment patches visible internally along vertebral column, interspaced with
yellowish cream areas, reddish brown pigment also visible along neural and
haemal spines in pigmented area of vertebral column and on alternating proxi-
mal radials of anal and dorsal fins.
Color in life (light morph). — Taken from a 35-mm Kodachrome of FMNH 86182.
Head and body virtually transparent, internal structures clearly visible; head
with reddish brown triangular-shaped bar extending from ventral portion of
orbit posteriorly along upper margin of maxillary (also transparent); another
reddish brown area under anteroventral portion of orbit extends ventrally across
both lips; a series of very small bluish white iridescent spots on side of head, one
spot between the two reddish brown areas and remainder posterior to orbit
across preopercle and opercle; the small black spots on side of head as in color in
alcohol; red gill membranes show through opercle; nine reddish brown pigment
patches visible internally along vertebral column, interspaced with transparent
areas; reddish brown pigment also present along every other neural and haemal
spine in the pigmented areas of vertebral column radials of fins alternate with
pigmented neural and haemal spines. Spines and soft rays of dorsal and anal
fins clear with three or four regularly spaced reddish brown bands. Pectoral,
pelvic, and caudal fins clear.
Color in alcohol. — FMNH 86182 (light morph), Figure 5. Head white with the
following black markings: triangular-shaped bar extending from ventral portion
of orbit posteriorly along upper margin of maxillary, widest under eye; interorbi-
tal, snout, anterior portions of lips and chin with a light peppering of
melanophores; a single black spot slightly less than one-fourth pupil diameter
posterior to orbit at level of dorsal margin of pupil, surrounded by a ring of
smaller spots; scattered black spots on nape, preopercle, and opercle; a distinct
black spot (about same size as postorbital spot) at junction of preopercle and
opercle at the level of ventral margin of orbit. Body white, lacking pigmentation,
except for one or two small black spots on pectoral-fin base. Pectoral, pelvic, and
caudal fins clear. Anal fin clear except for a few faint brown areas on some fin
rays. Dorsal fin clear except for a few faint brown areas on some fin rays and a
few scattered small black spots on fin membranes between first five spines.
Color morphs. — Emblemariopsis pricei occurs in two color morphs (fig. 5), dark
and light. Individuals collected from holes in living coral all have black heads
and dark bodies, whereas those collected outside of holes, resting on or swim-
ming over the coral, are all virtually transparent. The paler color morph was
discovered, incidentally, while attempting to collect a dark morph from a hole in
a large head of Montastrea, using quinaldine, at a depth of 100 ft at Glovers Reef.
Two small, transparent fishes were observed actively swimming along the sur-
64
GREENFIELD & JOHNSON: BLENNIOID FISHES 65
face of the coral head and darting around the entrance to the hole. One of the
transparent fishes was collected with quinaldine and was later identified as
£. pricei.
All of the known specimens of the dark morph are males. Examination of the
two light morphs revealed that, in contrast to our expectations, these specimens
were also males with well-developed genital papillae. This unusual situation
may correlate with the acquisition of a hole by males. Individuals that inhabit
holes have jet-black heads, in sharp contrast to the background of living coral in
which they live. This distinctive coloration may be related to aggressive territo-
rial display toward other males, attracting females to holes, or both. The nonter-
ritorial males with their transparent bodies are very cryptic. Males with black
heads would be conspicuous and thus vulnerable to predation if resting in the
open on top of a coral head. It seems likely that the dark coloration is not
attained until an individual is able to establish itself in a hole.
Ecological data. — Emblemariopsis pricei is apparently eurybathic over the depth
range represented by our samples. This species has been taken in less than 4 ft
and as deep as 100 ft. The common factor in all six stations is the presence of
richly developed extensive coral formations (table 23; the one station labeled H9
is F75-51, a very broad station in the Hog Islands in which both rock and coral
formations were sampled). At Glovers Reef £. pricei was taken twice in living
Acropora palmata (in depths of 2 to 4 ft) and twice in living Montastrea
(at depths of 50 to 100 ft). Emblemariopsis pricei was taken with E. leptocirrus
(table 22).
Material examined. — A total of 10 specimens from seven collections. HON-
DURAS: CC, 84580 (1). BELIZE: CBC, 89419 (1), 89420 (1); GR, 77481 + 77482 (2)
(holotype and paratype), 86180 (2), 86181 (2), 86182 (1).
(48) Emblemariopsis signifera (Ginsburg, 1942)
This species is known from the Bahamas, Lesser Antilles, near Rio de Janeiro
(Brazil), Curacao, Colombia, and Grand Cayman Island (Stephens, 1970; Nagel-
kerken, 1974; Palacio, 1974). This constitutes the first record for Belize and Hon-
duras and a significant range extension for the species.
Ecological data. — We have taken E. signifera on five occasions, and in each
collection only one specimen was present. Three of the four records from Belize
[those from CBC (2), GR (1)] are from 60 to 80 ft in the reef-front dropoff zone
(HI). The fourth record from Belize (BRG) is from a station in less than 6 ft on the
reef crest (H2). The one record from Honduras (CC, 50 ft, HI) is from a deep-
water richly developed coral formation at Northwest Cay. Emblemariopsis signif-
era was taken with E. leptocirris and shares highest affinity (GMI) values with
Lucayablennius zingaro (table 22).
Material examined. — A total of five specimens from five collections. HON-
DURAS: CC, 84579 (1). BELIZE: BRB, 86072 (1); CBC, 89421 (1), 89422 (1); GR,
86073 (1).
Hemiemblemaria Longley & Hildebrand, 1940
The genus Hemiemblemaria is monotypic.
66 FIELDIANA: ZOOLOGY
(49) Hemiemblemaria simulus Longley & Hildebrand, 1940
This species has been reported from Florida and the Bahamas (Stephens, 1963;
Bohlke & Chaplin, 1968). This constitutes the first record for Belize and Hon-
duras and a significant range extension for the species.
Hemiemblemaria simulus has been taken twice in Honduras at the Hog Islands
and once at Tom Owens Cay on the Barrier Reef in Belize. Uncatalogued collec-
tions at UMML were from two locations on the Barrier Reef in Belize. All collec-
tions are from depths of less than 10 ft and were from areas of coral rubble and
scattered isolated heads of living coral.
Material examined. — A total of five specimens from three collections. HON-
DURAS: CC, 84359 (3), 84544 (1). BELIZE: BRT, 86171 (1). Additional Belize
material examined: UMML, uncatalogued material collected by P. Colin:
Queen's Cay, 26 Oct. 1972 (1); Tobacco Reef, 28 Oct. 1972 (2).
Lucayablennius Bohlke, 1957
The genus Lucayablennius is monotypic.
(50) Lucayablennius zingaro Bohlke, 1957
This species has been reported from the Bahamas, Jamaica, Colombia, and the
western Caribbean, including Belize and Panama (Bohlke & Chaplin, 1968;
Greenfield, 1972; Colin & Gomon, 1973; Palacio, 1974). This constitutes the first
record for Honduras.
Ecological data. — Lucayablennius zingaro is a deep-water species. We have taken
it only in distinct reef-front dropoff zones in areas of rich coral development
(HI). All specimens were taken at depths exceeding 45 ft and at depths as great
as 80 to 110 ft, the limit of our sampling efforts. Lucayablennius zingaro was
observed at depths of 70 m off Belize and 106 m off Jamaica by Colin (1974). A
majority of records (14 of 24 records) and specimens (49 of 78) are from Glovers
Reef. Other Belize specimens are from the Barrier Reef (CBC) and Lighthouse
Reef (LR). In Honduras L. zingaro was taken only at the two deepest stations,
FMNH 75-15 (IR, 45 to 60 ft) and FMNH 75-48 (CC, 50 ft). Lucayablennius zingaro
was taken only with chaenopsid species that occur in deep water, viz., Acan-
themblemaria aspera (1 station), Emblemaria caldwelli (12 stations), E. hyltoni (1
station), Emblemariopsis leptocirris (1 station), and E. signifera (4 stations) (table
22). Lucayablennius zingaro shares highest affinity (GMI) values with Enneanectes
atrorus and Emblemaria caldwelli.
Material examined. — A total of 78 specimens from 24 collections. HONDURAS:
CC, 84542 (2); IR, 84541 (5). BELIZE: CBC, 89302 (3), 89312 (4), 89326 (1), 89333
(2), 89358 (3), 89360 (2); GR, 70823 (3), 70824 (4), 70825 (3), 70826 (11), 70827 (6),
70828 (1), 71008 (2), 71119 (1), 71141 (2), 71153 (3), 77567 (2), 86076 (4), 86077 (2),
86176 (5); LR, 77565 (2), 77566 (5).
BLENNIIDAE
Of the four families of tropical western Atlantic blennioid fishes treated in
this paper, the Blenniidae is unquestionably in greatest need of thorough review
and revision. We have not attempted to compile figures on the total number of
GREENFIELD & JOHNSON: BLENNIOID FISHES 67
Table 24. Co-occurrence of blenniid species. Values in upper half of matrix (whole
numbers) are actual number of records (number of stations) of co-occurrence (based on all
stations). Values in lower half of matrix are calculated GMI values (based on Class I
stations only).
Co-occurring species
Species
NT
N,
51
52
53
54
55
56
57
51. E. nigricans
52. H. aequipinnis
53. H. springeri
54. L. dispar
55. 0. atlanticus
21
3
7
4
20
3
5
4
0
.29
0
0
0
0
4
0
0
0
0
0
14
1
5
0
0
1
0
0
5
1
3
0
macclurei
30
30
.48
.01
.32
0
—
0
6
56. P. marmoreus
1
1
0
.29
0
0
0
—
0
57. S. cristata
7
7
.31
.03
.32
0
.32
0
—
Listing of species
only).
sharing highest GMI values with blenniid species
(five highest v,
dues
51. E. nigricans
52. H. aequipinnis
53. H. springeri
54. L. dispar
55. 0. atlanticus
9 (.53), 7 (.51), 4 (.49), 35 (.49), 37 (.49)
56 (.29), 11 (.04), 57 (.03), 8 (.02), 14 (.02)
57 (.32), 55 (.32), 49 (.29), 51 (.29), 37 (.29)
Not taken with any other blennioid species
macclurei
56. P. marmoreus
57. S. cristata
37 (.64),
52 (.29),
8 (.54),
4 (.62),
6 (.53),
15 (.57),
12 (.46),
35 (.55),
21 (.37),
51 (.48)
-(-)
27 (.37)
NT = total number of stations; N = number of Class I stations.
either species or genera in the tropical western Atlantic for this family. Names
used in this paper are in accord with the available literature (Jordan & Ever-
mann, 1898; Meek & Hildebrand, 1928; Norman, 1943; Tavolga, 1954; Springer,
1967, 1968; Springer & Smith-Vaniz, 1970; Bohlke & Chaplin, 1968; Randall,
1968; Dawson, 1970; Bath, 1977). Co-occurrence data for blenniid species is
presented in Table 24.
Entomacrodus Gill, 1859
There are four species of Entomacrodus in the Atlantic Ocean, but only E.
nigricans is known to occur in the Caribbean Atlantic area (Springer, 1967, 1972).
(51) Entomacrodus nigricans Gill, 1859
This species is known from Bermuda, the Bahamas, Florida, Greater and
Lesser Antilles, Venezuela, Curacao, Colombia, and Central America, including
Belize, Costa Rica, Nicaragua, Panama, and Yucatan (Rubinoff & Rubinoff, 1962;
Caldwell, 1963; Caldwell & Caldwell, 1964; Cervigon, 1966; Springer, 1967;
Birdsong & Emery, 1968; Bohlke & Chaplin, 1968; Mago, 1970; Nagelkerken,
1974). This constitutes the first record for Honduras.
Ecological data. — Entomacrodus nigricans was taken in only one station in which
sampling extended deeper than 10 ft (F75-45, 2 to 18 ft, H2), and the vast
majority of records (16 of 21) and specimens (308 of 367) are from less than 5 ft.
Entomacrodus nigricans was taken in a variety of shallow-water habitats (table 25).
Nearly half of the records (10 of 21) were in shallow fringing reefs or reef crest
(H2) areas (including stations at CC, IR, BRT, CBC, GR, LR). A large majority of
G
c
(8
C
tn
ID
s-
s
o>
XI
c
E
$
3
01
>-
C
Oh
01
o
1-.
c
'VI
JC
o
•n
*
O <N 00 t-i rH
I I
o £
■£ *
>^ o
o c
Si to
TO
TO > 2
3
s 8
0> TO
i-c dj
x;
u in
- B
IA> TO
111
a, . 6
:S ^ .h
c .3
c <-> "to
J2i C 2
-s g I
O "3 01
u 3 i
oi -a p
5-5 5
^•S-i
*OT3
° a "?
•X3 O ,C
sis
W OB
P 5
u ex,
■>* CM rH CO
co cr> -^f m
CO rH rH
B 9
I I
H N CO00 t-xON ^ ON ON N tn
i rH CO K CO
CO
IT)
•**
5
01
■ 2?
a
e
en
'C
«X5
•43 3
5- JS
s-6
s?
68
GREENFIELD & JOHNSON: BLENNIOID FISHES 69
the total specimens (88% of 367) came from reef crest pools (H6) at Glovers Reef
or from shallow, rocky areas (H4) at the Hog Islands (including one exception-
ally broad station, F75-51, in which both rock and coral reefs were sampled). In
the Bahamas, according to Bohlke & Chaplin (1968, p. 563), E. nigricans
". . . lives in tide pools, on rocky slopes, and in locations where there are
boulders on the bottom." All but one of the records for this species from Glovers
Reef are from tide pool areas.
At Long Cay, Glovers Reef, we have observed E. nigricans at night (9:30 PM)
resting in small depressions in pieces of coral rubble along the shore, which were
completely out of the water. Their bodies were only partly covered with water,
and respiratory movements were very rapid. Stomach analyses of these indi-
viduals showed that they were not feeding.
Entomacrodus nigricans has been taken with three other blenniid species (table
24), but a number of these records of co-occurrence may reflect a basic artifact of
our sampling method (it is still the only practical method), viz., the inclusion of
many species in one sample which may differ greatly in microhabitats. For
example, E. nigricans has been taken seven times in Belize at H2 stations at which
Ophioblennius atlanticus macclurei was also taken. All of these stations were rela-
tively broad, ranging from close to the shoreline out into slightly deeper water
(less than 5 ft maximum depth, except for one station extending to 15 ft
maximum depth). At Glovers Reef, O. a. macclurei appears to replace E. nigricans
in the deeper water areas of the wave-swept shore, away from the intertidal
area. Thus the relatively high calculated affinity value (GMI = 0.480) shared by
these two species may in part reflect the scale over which our samples were
taken. Entomacrodus nigricans shares highest affinity (GMI) values with Malacoc-
tenus aurolineatus and Labrisomus nigricinctus (table 24).
Material examined. — A total of 367 specimens from 21 collections. HON-
DURAS: CC, 84556 (82), 84557 (44), 84558 (1), 84559 (5), 84560 (22), 84561 (38),
84562 (16); IR, 84555 (12). BELIZE: BRT, 86172 (5); CBC, 89282 (7), 89323 (3),
89344 (5), 89377 (35); GR, 77529 (14), 77574 (1), 77576 (9), 77801 + 89423 (14),
86074 (1), 86075 (83); LR, 77573 (1), 77575 (1). Additional Belize material: CBC,
JWC-18 (1); LR, UMML 9981.
Hypleurochilus Gill 1861
The most recent partial revisions of Hypleurochilus are those of Randall (1966)
and Bath (1977). According to Bath there are seven species of Hypleurochilus, four
of which occur in the Caribbean Atlantic area. Two of these four species are
unknown from the western Caribbean: H. geminatus Wood (South Atlantic and
Gulf coasts of the United States, a complex consisting of more than one species
according to Randall) and H. bermudensis Beebe & TeeVan (Bermuda, the
Bahamas, Florida; records from Jordan & Evermann, 1898; Randall, 1966; Bohlke
& Chaplin, 1968). The two species remaining, H. aequipinnis (Gunther) and H.
springeri Randall, are represented in our collections from Belize and Honduras.
(52) Hypleurochilus aequipinnis (Gunther, 1861)
This species was described from one specimen taken in west Africa. In the
western Atlantic H. aequipinnis is known from the Bahamas, Florida, Greater and
Lesser Antilles, Venezuela, and Central America, including Guatemala and
70 FIELDIANA: ZOOLOGY
Yucatan (Cervigon, 1966; Randall, 1966; Bohlke & Chaplin, 1968; Mago, 1970;
Bath, 1977). This constitutes the first record for Belize and Honduras.
Ecological data. — Hypleurochilus aequipinnis has been collected by us once in
Belize (G74-14, FC) in mangrove roots (H9) at a cay inside of the Barrier Reef,
once in a mangrove channel (H9) at Roatan, and once on a pier piling (H8) at
Trujillo, Honduras. All three stations were at depths of 3 to 10 ft. In Belize H.
aequipinnis was taken with Parablennius marmoreus (at FC) and in Honduras was
taken with Scartella cristata and Ophioblennius atlanticus macclurei at Trujillo.
Material examined. — A total of 18 specimens from three collections. HON-
DURAS: IR, 84569 (2); mainland: main pier at Trujillo, 84570 (13). BELIZE: FC,
86170 (3).
(53) Hypleurochilus springeri Randall, 1966
This species is known from the Bahamas, Florida, Greater and Lesser Antilles,
Venezuela, Curacao, Grand Cayman Island, and Central America, including
Albuquerque Cay off Nicaragua (Randall, 1966; Birdsong & Emery, 1968; Bohlke
& Chaplin, 1968; Mago, 1970; Nagelkerken, 1974). This constitutes the first re-
cord for Belize and Honduras.
Ecological data. — We have taken H. springeri on seven occasions at stations in
which sampling extended (maximally) to 15 ft. Although the Honduras collec-
tions, all rotenone stations, suggest co-occurrence with three other blenniid
species (table 24), this may in part reflect the scale of sampling (see discussion
under Entomacrodus nigricans). In two of the four Belize records, stations in which
sampling was done using quinaldine, specimens of H. springeri were found only
on the upper portions of living trees of Acropora palmata. Only in the Belize
collections in which rotenone was used (CBC, G 78-13; GR, G 73-50) were two
additional blenniid species taken — Entomacrodus nigricans and Ophioblennius
atlanticus macclurei. Bohlke & Chaplin (1968, p. 568) describe the habitat of H.
springeri as follows: "All known specimens are from water less than 10 ft deep,
most from rocky bottom along a sandy shoreline, in clear water that is relatively
calm." Except for the sandy shoreline, this sentence, as far as it goes, could
describe the H4 stations in Honduras at which H. springeri was taken. In Belize
H. springeri was taken at the Barrier Reef (CBC) and Glovers Reef. In Honduras
H. springeri was taken at the Hog Islands. Hypleurochilus springeri shares highest
affinity (GMI) values with Scartella cristata and Ophioblennius atlanticus macclurei.
Material examined. — A total of nine specimens from seven collections. HON-
DURAS: CC, 84571 (2), 84572 (1), 84573 (1). BELIZE: CBC, 89386 (1); GR, 86167
(1), 86168 (2), 86169 (1).
Lupinoblennius Herre, 1942
According to Bath (1977) the genus Lupinoblennius contains two species:
L. nicholsi (Tavolga, 1954) (known from Florida and the Gulf of Mexico at
Tamaulipas, Mexico) and L. dispar.
(54) Lupinoblennius dispar Herre, 1942
This species is known from Jamaica, Antigua, Trinidad, and Central America,
including Guatemala, Nicaragua, Panama, and Yucatan. In Panama L. dispar has
GREENFIELD & JOHNSON: BLENNIOID FISHES 71
also been taken in the Miraflores Third Lock, Panama Canal, Pacific side (records
from Dawson, 1970). This constitutes the first record for Belize and Honduras
and is an addition to Miller's (1966, 1976) lists of the freshwater fishes of Central
America.
Ecological data. — Lupinoblennius dispar is the only species of blennioid fish taken
by us that was not taken with any other blennioid species. All collections of L.
dispar are from the mainland (H8), from shallow water (less than 4 ft), and from
low salinity areas. The specimen taken at Gales Point (86080) was taken in
freshwater.
Material examined. — A total of 19 specimens from four collections. HON-
DURAS: mainland: Brus Lagoon, Cannon Island, 84550 (7). BELIZE: mainland:
beach at St. John's College, Belize City, 86079 (5); mangrove shore just outside
entrance to boat canal at St. John's College, Belize City, 89412 (6); southern
lagoon at Gales Point, 86080 (1).
Ophioblennius Gill, 1860
The genus Ophioblennius contains two species, each represented by two sub-
species (Springer, 1962). In the eastern Pacific are O. steindachneri steindachneri
Jordan & Evermann, 1896 (widespread in eastern Pacific, California to Peru) and
O. s. clippertonensis Springer, 1962 (Clipperton Island). In the Atlantic are O.
atlanticus atlanticus (Valenciennes, 1836) (eastern Atlantic and Brazil) and O. a.
macclurei (Silvester, 1915) (widespread in Caribbean Atlantic area).
(55) Ophioblennius atlanticus macclurei (Silvester, 1915)
This subspecies is known from Bermuda, the Bahamas, North Carolina, Geor-
gia, Florida, Greater and Lesser Antilles, Venezuela, Curacao, and Central
America, including Belize, Costa Rica, Nicaragua, Panama, and off Yucatan
(Rubinoff & Rubinoff, 1962; Springer, 1962; Caldwell, 1963; Cervigon, 1966;
Birdsong & Emery, 1968; Bohlke & Chaplin, 1968; Mago, 1970; Nagelkerken,
1974). This constitutes the first record for Honduras. All specimens counted
(N = 15) in our western Caribbean material agree with Springer's (1962) values
for O. a. macclurei in having 31 or 32 total dorsal-fin elements and 22 or 23 total
anal-fin elements.
Ecological data. — We have taken Ophioblennius atlanticus macclurei in stations in
which sampling extended to 20 ft, but the great majority of records (28 of 30) and
specimens (338 of 377) are from stations in less than 15 ft, and 21 of 30 records are
from stations in less than 6 ft. Ophioblennius atlanticus macclurei was taken in a
variety of shallow- water habitats (table 25). In Belize a majority of records (12 of
21) are from rich shallow wave-swept coral reef sites (H2). Two habitats, H2 and
H4, accounted for 94% of the 377 specimens taken (includes the two exception-
ally broad stations at the Hog Islands, F 75-47 and F 75-51). Although rep-
resented by more specimens in our collections than any other blennioid species
(table 1), abundance figures for O. a. macclurei are underestimates in that in a
number of cases not all of the individuals killed at a particular rotenone station
were collected (the collectors were greatly limited by time and manpower, and
thus "rare" species may well be overrepresented in terms of abundance figures).
A detailed description of the habitat preferences and territorial behavior of O. a.
72 FIELDIANA: ZOOLOGY
macclurei is provided by Nursall (1977). We have taken O. a. macclurei in Belize at
cays inside the Barrier Reef (BC, FC, SC, TC), at Barrier Reef sites (BRG, BRT,
CBC), and at the atolls (GR, LR). In Honduras we have taken this species at the
Hog Islands and on the mainland (pier at Trujillo). Ophioblennius atlanticus
macclurei was taken with four other blenniid species (table 24) and shares highest
affinity (GMI) values with Enneanectes pectoralis and Labrisomus guppyi.
Material examined. — A total of 377 specimens from 30 collections. HON-
DURAS: CC, 84351 (65), 84352 (9), 84353 (3), 84354 (27), 84355 (4), 84356 (12),
84357 (3), 84358 (19); mainland: main pier at Trujillo, 84350 (1). BELIZE: BC,
86159 (2); BRG, 86078 (32), 86163 + 86175 (2); BRT, 86162 (16); CBC, 89287 (23),
89310 (7), 89324 (7), 89352 (34), 89367 (2), 89383 (1), 89397 (28); FC, 86173 (2); GR,
77530 (2), 77531 (1), 77532 (1), 77578 (15), 77579 (18); LR, 77577 (2), 77802 (36); SC,
86161 (1); TC, 86160 (2). Additional Belize material examined: CBC, JWC-11 (1);
LR, UMML 9557; SWC, JWC-9 (1), JWC-24 (4).
Parablennius Riberio, 1915
According to Bath (1977), the genus Parablennius contains nine species, only
one of which, P. marmoreus, occurs in the Caribbean Atlantic area.
(56) Parablennius marmoreus (Poey, 1875)
This species4 has been reported from the Bahamas, Florida, Greater and Les-
ser Antilles, Trinidad, Venezuela, and the Gulf of Mexico. It has also been
reported from the east coast of the United States as far north as New York
(Cervigon, 1966; Bohlke & Chaplin, 1968; Randall, 1968; Mago, 1970; Smith,
1976; Bath, 1977). This is apparently the first record of this species for Belize and
for the western Caribbean.
Ecological data. — Our single specimen of P. marmoreus was taken at French-
man's Cay (Belize) in less than 7 ft of water in an area of mangrove roots
(covered with an attached fauna of sponges and tunicates) and a soft mud
bottom (H9). Three specimens of Hypleurochilus aequipinnis were taken at the
same station (G 74-14).
Material examined. — A total of one specimen from one collection. BELIZE: FC,
86165 (1).
Scartella Jordan, 1886
According to Bath (1977) there are four (possibly five) species assignable to the
genus Scartella, of which only one, S. cristata, occurs in the western Atlantic.
(57) Scartella cristata (Linnaeus, 1758)
This species is known from both sides of the Atlantic. In the western Atlantic it
is known from Bermuda, the Bahamas, Florida, Greater and Lesser Antilles, and
south to Brazil, Venezuela, and the Gulf of Mexico (Pinto, 1954; Cervigon, 1966;
Bohlke & Chaplin, 1968; Randall, 1968; Mago, 1970). It has been reported from
Catelonia Bay, Panama (Breder, reported in Meek & Hildebrand, 1928). This is
apparently the first record for Belize and Honduras.
4C. R. Gilbert (personal communication) has informed us of the likelihood that more
than one species is currently subsumed under this name.
GREENFIELD & JOHNSON: BLENNIOID FISHES 73
Ecological data. — All of our records for S. cristata are from less than 10 ft of
water, and five of seven records are from less than 5 ft of water. We have
collected S. cristata on only one occasion in Belize [SC, G 74-16 (1)] in an area of
coral rubble and algae (H5). We have collected S. cristata on five occasions at the
Hog Islands (table 25) in areas of coral rubble and rock with or without adjacent
scattered heads of live coral (H4, and H9 in the case of F 75-51, an exceptionally
broad station). Scartella cristata was also taken from a pier piling at Trujillo on the
mainland (H8). Scartella cristata was taken with four other blenniid species: En-
tomacrodus nigricans (CC only), Hypleurochilus aeauipinnis (Trujillo pier only), H.
springeri (CC only), and Ophioblennius atlanticus macclurei (CC, SC) (table 24).
Scartella cristata shares highest affinity (GMI) values with Labrisomus nuchipinnis
and L. kalisherae.
Material examined. — A total of 33 specimens from seven collections. HON-
DURAS: CC, 84564 (11), 84565 (4), 84566 (3), 84567 (1), 84568 (10); mainland:
main pier at Trujillo, 84563 (3). BELIZE: SC, 86166 (1).
DISCUSSION
We have attempted to elucidate and study certain general patterns in the
distribution of the 57 blennioid species we have collected in Belize and Hon-
duras. We were particularly interested in (1) comparisons between localities
directed toward quantifying observed inshore to offshore changes in the com-
position of blennioid species assemblages and in species richness, (2) determin-
ing the composition of groups of blennioid fishes that commonly co-occurred in
our samples, (3) examining locality data for possible environmental correlates of
the distribution of the groups recognized, and (4) examining the composition of
the blennioid fauna of Belize and Honduras in terms of supposed broadscale
distribution patterns of tropical western Atlantic shorefish species.
Comparisons Between Localities
Unevenness of sampling effort and especially the paucity of collections at
inshore sites forced us to pool data prior to attempting comparisons between
localities. We chose to group the 18 specific collecting sites (table 3) at which we
captured blennioid species into six "general" localities, two in Honduras — (A)
Hog Islands and (B) Roatan; one shared by Belize and Honduras — (C) mainland
localities; and three in Belize — (D) cays inside of the Belize Barrier Reef, (E)
Belize Barrier Reef sites, and (F) atolls. We refer to localities C and D as "in-
shore" localities; localities A, B, E, F as "offshore" localities. We have compiled
species lists (based on all 127 stations) for each of the six general localities and
determined the number of species shared for each of the 15 possible pairwise
comparisons (table 26). These data allow computation of similarity coefficients
(SI) for each of the pairwise comparisons, and values for the Jaccard Coefficient
(table 26) are presented in rank order in Table 27.
Three of the general localities, (A) Hog Islands, (E) Belize Barrier Reef sites,
and (F) atolls, stand out (tables 3, 26) in sharing three attributes: (1) the greatest
total number of species (respectively, 38, 38, 39) were taken at these three
localities, (2) the greatest proportion of sampling effort (107 of 127 stations) was
expended at these three localities, and (3) the three highest totals for number of
shared species and three largest shared SI values are represented in the three
74
FIELDIANA: ZOOLOGY
Table 26. Similarity between various localities in Belize and Honduras based on num-
bers of shared species. Each "general locality" indicated includes the same specific sam-
pling sites listed in Table 3, except that results for the Hog Islands and Roatan are given
separately. The upper half of the matrix indicates number of shared species for each pair of
localities. The lower half of the matrix gives calculated SI values for each pair (based on
Jaccard Coefficient, see text for additional explanation).
Locality
N
Locality
A
B
C
D
E
A.
Honduras: Hog Ids.
38
A
—
19
4
13
30
B.
Honduras: Roatan
25
B
.43
—
3
10
19
C.
Belize and Honduras:
mainland localities
7
C
.10
.10
_
5
2
D.
Belize: cays inside
of Barrier Reef
16
D
.32
.32
.28
_
10
E.
Belize: sites on
Barrier Reef
38
E
.65
.43
.05
.23
_
F.
Belize: atolls
39
F
.54
.46
.05
.22
.71
N = total number of species taken (based on all stations).
Table 27. Correspondence between calculated SI values for gen-
eral localities in Belize and Honduras and the total number of
stations occupied at each pair of localities. Localities are those de-
fined in Table 26. The total number of stations (NST) is the
summed total number of stations at each member of a pair of
localities.
Calculated
Total number
SI values
of stations
icalities
SI
Rank
NST
Rank
E-F
.711
1
96
1
A-E
.652
2
38
6
A-F
.540
3
80
2
B-F
.455
4
77
3
A-B
.432
5.5
19
10
B-E
.432
5.5
35
7
B-D
.323
7
13
14
A-D
.317
8
16
12
C-D
.278
9
12
15
D-E
.227
10
32
9
D-F
.222
11
74
5
B-C
.103
12
15
13
A-C
.098
13
18
11
C-E
.047
14
34
8
C-F
.045
15
76
4
SI vs. NST, Tau15 = 0.267, .10 < p < .20.
possible comparisons between these localities (tables 26, 27). These facts might
imply that SI values calculated in pairwise comparisons may directly reflect
sampling effort. Given the unevenness of sampling effort represented in our
data, this result would not be surprising but neither would it be of any biological
interest. We tested for a direct relationship between SI values and sampling
effort (measured by number of stations) over the 15 pairwise comparisons (table
27). The resulting tau value (Kendall's rank-correlation coefficient, corrected for
GREENFIELD & JOHNSON: BLENNIOID FISHES 75
ties, see Tate & Clelland, 1957) is not statistically significant, Tau15 = + 0.267, .10
< p < .20. We use this result to justify, with hesitation, further discussion.
There appears to be an inshore to offshore gradient in SI values, with the
mainland locality clearly the most distinctive (fig. 6). The mainland locality (C)
shares a moderately high SI value (0.278) only with locality (D) cays inside of the
Belize Barrier Reef and shares very low SI values with the other localities (tables
26, 27). If these results are not explainable solely in terms of differences in
sampling effort at different localities, two additional explanations come to mind:
(1) that the inshore to offshore gradient in SI values reflects an inshore to
offshore gradient in species richness (total number of blennioid species present),
and (2) that the inshore to offshore gradient reflects changes in the species of
blennioid fishes present at the different localities.
Both explanations are probably true. There is no doubt that many more
species of blennioid fishes occur at the offshore, insular localities than occur at
the inshore, continental localities we have sampled in Belize and Honduras. We
are certain that additional sampling at the Hog Islands and at Roatan would
increase the SI values shared by each with each other and with the offshore
localities in Belize (e.g., there was only one station each at the Hog Islands and at
Roatan in which sampling extended to depths greater than 20 ft). We greatly
doubt, however, that additional sampling would substantially increase the SI
values shared by the mainland and offshore localities. This is particularly true for
Belize where the entire mainland coast is fringed by mangrove swamps.
In addition to the apparent inshore to offshore gradient in species richness
there is also some evidence for changes in species actually present along the
inshore to offshore axis. For example, despite the great expenditure of sampling
effort at offshore (Barrier Reef + atolls) sites in Belize (76% of all stations, 91% of
all Belize stations), seven species taken at mainland sites and/or sites at cays
behind the Barrier Reef were never taken at Barrier Reef sites or at the atolls. The
seven species are as follows: (6) Labrisomus kalisherae, (11) Malacoctenus delalandei,
(18) Paraclinus fasciatus, (52) Hypleurochilus aequipinnis, (54) Lupinoblennius dispar,
(56) Parablennius marmoreus, and (57) Scartella cristata. Were it not for one record
(and one specimen) from the Barrier Reef at BRG, (27) Starksia occidentalis would
be a member of this group. All of these species (except Parablennius marmoreus,
represented in our collections by a single specimen) were also taken at Roatan or
the Hog Islands or both. Lupinoblennius dispar is unique among the 57 species
reported in this paper in that it was taken only at mainland localities and only in
brackish water or freshwater.
These results support our belief that part of the inshore to offshore gradient in
SI values is related to changes in blennioid species actually occurring at different
sites along the inshore to offshore axis. This would be in general agreement with
discussions of the continental vs. insular shorefish faunas provided by Gilbert &
Kelso (1971), Robins (1972), and Gilbert (1973).
Part of the explanation for the trend in species richness and for changes in the
composition of blennioid assemblages along the inshore to offshore axis might
relate to differences in diversity of available habitats along the inshore to
offshore axis. Unfortunately, given the distribution of sampling effort rep-
resented in our collections (table 4), any attempt to test this suggestion will
depend on our opportunity to considerably broaden our collection base, par-
ticularly at inshore sites.
HI
m
o
CM
10
CO
10
10
CO
CM
CO
CO
CM
^
o
o
o
o
(O
o
o
10
CO
o
o
o
o
CO
o
H o
CM
s
T3UJ-
76
GREENFIELD & JOHNSON: BLENNIOID FISHES 77
Recurrent Group Analysis
Our usage of recurrent group analysis had two purposes: (1) to determine the
composition of groups of blennioid fishes that commonly co-occurred in our
samples, and (2) to provide a basis for examination of locality data for those
stations at which a large proportion of the members of a recurrent group were
taken to determine possible environmental features (site, depth, substrate, etc.)
associated with the distribution of the groups formed by the analysis. In at-
tempting to determine groups of "associated" blennioid fishes in Belize and
Honduras, i.e., groups of species "commonly" taken together in the same sta-
tion, we had to choose between using presence and absence data or abundance
data (number of individuals/station, either absolute or by rank) or both. Much of
the recent work on fish associations in which attempts have been made to
determine the composition of groups of associated species (or species similar in
habitat requirements) have depended on the use of abundance data (representa-
tive examples: Clarke, 1977; Ebeling et al., 1970; Echelle & Schnell, 1976; Smith &
Powell, 1971). There is enormous potential for bias on a station-to-station basis
in broadcast ichthyocide collecting, and possible (in fact likely) bias in our data
was compounded because the initial purpose of collecting efforts in Belize was
faunistic rather than ecological (resulting in, among other things, the likelihood
of under-representation of "common" species in our samples). For these reasons
we felt that the use of a grouping strategy based on presence and absence data
was likely to be more reliable in the present study. Our use of abundance data in
a subsequent section of this discussion is on a pooled-sample basis rather than
on a station-to-station basis.
The recurrent group method (Fager, 1957; Fager & McGowan, 1963; Fager &
Longhurst, 1968; Hayes, 1978) is based on presence and absence data. It has the
advantage of being definable and repeatable. The groups formed contain the
largest possible number of species, all members of which share a minimum
(pre-established and arbitrary) level of "affinity" (GMI value in the case of this
paper). The basic datum for the recurrent group method is the presence or
absence of a particular species at a particular station. Thus the method has the
disadvantage of being severely influenced by the method(s) of sampling em-
ployed, the distribution of sampling effort (among collecting sites, depth zones,
habitat categories, etc.), the dichotomous index of affinity chosen (here the
GMI), and the pre-established minimum index value chosen as the breakpoint
for testing affinity. Note that disadvantages relating to sampling technique and
distribution of sampling effort apply to any grouping strategy chosen. Although
Fager (1957) discusses analytical techniques for studying co-occurrence based on
numbers of individuals taken, these techniques, in recurrent group analysis, are
based on the results of the analysis and not the reverse.
Recurrent group analyses were performed for data based on Glovers Reef
samples only, Belize samples only, and Belize and Honduras samples. Only
Class I stations were included in the analysis. We report on results for runs at
three pre-established affinity levels (GMI = 0.500, GMI = 0.400, GMI = 0.300)
but discuss in detail only those groups formed at GMI = 0.300. Recurrent groups
formed at GMI = 0.300 are indicated in Figures 7, 8, and 9. Recurrent groups
formed at higher pre-established affinity levels are listed in Table 28.
Recurrent groups based on Belize and Honduras data. — At GMI = 0.300 it is possi-
ble to form four recurrent groups (fig. 7) with a total of 23 species (plus two
78
FIELDIANA: ZOOLOGY
GROUP
3, 4, 7
9, 15, 35
37, 51, 55
GROUP II
26 34
43 50
a/
b
\e
\ d
e\
A
GROUP III
22, 39, HI
GROUP V
6, 8
21, 27, 57
Fig. 7. The composition of recurrent groups of blennioid species from Belize and Hon-
duras, based on all Class I stations (N = 104). Small satellite circles indicate associates (see
text for additional explanation). Lines linking the groups (large circles) show intergroup
connections with level of connection (keyed to letters as listed below) indicated by ratio of
observed species-pair connections (at or above GMI — 0.300) to the maximum possible
number of such connections: a = 0/36, b = 20/45, c = 7/27, d = 0/12, e = 0/20, and f = 0/15.
associated5 species) based on the 104 Class I stations at which blennioid fishes
were taken in Belize and Honduras.
Group I
Members: (3) Labrisomus gobio, (4) L. guppyi, (7) L. nigricinctus; (9) Malacoctenus
aurolineatus, (15) M. triangulatus; (35) Ennecmectes boehlkei, (37) E. pectoralis; (51)
Entomacrodus nigricans; (55) Ophioblennius atlanticus macclurei.
Associated species: (28) Starksia sluiteri.
All nine members of Group I were taken at three stations [CBC (2), GR (1)], all
in shallow water (less than 15 ft). Only one habitat category was represented
(H2). Six, seven, or eight members of Group I were taken at 12 additional
stations [BRG (2), BRT (1), CBC (1), LR (2), CC (5), IR (1)], all in shallow water
(less than 18 ft, 11 of 12 stations in less than 10 ft). Four habitat categories were
5Species associated with a recurrent group exhibit affinity (established by GMI value)
exclusively with some (not all) members of that group.
GREENFIELD & JOHNSON: BLENNIOID FISHES
79
Fig. 8. The composition of recurrent groups of blennioid fishes based on Belize data only
(number of Class I stations = 83). Other symbols as in Figure 7. Key to intergroup connec-
tions: a = 0/36, b = 7/27, c = 9/27, d = 0/12, e = 0/12, and f = 0/9.
represented [H2 (7), H3 (1), H4 (3), H9 (1)]. The station (F 75-51) classed H9 was
at the Hog Islands in an area with substrates corresponding to H2 and H4 habitat
categories well-represented.
At the affinity level GMI = 0.500, most of the members of Group I formed one
of two groups labeled (table 28) Group I-A and Group I-B. Three or more
members of Group I-A occurred at 28 stations, all in shallow water (less than 20
ft, 23 of 28 in less than 10 ft), with most records (22 of 28) from stations classed
H2 (15) or H3 (7). Two or more members of Group I-B occurred at 14 stations, all
in shallow water (less than 18 ft, 12 of 14 in less than 10 ft), with most records (9
of 14) from stations classed H2. As a group, Group I-B was never taken in patch
reef (H3) habitats.
At the affinity level GMI = 0.400, all Group I species except (3) Labrisomus gobio
form a recurrent group (table 28).
Group II
Members: (26) Starksia nanodes, (34) Enneanectes atrorus, (43) Emblemaria cald-
welli, (50) Lucayablennius zingaro.
Associated species: (19) Paraclinus infrons.
80
FIELDIANA: ZOOLOGY
GROUP I
1. 15, 35
GROUP II
25, 26, M
43, 50
*/
b
\e
\ d
®\
A
GROUP III
5 22
33 39
GROUP IV
2, 13, 51
Fig. 9. The composition of recurrent groups of blennioid fishes based on Glovers Reef
data only (number of Class I stations = 42). Other symbols as in Figure 7. Key to inter-
group connections: a = 0/15, b = 0/9, c = 2/12, d = 3/20, e = 0/15, and f = 0/12.
All four members of this group were taken at four stations, all at Glovers Reef,
all at depths exceeding 50 ft, all in HI habitats. Three of the four members of
Group I were taken at eight additional stations [CBC (2), GR (5), IR (1)], all at
depths exceeding 45 ft, all in HI habitats.
None of the members of Group II formed part of a recurrent group at GMI =
0.500. Three of the four members of Group II form a recurrent group at GMI =
0.400 (table 28).
Group III
Members: (22) Starksia atlantica, (39) Acanthemblemaria greenfieldi, (41) A.
spinosa.
All three members of Group III were taken at five stations [CBC (3), GR (2)], all
in shallow water (less than 15 ft). Two habitat categories were represented [H3
(4), H9 (1)]. The station (G 78-6) classed H9 was at Carrie Bow Cay on and
behind (leeward to) the reef crest, with substrates corresponding to H2 and H3
habitat categories well represented. Two of the three members of Group III were
taken at 10 additional stations [CBC (5), GR (5)], all in shallow water (less than 15
ft). Two habitat categories were represented [H2 (3), H3 (7)].
No member of Group III was part of any group formed at GMI = 0.500 or GMI
= 0.400.
>>
<v
vO
1
"8
E
c
Q
9\
U3
8
o
ir.
—
II
Ov
in
II
o
o
u I
3 0i
O >
in
t
«
M
(0
II
>
o
u 0i
in
3
5
rn
•g
c
3
0.9
I-H
in
t/5
<
J".
s.
<
1
o.H <^
03 en
in in *; tj<
s
■<* <N <
sr
■8
^ i
oj x
i^
in
CO r->
■a
03
CD
in
CI
in
•g
o
in
C "0
H CT>
x
rH
o e
<n
U a
"<* IN
<
Ov
a,
3
O
Im
60
O
2
o
E
(J
^ >
O .2
81
82 FIELDIANA: ZOOLOGY
Group IV
The group of species (2, 13, 51) herein referred to as Group IV formed a
recurrent group only when Honduras data were excluded from the analysis.
Discussion of Group IV is deferred to results for runs based on Belize data only.
Group V
Members: (6) Labrisomus kalisherae, (8) L. nuchipinnis, (21) Paraclinus nigripinnis,
(27) Starksia occidentalis, (57) Scartella cristata.
All five members of Group V were taken at five stations [SC (1), CC (4)], all in
shallow water (less than 10 ft). Three habitat categories were represented [H4
(3), H5 (1), H9 (1)]. The station classed H9 was at the Hog Islands in an area with
substrates corresponding to H2 and H4 habitat categories well represented.
Three or four members of Group V were taken at three additional stations [BC
(1), TC (1), CC (1)], all in shallow water (less than 5 ft). Two habitat categories
were represented [H4 (1), H5 (2)].
No member of Group V was part of any recurrent group formed at GMI =
0.500. At GMI = 0.400 four recurrent groups are possible [(a) 6, 8, 21; (b) 6, 8, 57;
(c) 6, 21, 27; (d) 8, 13, 21] of which three (a, c, d) are equally acceptable in that in
each case all three species were taken at a total of seven stations. Of the six
species represented in these four groups, only one species, (13) Malacoctenus
gilli, is not recognized as a member of Group V.
Recurrent groups based on Belize data. — At GMI = 0.300, it is possible to form
four recurrent groups (fig. 8) with a total of 22 species (plus two associated
species) based on the 83 Class I stations at which blennioid fishes were taken in
Belize. Groups I, II, and III thus formed are identical in composition with groups
based on the total (Belize + Honduras) data except that species (51) Entomacrodus
nigricans is replaced by (28) Starksia sluiteri as a member of Group I (cf. figs. 7, 8).
Two equally acceptable groups [(a) 6, 8, 21; (b) 6, 21, 27] corresponding to Group
V (fig. 7) can be formed from the Belize data. In each case (a and b) all three
species were taken at three stations, the same three stations: G 74-11, BC, 0 to 4
ft, H5; G 74-15, TC, 0 to 3 ft, H5; G 74-16, SC, 0 to 5 ft, H5.
Group IV
Members: (2) Labrisomus bucciferus, (13) Malacoctenus gilli, (51) Entomacrodus
nigricans.
All three members of Group IV were taken at six stations [Belize only: BRT (1),
CBC (1), GR (3), LR (1)], all in less than 4 ft of water. Two habitat categories were
represented [H2 (3), H6 (3)]. Two of the three members of Group IV were taken
at four additional stations [Belize only: CBC (1), GR (3)], all in less than 4 ft of
water. Four habitat categories were represented [H2 (1), H3 (1), H6 (1), H9 (1)].
The station (G 74-2) classed H9 was at Glovers Reef in less than 2 ft of water.
This station should probably be classed H6, but complete habitat information
was not recorded by the collectors.
Group IV was formed only in those runs in which Honduras data were
excluded (cf., figs. 7 vs. 8, 9). This group may represent an artifact, a failing of
the methodology, in that species (51) Entomacrodus nigricans would join Group I
(as based on Belize data only) were it not for the low affinity level (GMI = 0.299,
just below the required 0.300) shared with just one Group I species: (15)
Malacoctenus triangulatus. The GMI value shared by this pair is 0.472 for Belize +
GREENFIELD & JOHNSON: BLENNIOID FISHES 83
Honduras data, and 0.000 for Glovers Reef data (the two were taken together at
only one Glovers Reef station). Because of the apparent high association (relative
to sampling effort) of this group with very shallow reef-crest intertidal or im-
mediately subndal pools (H6), we choose to recognize Group IV as a separate
group for purposes of discussion in this paper.
Recurrent groups based on Glovers Reef data. — At GMI = 0.300, it is possible to
form four recurrent groups (fig. 9) with a total of 15 species (plus one associated
species) based on the 42 Class I stations at which blennioid fishes were taken at
Glovers Reef.
Based on Glovers Reef data only three species corresponding to Group I
formed a recurrent group, viz., (4) Labrisomus guppyi, (15) Malacoctenus trian-
gulatus, (35) Enneanectes boehlkei. Two or three members of this group were taken
at five Glovers Reef stations, all in less than 15 ft. Two habitat categories are
represented [H2 (2), H3 (3)].
In the run restricted to Glovers Reef data five species corresponding to Group
II formed a recurrent group, viz., (25) Starksia lepicoelia, (26) S. nanodes, (34)
Enneanectes atrorus, (43) Emblemaria caldwelli, (50) Lucayablennius zingaro, with (19)
Paraclinus infrons as an associated species. Three or more members of this group
were taken at a total of 11 Glovers Reef stations, all at depths exceeding 50 ft.
Only one habitat category (HI) was represented.
Four species corresponding to Group III formed a recurrent group based on
Glovers Reef data, viz., (5) Labrisomus haitiensis, (22) Starksia atlantica, (33) En-
neanectes altivelis, (39) Acanthemblemaria greenfieldi. Three or four members of this
group were taken at a total of seven Glovers Reef stations, all at depths less than
20 ft. Only one habitat category (H3) was represented.
Group IV as formed from Glovers Reef data (fig. 9) is identical to Group IV
based on all Belize data (fig. 8). All three members of Group IV were taken at
three Glovers Reef stations, all in less than 4 ft of water, from stations repre-
senting only one habitat category (H6). Two of the three members of Group IV
were taken at three additional Glovers Reef stations, all in less than 4 ft of water.
Three habitat categories were represented [H3 (1), H6 (1), H9 (1)]. The station (G
74-2) classed H9, in less than 2 ft of water, should probably be classed H6, but
complete habitat information was not recorded by the collectors.
Only two of the five species assigned to Group V (fig. 7) were taken at Glovers
Reef — species 8 and 21 — and neither species is part of any recurrent group
formed on the basis of Glovers Reef data alone.
Summary of composition of recurrent groups. — Based on the three sets of analyses
(Belize + Honduras, Belize only, Glovers Reef only) we choose to recognize, for
purposes of further discussion, five groups of blennioid fishes from Belize and
Honduras:
Group I —8 species (3, 4, 7, 9, 15, 35, 37, 55)
Group II —4 species (26, 34, 43, 50)
Group III— 3 species (22, 39, 41)
Group IV— 3 species (2, 13, 51)
Group V —5 species (6, 8, 21, 27, 57)
Groups I, II, III, and V are identical to the foregoing definition (fig. 7) except, for
reasons discussed elsewhere, species (51) Entomacrodus nigricans is excluded
84 FIELDIANA: ZOOLOGY
from Group I and recognized as a member of Group IV. Group IV is defined in
the two preceding sections (Belize data only, Glovers Reef data only). We realize
that we are departing from normal procedure in recognition of five groups based
on three separate sets of analysis, but we believe that distributional information
and analysis detailed below justifies this departure. All subsequent discussion is
based on the five groups as defined in this paragraph.
Analyses based on pooled data. — The results of the recurrent group analysis are
summarized in Tables 29, 30, and 31. Table 29 gives the number of stations at
which 60% or more of the members of the indicated groups were taken with
respect to the nine habitat categories and the six geographic localities defined in
Table 26. None of the groups occurred (as a group) at any mainland station, and
only Group V was found to occur as a group at cays inside the Barrier Reef. Each
group is distinctive in the habitat category or categories at which it was best
represented.
Capture data (based on Class I stations only) for the 23 species included in the
five recognized groups is summarized in Tables 30 and 31. In all cases the
number of records and proportion of individuals given is based on all 104 Class I
stations, despite the fact that Group IV was formed from Belize data only.
Justification for this is provided by the concordance calculations discussed
below.
One result that stands out, especially in the cases of Groups I and V (table 30),
is the remarkable success of our sampling efforts at the Hog Islands. Relative to
effort expended — Hog Island stations account for 11 of 104 Class I stations — the
number of individuals of Group I and Group V species taken at the Hog Islands
is very large. The same is true for one or two species included in Group IV.
Equally large were the total number of species (species richness) represented at
several Hog Island stations. Of the 127 collections on which this paper is based,
only 20 collections contained 10 or more blennioid species, and eight of these
were at Hog Island stations. A total of 23 blennioid species was taken at one Hog
Island station (F 75-51), and totals of 19 species each were taken at two additional
Hog Island stations. There three stations are the richest in terms of numbers of
species taken of any of the 127 stations in Belize and Honduras at which blen-
nioid species were taken.
There are a number of factors contributing to the richness of the Hog Island
stations: (1) all were obtained by a field party of up to six persons, all with
extensive experience in collecting (many of the Belize collections were obtained
by a field party of only two persons), (2) all of the richest collections were in
shallow water (less than 20 ft), allowing time to obtain very large samples, and
(3) the geography of collecting sites at the Hog Islands was such that a number of
Hog Island stations were in areas exhibiting a richer variety of habitats than at
any site visited by us in Belize; e.g., at two sites in the Hog Islands (F 75-47, F
75-51), collecting ranged from rocky ledges, cobble, and rubble inshore to small
but richly developed coral formations offshore. This mixture of habitat types was
most pronounced at the two stations listed but was more or less true for a
number of other shallow-water Hog Island stations. Both the Hog Islands and
Roatan are "high" rocky islands and are without parallel in Belize. All of this
would seem to contribute to a possible explanation for the fact that a number of
species taken together at Hog Island stations were never taken together in Be-
'£ '£
a. a.
re re
60 60
O O
& a*
C c -^
re i; x
(fl re **
o> *- c
•C a; •=
00 ^ ni
Si * C
re a.
.2 }Tr <n
5 -^ ro
SCO,
052
~ 3*
01 ^ c
q, re ai
0) o c
•" en (X
* 6 5b
» £ c
§j= i
00 0
IT'S *
Si
§!§
E O "2
C X <«
0 y —
re > t«
"ft «u
0 ■ c
v- C O
3 « s
C 0)°
•2 £S
0.° re
re 73 H
(J 01 c
<f> .3
<3< _Q I/)
fN ~ re
Uj 1/5 (fl
3 01 01
S ;.o ;xj
Hre re
o 5
4) >— •
^ w in (N m o 00
3-2
Z re
CO
3 0) *J
O [i, C) Ov N N
^pg vo n 00 n
00
o
* rH
s
5*
60
G
re jn
"I
3
re SB
E
CD
E
I * I
I I
3
0
B 1
CO
a 1 1 1 g
a.
3
8 ~
O
"8
« S £ > P
85
c |~I cic I iG-Ci-c-
>
C (0
CQ
" (« .,
S '3 «
S 3 0i
v re -C
o
£ ^""
U § 2
. s- 01
» U *5
<8 3
S * <u
re
tfl
3~ S
C ffl'o
SO Du5
3 C a
.a o c
re
to
K .H
«5 3
eJ
Ml
c P5
r-1 *— ^ /«
S
0)
gi
.S 3
a 5 '5
O XT)
■M <U C
■p »:
c ~ 2 •
.spa s1?
C O 2
re
£•8
as |
«3
111!
Sfi en « w
I *^ ^
U 3 1! c
O "5 s
■ C Oi o
S n £? '3
-J ** „
H §3 . (§
It, <N (J
Mi 01
.5 es
* in
re C
S I/!
re oi
31
fT fn" ^" a^* fvT \D oo
(NtNinmroi-iro o I
| rT r^tinf^" fx 3" in So" lo
m o <n in rifl't
5"" o? nT t$ <3" oo tn ^o nTin r^~3" ct^\o en f^f-Tf^- | (vTrn |
in nT 5^ In" So cn" oo £? S1 | co^S" cNTo^in ^In in
m
1 I^ ON
\o i-h m
0> \ON
1 9>
(N
in
1 CN rn
mN^
(N O O
1 m
1 O
en | cn"^
I I
I I
s
^nfOTfrn
oo i—i oo i— i m i-
I I I IB I I I I
II IS I IB
I I I I III III
<r> »orHtnnnNts t^oot^oo ono (N^ro *ririnn
Z^N^KinxooN mn^oN t^votN o in in ^ n m in to
tNi-lfNl-lfn i-lfOr-lrH
8 „ s a £ >
"C an^soMninMn a^^no ^ cn on r-i hNnn fevoooHNN
«K nncnm KNrttm gMto^ 5 hid 5 Ntsm
£ S 2 S S S
OT U U U U <J
86
8
g ■
c
o
a
2
c
35
■a
(0
1
o
</i
D.
3
o m
01
>
i
«:
2
1
&
IS
gj
-
I
2
a.
6
N
I I
13
IS
-
PP
rsl i— n
r^ <N CM r- r^
5
o o
r^ i- r-
O O n
r-i n in sn
O O (N i-i O
33 I
S3 I
m ^ oo so
en OrH Tf
i— i <N i-h n Ci <N fN
OvOtTf
13
18
•^ m m tj< i-c
^ ^ rt fi ^
r*"^ |
IS
IS
n n n Tt n n n
nT^^" | fvTo^f^in S?
vC \D (N I (N vO n m a~>
O »-< O I r((NrlO o
So 31 S~ f^ ^ ^ ^ in" ^
^oooNNtr^^ nj
I I I I S I I I s*^ 3 11 III lll£
I I
(N <N (N in
<« oSP^pfSQf^t* t^ootxoo ono n*(«j vcniHnn
CMi-ifMrirO rHCDrHrH
i
<
B ~ a 5 £ >
Sam^NOMfiinMn 5j vO ^ fT> Q £j <N Os >-« ^(Njfrjr-c fc vOMrits N
g r<nnin grjrtf J gNnTf g r-i in g NNin
U
a:
87
88 FIELDIANA: ZOOLOGY
lize. For example, Labrisomus kalisherae was taken with 14 other blennioid species
only at Hog Island stations (2, 4, 5, 7, 9, 12, 33, 35, 41, 46, 47, 49, 51, 53— this list
does not include species that were only taken at Hog Island stations). A number
of similar examples could be found.
Our purpose in conducting recurrent group analysis was to find, for the
blennioid species included in this paper, groups of species commonly taken
together, in Fager's (1957) words, "... species which very frequently form a part
of each other's environment." Our unit of sampling is the station or
collection — a broadcast method that may well result in the grouping of species
that occupy strikingly distinctive microhabitats. One possible example of this is
indicated in the discussion of (51) Entomacrodus nigricans. Another seemingly
good example is the case of (39) Acanthemblemaria greenfieldi and (41) A. spinosa,
which with (22) Starksia atlantica form Group III. Adults of the two species of
Acanthemblemaria occupy strikingly different (yet closely adjacent) microhabitats
(Greenfield & Greenfield, in press) yet are commonly taken together in the same
sample. Thus, the scale of sampling is relatively coarse, and the results must be
so interpreted.
With due knowledge of the coarseness of the groupings obtained through the
recurrent group analysis, we were interested in determining the level of concor-
dance among species belonging to a recurrent group with respect to captures
tabulated by geographic localities and habitat categories. We follow Fager (1957)
and Fager & McGowan (1963) in using numbers of individuals taken in deter-
mining this concordance; however, the uneven distribution of sampling effort
(tables 3, 4) as well as biases inherent in our sampling methods suggested that
ranking within species on a station by station basis, as done by Fager (1957, p.
593), would not produce interpretable results. Rather we tested for concordance
(Kendall's coefficient of concordance, W, corrected for ties, see Tate & Clelland,
1957) among the species of a recurrent group based on the pooled data presented
in Tables 30 and 31. For each species within each group we ranked captures
(total number of individuals taken at Class I stations) over the six geographic
localities (table 30) and nine habitat categories (table 31) and then calculated the
level of concordance among the species forming the group. We are well aware of
the possible, in fact probable, effects that differences in sampling effort among
the six geographic localities and nine habitat categories might have had on our
results. Our sole purpose in making these calculations was to determine
whether or nor there existed a statistical basis for a discussion of apparent
differences in distribution among the five groups with respect to the geographic
localities and habitat categories. The results (table 32) show very high levels of
concordance in every case. We interpret this to mean that within the recurrent
groups as defined in this paper rank-abundance (based on pooled data) within
species at a given geographic locality or for a given habitat category tends to
remain the same from species to species. This result would be expected for
species supposedly forming a recurrent group. The very high concordance levels
(table 32) between the members of recurrent groups allows further pooling of the
data by geographic locality (table 33) and habitat category (table 34). The results
are summarized below for each group.
Group I. — The species forming Group I are primarily associated with offshore,
shallow-water, coral-rich habitats. Fifty-eight percent of all Group I specimens
GREENFIELD & JOHNSON: BLENNIOID FISHES 89
Table 32. Calculated concordance values, W, for species belonging to five recur-
rent groups. Based on data given in Tables 30 and 31. See text for additional expla-
nation.
Basis for concordance
Group
calculation
M
N
W
Significance
I
Geographic areas
Habitat categories
8
8
6
9
.80
.74
**
**
II
Geographic areas
Habitat categories
4
4
6
9
.78
.81
**
**
III
Geographic areas
Habitat categories
3
3
6
9
.81
.79
**
**
IV
Geographic areas
Habitat categories
3
3
6
9
.73
.74
*
**
V
Geographic areas
Habitat categories
5
5
6
9
.58
.64
**
**
Key:
M
N
»
*- it-
= number of species.
= number of geographic areas
= statistically significant at .01
= statistically significant at p <
(6) or habitat
< p < .05.
.01.
categories (9).
were taken at stations classed H2 (table 34), and 10 of 15 records for the group
(table 29) are from stations classed H2. In Belize 98% of all Group I specimens
were collected at offshore sites, either at the Barrier Reef or the atolls (table 33).
The rather high intergroup connection between Group I and Group V (fig. 7) is
almost entirely the result of broad collections in mixed habitats at the Hog
Islands.
Group 11. — The species forming Group II are associated with offshore (insu-
lar), deep-water, coral-rich habitats. Virtually all (> 90%) Group II specimens
are from stations in depths exceeding 45 ft. Ninety-five percent of all Group II
specimens were taken at stations classed HI (table 34), and 12 of 12 records for
the group (table 29) are from stations classed HI. In Belize 100% of all Group II
specimens were taken at offshore sites, either at the Barrier Reef or the atolls
(table 33).
Group III. — The species forming Group III are primarily associated with
offshore, shallow-water, coral-rich habitats. Ninety-one percent (table 34) of all
Group III specimens were taken at stations classed H2 (42%) or H3 (49%). Most
records (11 of 15) for Group III as a group are from stations classed H3 (table 29).
In Belize 100% of all Group III specimens were taken at offshore sites, either at
the Barrier Reef or the atolls (table 33).
Group IV. — Group IV could be formed only in those runs in which Honduras
data was excluded, due in large part, we believe, to the diverse range of habitats
represented at a number of Hog Island stations. In Belize 184 of 267 (69%) Group
IV specimens were taken at stations classed H6. Stations classed H6 accounted
for only 6% of the 83 Class I stations in Belize. As a group (Belize data only)
Group IV was equally represented at stations classed H2 and H6 (table 29), but
all three members of Group IV were taken together only at stations classed H6.
In Belize 97% of all Group IV specimens were taken in less than 4 ft of water. In
Belize 97% of all Group IV specimens were taken at offshore sites, either at the
60 »>
o -a
SC §
<2
5 NN
33
fcs s
TO 01
CO g
w
a
4j
• - 01
<2 <u
EJj
.2 «
la
3 «
I I
T3
Si
s I I
tn £ co Tt in os i%
2 hJ 00 N H tN <N
O
CO (XI
O <
a, >-i
a w
5 «
!5£>
90
GREENFIELD & JOHNSON: BLENNIOID FISHES 91
Table 34. Pooled capture data for five recurrent groups of blennioid fishes tabulated
with respect to habitat category. Based on data given in Table 31. Values for each group for
each locality given as proportion of total specimens (Ns). See text for additional explana-
tion.
Group
Ns
HI
H2
H3
H4
H5
H6
H7
H8
H9
I
1,432
.002
.58
.11
.21
.01
.003
.002
.001
.09
II
240
.95
.03
.02
III
153
.01
.42
.49
.05
—
—
—
—
.03
IV
599
—
.13
.01
.33
.02
.34
.10
—
.08
V
474
.01
.05
—
.24
.52
.07
.01
.01
.09
Effort
(Class I
stations)
104
.29
.18
.18
.04
.06
.06
.05
.07
.08
Barrier Reef or the atolls (table 33). In Belize, therefore, this group appears to be
primarily associated with very shallow, reef-crest, intertidal pools or adjacent
subtidal areas.
Group V. — The species forming Group V are apparently associated with more
inshore (but not mainland), shallow-water habitats and are not strongly associ-
ated with richly developed coral reefs. Eighty-six percent (table 34) of all Group
V specimens were taken at stations classed H4 (24%) or H5 (52%). Stations
classed H4 or H5 account respectively for only 4% and 6% of the total 104 Class I
stations. Only 6% of all Group V specimens came from stations classed HI, H2,
or H3, despite the fact that stations classed as one of these three account for 65%
of the 104 Class I stations. No Group V specimens were taken at stations classed
H3. Most records (7 of 8, see table 29) came from stations classed H4 (four
stations at the Hog Islands) or H5 (three stations, all at cays behind the Barrier
Reef)- In Belize the overwhelming majority of Group V specimens (76%) was
taken at cays inside of the Barrier Reef (table 33).
ZOOGEOGRAPHIC IMPLICATIONS
The shorefishes of the tropical western Atlantic are perhaps better known than
those of any other comparably sized tropical marine area. Despite this knowl-
edge, there exists little agreement on major patterns of shorefish distribution in
this area as evidenced by a number of competing and noncomplimentary
schemes for major faunal assemblages, provinces, or generalized tracks (e.g.,
Briggs, 1974; Miller, 1968; Robins, 1971b; Gilbert, 1973; Rosen, 1975). Three
factors, at least, contribute to this situation: (1) lack of agreement among authors
on the purposes of and best methodology for zoogeographic analysis, (2) a
continuing need for synoptic systematic studies on numerous groups, and (3) a
lack of adequate study material from a sizable proportion of the tropical western
Atlantic, particularly the Caribbean coasts of Central America and northern
South America (Gilbert & Kelso, 1971; Robins, 1971b, 1972). The purpose of
these paragraphs is to strongly argue for the need for additional synoptic col-
lecting in the western Caribbean by detailing the impact of our information on
blennioid fishes from Belize and Honduras on one particular scheme for faunal
provinces within the tropical western Atlantic area — that of Briggs, 1974.
92 FIELDIANA: ZOOLOGY
Of the 57 blennioid species represented in our collections from Belize and
Honduras, 20 species (table 1) represent new records for Belize, 40 species new
records for Honduras, and 15 species are recorded for the first time from the
Caribbean coast of Central America. Included in the material were representa-
tives of four blennioid species new to science: (27) Starksia occidentalis Greenfield,
1978, (39) Acanthemblemaria greenfieldi Smith- Vaniz & Palacio, 1974, (44) Emble-
maria hyltoni Johnson & Greenfield, 1976, and (47) Emblemariopsis pricei Green-
field, 1975.
Briggs (1974) divided the tropical western Atlantic into three zoogeographic
provinces: Caribbean, Brazilian, and West Indian (Briggs, 1974, p. 63). Possibly
the most striking feature of Briggs' proposal is that his West Indian Province
[Bermuda, Bahamas, Greater and Lesser Antilles (to Granada)] is entirely insu-
lar. Briggs' disjunct Caribbean Province consists of southern Florida (east and
west coasts), the Caribbean coasts of Central and northern South America, and
the Mexican coast bordering the Gulf of Campeche. To Briggs (1974, p. 74),
"... In zoogeography, the ultimate proof, as far as determination of provinces is
concerned, is in the extent of endemism." We do not in this paper attempt to
argue the merits of this dictum, but we are prepared to show that even in the
supposedly well-known tropical western Atlantic arguments based on supposed
levels of endemism may need to be severely modified.
Briggs (1974) defends his recognition of an entirely insular West Indian
zoogeographic province (as distinct from his Caribbean Province) partly on the
basis of evidence provided by Bohlke & Chaplin (1968) in their monumental
contribution to our knowledge of the shorefishes of the Bahamas. Briggs states
that of 466 shorefish species treated by Bohlke & Chaplin, 87 (about 19%) were
unknown outside of the area of his West Indian Province.
Bohlke & Chaplin (1968) record 54 species of blennioid fishes from the
Bahamas. According to information provided by them, one would conclude that
18 (33%) of these species were limited to Briggs' West Indian Province. Rep-
resented in our collections from Belize and Honduras are 136 of those 18 species,
and a 14th (Emblemariopsis bahamensis) has also been reported from the western
Caribbean (Birdsong & Emery, 1968). This leaves only four species of the 54
treated by Bohlke & Chaplin (1968) as putative West Indian endemics. Of these,
one occurs at Isla de Providencia off Nicaragua (G. H. Burgess, pers. comm.),
leaving only three species as putative West Indian endemics. Briggs (1974), to
our knowledge, never states how high the percentage of endemism has to be in
order to be ". . . the ultimate proof, as far as determination of provinces is
concerned . . .," but we take it (Briggs 1974, e.g., pp. 92, 95) that he would not
consider 5.6% endemism (37 of 54) as sufficient supporting evidence to argue for
a major zoogeographic province.
Three additional blennioid species8 described subsequent to publication of
Bohlke & Chaplin's (1968) book and known (at the time of original description)
to occur only in the West Indian area, have now also been taken off Central
America. Although we have identified only a small portion of the gobies from
our collections in Belize and Honduras, preliminary results, summarized by
Greenfield, 1978, suggest a developing picture similar to that for the blennioid
6The 13 species: 12, 16, 19, 22, 25, 26, 30, 33, 34, 36, 40, 42, 50.
'Now 2 of 54; see addendum (Paraclinus naeorhegmis) .
8Viz., Emblemaria caldwelli, Emblemariopsis leptocirris, Starksia elongata.
GREENFIELD & JOHNSON: BLENNIOID FISHES 93
fishes. Yet it is groups such as the blennioids and gobiids that we might expect to
best show a trend toward localized distributions {e.g., Rosenblatt, 1963).
It seems likely that as additional material from new collections along the
Caribbean coast of Central and northern South America becomes available and is
studied, the distinctness of Briggs' West Indian Province, at least as based on
Briggs' species-tally methodology, will further decline. This is not to say that
regional endemism within the tropical western Atlantic area is lacking; there is
good evidence to the contrary (see Greenfield, 1978). We do believe that until
adequate synoptic collections are available from the western Caribbean area,
attempts to use levels of endemism to define fish faunal provinces in the tropical
western Atlantic region will prove to be exercises in futility.
In discussing the many curious disjunctions in distribution of species within
his insular province, Robins (1971b) suggested that a combination of habitat
specificity (the degree to which a particular species is associated with and re-
stricted to a given set of habitat-defining parameters) and rather broadscale
habitat patchiness might prove to be key clues to understanding such distri-
butions. The discovery of typical coral-reef-associated fishes far north in the
Gulf of Mexico at deep reefs is one example of this dependence on habitat
availability rather than geography within the western Atlantic area {e.g.,
Cashman, 1973; Smith et al., 1975; Sonnier et al., 1976; Smith, 1976). Our own
work on blennioid fishes from Belize and Honduras suggests a spectrum in
degree of habitat specificity, from species rather broadly distributed over the
habitat categories recognized herein to species closely associated with a specific
habitat type. Given that the parameters used to define a habitat category will
greatly affect any view of degree of habitat specificity, it is also true that any test
of the general applicability of Robins' suggestion will require additional attempts
to determine environmental correlates of distribution for species in selected
shorefish groups. Especially important will be attempts to define the degree of
habitat specificity in a suitably broad and diverse area. Because of the diversity of
habitats represented within a relatively restricted inshore to offshore distance,
we believe Belize to be an ideal area in which to make such an attempt.
ADDENDUM
(October 1980)
We have taken and examined much additional blennioid material from Belize
since this report was accepted for publication. For most of the species this new
material confirmed distributional patterns described herein, and time did not
permit the incorporation of all new records. In the new material, however, were
specimens of six species previously unknown from Belize, including three not
previously represented in any of our western Caribbean material. Locality rec-
ords and, where appropriate, descriptive information are presented for these six
species in this addendum. New records are also presented for 12 species previ-
ously represented in our material by four or fewer lots. Measurements (where
given) are expressed as thousandths of the SL.
LABRISOMIDAE
(1) Labrisomus albigenys Beebe & Tee Van, 1928
This apparently rare species was previously reported from Belize by Birdsong
& Emery (1968) — a single collection from Lighthouse Reef (UMML 9555) that we
94 FIELDIANA: ZOOLOGY
have not had the opportunity to examine. One of us (RKJ) collected a single
specimen (FMNH 93837, 40.9 mm SL) at Carrie Bow Cay in March 1980. The
specimen was taken in the spur and groove province directly to the east of Carrie
Bow Cay in 24 to 26 ft of water in an area of rich coral and moderately high relief.
Data for the specimen (see table 7) follow: dorsal-fin elements, XVIII, 11: anal-fin
elements, II, 16; pectoral-fin rays, 13/13; lateral line scales, 42/42; gill rakers on
first arch (total), 15; head length, 367; interorbital width (bony), 39; eye diameter
(fleshy orbit), 98; snout length, 100; dorsal fin, length of first spine, 98; pelvic fin,
length of longest segmented ray, 269; upper jaw length, 125; ratio A (pelvic fin),
2.20; ratio B (nuchal cirri), 1.00.
(11) Malacoctenus delalandei (Valenciennes, 1836)
Additional material includes 86 specimens from six collections, all from sites
well inshore of the Barrier Reef, five of the six collections from inshore cays.
Coral rubble, extensive Thalassia, and a relatively firm bottom were common to
all six sites.
Additional material: Toledo District, 78146 (13), Tarpon Cay, 0 to 5 ft; 78147
(6), West Snake Cay, 0 to 4 ft; 78148 (21), "Kulu" Cay, ca. 2 miles WNW of Moho
Cay, mangrove-lined channel, extensive Thalassia offshore, coral rubble and
occasional isolated clumps of Porites inshore, 0 to 4 ft; 78149 (27), Wilson's Cay, 0
to 4 ft; Belize District, 90279 (18), Spanish Cay, ca. 1.5 miles ESE of Robinson
Point, coral rubble and Thalassia, adjacent to a pier, 0 to 3 ft; 90280 (1), mainland,
point ca. 1 mile west of Little Rocky Point, in and around breakwater composed
of large cobbles and surrounded by an extensive bed of Thalassia, 0 to 3 ft.
(16) Paraclinus barbatus Springer, 1955
Additional material includes 15 specimens from 10 collections, all from areas
of rich coral development and high vertical relief, all from the deep spur and
groove or dropoff zones. The specimens taken off Ambergris Cay represent the
first record of this species from the Barrier Reef.
Additional material: Glovers Reef, 87839 (3), Northeast Cay, spur and groove,
70 to 90 ft; 87840 (1), Southwest Cay, spur and groove, 45 to 60 ft; 87841 (1), Long
Cay, dropoff, 70 to 90 ft; 87842 (2), Long Cay, dropoff, 60 to 80 ft; 87843 (1),
dropoff at western (leeward) reef, due west of Long Cay, 60 to 80 ft; 90299 (1),
Long Cay, dropoff, 70 to 90 ft; 90300 (1), Middle Cay, spur and groove, 40 to 50
ft; 90301 (1), Northeast Cay, spur and groove, 80 to 100 ft; 94018 (2), Southwest
Cay, dropoff, 70 to 90 ft. Corozal District, Ambergris Cay, 94019 (2), dropoff, 60
to 100 ft.
(17) Paraclinus cingulatus (Evermann & Marsh, 1900)
Additional material includes 10 specimens from three collections, constituting
the first records for this species from Belize.
Additional material: Glovers Reef, 90302 (7), reef crest on windward side at
Long Cay, limestone rock with rich algal turf grading to rich shallow reef (mostly
Acropora palmata), 0 to 4 ft. Belize District at the Barrier Reef, 90303 (1), Sergeant's
Cay, rich, shallow reef on windward side, 0 to 8 ft; 90304 (2), Goff's Cay, reef
crest on windward side, rock and coral rubble grading to rich, shallow reef, 0 to
3 ft.
GREENFIELD & JOHNSON: BLENNIOID FISHES 95
( ) Paraclinus naeorhegmis Bohlke, 1960
Two specimens (94020) from a single station at Ambergris Cay (1980) represent
the first record of this species for Belize and a significant range extension for the
species. Previously reported only from the Bahamas and from very few speci-
mens (Bohlke & Chaplin, 1968, p. 519), P. naeorhegmis has also been taken at Isla
de Providencia (G. H. Burgess & C. R. Gilbert, personal communication). The
specimens from Ambergris Cay were taken from the "spur and groove prov-
ince," a zone of eroded reef rock, sand channels, low relief, and few living
corals, ca. 150 yd offshore from the crest of the Barrier Reef, in 34 to 36 ft of
water. Data for the specimens (16.0, 16.1 mm SL, data given in that order)
follow. Counts: dorsal-fin spines, xxvi (both); anal-fin elements, II, 15; II, 16;
pectoral-fin rays, 13/13 (both); pelvic-fin elements, I, 2 (both); caudal fin, number
of segmented rays, 13 (both). Measurements: body depth at dorsal-fin origin,
250, 248; body depth at anal-fin origin, 250, 261; caudal peduncle depth, 113,
118; dorsal fin, length of first spine, 200, 168, length of third spine, 100, 118;
pectoral fin, length, 238, 255; pelvic fin, length, 206, 199; caudal fin, length, 263,
261; head length, 325, 342; eye diameter, 100, 112; upper jaw length, 163, 143;
snout length, 69, 87.
(23) Starksia elongata Gilbert, 1971
Additional material includes a single specimen: 90342 (1), Glovers Reef, spur
and groove province on the windward reef, rich coral development, 23 to 25 ft.
(24) Starksia hassi Klausewitz, 1958
Additional material includes a single specimen: 87855 (1), Glovers Reef,
dropoff at western (leeward) reef, due west of Long Cay, 60 to 80 ft.
(28) Starksia sluiteri (Metzlaar, 1919)
Additional material includes 58 specimens from eight collections, all in less
than 6 ft of water, all but one in areas of windward, rich, shallow reefs.
Additional material: Glovers Reef, 87876 (8), reef crest at Long Cay, 3 to 6 ft;
87877 (3), reef crest south of Long Cay, 2 to 6 ft; 90369 (21), reef crest at Long
Cay, 0 to 4 ft; 90370 (8), reef crest at Long Cay, 0 to 5 ft. Belize District, at the
Barrier Reef, 90371 (1), Sergeant's Cay, cobble, broken chunks of concrete,
well-developed algal turf, 0 to 4 ft; 90372 (2), Goff's Cay, windward side, reef
crest, rock and coral grading to rich, shallow reef, 0 to 3 ft. Corozal District,
Ambergris Cay, barrier reef, 94021 (5), reef crest, 0 to 4 ft; 94022 (10), reef crest, 0
to 4 ft.
(29) Starksia starcki Gilbert, 1971
Additional material includes eight specimens from six collections constituting
the first records for Belize. All specimens were taken in areas of rich coral
development, high vertical relief, and in the spur and groove or dropoff prov-
inces at two Barrier Reef sites (Carrie Bow Cay, Ambergris Cay) and at Glovers
Reef. The material includes two adults (23.0 to 23.1 mm SL), both females, and
six juveniles (10.0 to 17.3 mm SL). The color pattern (fig. 10) is striking, espe-
cially in the small juveniles — a series of 10, narrow, chocolate bars strongly
contrasting with the pale cream background. The first (anteriormost) bar is con-
96
GREENFIELD & JOHNSON: BLENNIOID FISHES 97
tinuous across the nape and extends ventrally to a point on the gill covers
directly anterior to the insertion of the two to three ventralmost pectoral-fin rays.
The next three bars are continuous from the dorsal to the ventral contours of the
body. The next five bars are broken midlaterally on the body, with the epaxial
portion anteriorly offset from the corresponding hypaxial portion. The final bar
consists of the "hypural" markings. In the two adult specimens the barred
pattern is much less evident, due in part to the more even spread of pigmenta-
tion over the body. The nine bars on the body so evident in the juveniles are best
marked in the adults by the presence of blotches of pigment on the dorsal and
ventral contour adjacent to the dorsal- and anal-fin ray bases and by a barred
pattern on the dorsal and anal fins. Additional concentrations of pigment on the
head and associated with the pectoral fin are as described in Gilbert (1971,
p. 201).
There are 11 pores in the infraorbital series (based on the two adult specimens
only), extending from a point just anterior to the orbit and on a horizontal line
through the center of the eye (3:00 position; descriptions of pore positions based
on lateral view of right side) to a point adjacent to the dorsoposterior margin of
the orbit on a horizontal line through the base of the orbital cirrus (11:00).
Contrary to Gilbert's (1971, p. 200) description, pores 4 and 5 (6:00) and 8 and 9
(9:00) are paired in the Belize material. One para type (UF 17279, 20.5 mm SL) is
identical to the Belize specimens in pore number and configuration. Two larger
para types (ANSP 109800, 27.2 mm SL; USNM 205200, 27.0 mm SL) have one or
more additional pores, with the configuration differing slightly between the
right and left side (in each case), but the pores at the 6:00 and 9:00 positions are
paired as in the Belize material.
Meristic characters (based on all eight specimens): dorsal-fin elements, XX, 8
(1); XX, 9 (7); anal-fin elements, II, 18 (all); pectoral-fin rays, 13 (all); pelvic-fin
elements, I, 2 (all). Morphometric characters (expressed as thousandths of the
SL, based on four specimens, 16.3, 17.3, 23.0, 23.1 mm SL, and listed in that
order): head length, 344, 341, 335, 333; snout length, 61, 64, 74, 74; upper jaw
length, 135, 127, 157, 134; horizontal eye diameter, 92, 110, 113, 108; greatest
depth of body, 202, 202, 222, 216; dorsal fin, length of first spine, 86, 87, 91, — ;
pectoral fin, length of longest ray, 301, 289, 291, 281; pelvic fin, length of longest
segmented ray, 245, 243, 222, 221; caudal fin, length of longest ray, — , 289, — ,
255.
Additional material examined: Glovers Reef, 94023 (1), western (leeward reef),
spur and groove, 30 to 35 ft. Corozal District, Ambergris Cay, 94024 (1), spur and
groove, 58 to 62 ft; 94025 (1), dropoff, 70 to 100 ft. Stann Creek District, at the
Barrier Reef, 93834 (2), Carrie Bow Cay, spur and groove, 24 to 26 ft; 93835 (2),
Southwater Cay, spur and groove, 35 to 45 ft; 93836 (1), Curlew Cay, spur and
groove, 23 to 26 ft.
(30) Stathmonotus gymnodermis Springer, 1955
Additional material includes five specimens from one collection: 90318 (5),
Glovers Reef, reef crest on windward side at Long Cay, limestone rock with rich
algal turf grading to rich, shallow reef (mostly Acropora palmata), 0 to 4 ft. Meristic
characters: dorsal-fin spines, XLI (1), XLII (4); anal-fin elements, II (all), 22 (1), 23
(3), 24 (1); pectoral-fin rays, 8 (all); caudal fin, segmented rays, 11 (all).
98 FIELDIANA: ZOOLOGY
(31) Stathmonotus hemphilli Bean, 1885
Additional material includes four specimens from three collections, constitut-
ing the first records of this species from Belize: 90319 (2), Glovers Reef, spur and
groove province on windward side, south of Long Cay, rich coral development
with high vertical relief, 22 to 25 ft. Corozal District, Ambergris Cay, 94026 (1),
cut in barrier reef, steep-sided channel, eroded reef rock, little living coral, 35 to
38 ft, 94027 (1), dropoff, 80 to 92 ft.
CHAENOPSIDAE
(40) Acanthemblemaria maria Bohlke, 1961
Additional material includes 11 specimens from a single station, Corozal Dis-
trict, Ambergris Cay, at the Barrier Reef, 93977, reef crest; rich, living coral
formations bordered by extensive areas of cobble and coral rubble, 0 to 4 ft.
( ) Chaenopsis ocellata Poey, 1865
This species is known from the Bahamas, Florida, and the Greater Antilles
(Bohlke & Chaplin, 1968). This constitutes the first record for Belize and a sig-
nificant range extension for the species. Our three collections (83929, 6 speci-
mens, 34.7-44.5 mm SL; 83930, 8 specimens, 34.5-64.1; 83931, 17 specimens,
32.9-62.8) were all collected from a Thalassia bed in 6 to 25 ft of water on the
leeward side of a patch reef in the lagoon at Glovers Reef, about 1.5 miles west of
Long Cay. The colony was living in holes in the fine sand among the Thalassia.
Individuals were usually observed with the anterior portion of their body ex-
tending from a hole, but occasionally individuals were observed swimming
through the Thalassia or hanging in the water among the Thalassia blades.
(42) Coralliozetus cardonae Evermann & Marsh, 1899
Additional material includes two specimens (14.4, 18.8 mm SL) taken at a
single station, Corozal District, Ambergris Cay, at the Barrier Reef, 94017, reef
crest, rich, living coral formations bordered by extensive areas of cobble and
coral rubble, 0 to 4 ft. Counts, identical in the two specimens, follow: spinous
dorsal, 18; soft dorsal, 11; total dorsal-fin elements, 29; anal-fin elements, II + 19;
pectoral-fin rays, 13.
(45) Emblemaria pandionis Evermann & Marsh, 1900
Additional material includes 14 specimens from five collections, none of them
areas of rich coral.
Additional material: Glovers Reef, windward side, lagoon edge of leeward
reef flat, 90510 (1), in lagoon off Middle Cay, extensive Thalassia, sand, and
isolated pieces of rubble, 5 ft; 93987 (1), in lagoon between Long Cay and Middle
Cay, extensive Thalassia, sand, and isolated pieces of rubble, 4 to 5 ft. Glovers
Reef, 93988 (1), channel between Northeast Cay and Long Cay, "blowout" (sand
and rubble-lined basin) in extensive Thalassia bed, 4 to 6 ft. Corozal District,
Ambergris Cay, 93989 (6), "spur and groove province," a zone of eroded reef
rock, sand channels, low relief, and few living corals, 34 to 36 ft; 93990 (5), cut in
Barrier Reef, steep-sided channel, eroded reef rock, little living coral, 35 to 38 ft.
GREENFIELD & JOHNSON: BLENNIOID FISHES 99
( ) Emblemaria cf. piratula Ginsburg & Reid, 1942
A small (12.7 mm SL, male) spedmen of Emblemaria taken by Dr. Brian
Kensley (Station K-22, 7 February 1978, in a rubble sample from 90 ft close to the
outer wall of the deep channel off Carrie Bow Cay) is herein tentatively iden-
tified as Emblemaria piratula. This species has heretofore been reported only from
the northeastern Gulf of Mexico (Stephens, 1963, p. 89; Shipp, 1975, p. 16,
Johnson & Greenfield, 1976, p. 27). This record from Belize, assuming the cor-
rectness of our identification, represents a significant range extension for the
species.
The Belize specimen exhibits lower dorsal- and anal-fin counts than reported
values for Florida specimens: dorsal-fin elements, XIX, 13, total = 32 (vs. XVIII
to XIX, 14 to 16, total = 33 to 34); anal-fin elements, II, 19, total = 21 (vs. II, 21,
total = 23). Values for the pectoral fin, 13 (vs. 12 to 13) and pelvic fin, I, 2 (the
third soft pelvic-fin ray is vestigial in £. piratula, only two segmented rays are
evident) agree with reported values for Florida specimens. As in £. piratula
(Johnson & Greenfield, 1976, p. 19), there are apparently three CP pores, and a
line connecting the two posteriormost ST pores lies entirely anterior to the
dorsal-fin origin. The orbital cirri are elongate, ca. 50% of the head length. Aside
from a few scattered melanophores on the head, anterior body, and anterior
dorsal fin, the Belize specimen lacks pigmentation. The small size of the speci-
men and lack of any additional Belize material precludes any final taxonomic
decision.
(49) Hemiemblemaria simulus Longley & Hildebrand, 1940
Additional material includes six specimens from five collections: Honduras,
90522 (1), Little Hog Island, rock ledges and cobble inshore grading to small but
well-developed coral formations offshore, 0 to 20 ft; Belize, Glovers Reef, spur
and groove province off Northeast Cay, 80 to 90 ft, 90523 (1), 94009 (1). Belize
District, at Barrier Reef, 90524 (1), Sergeant's Cay, rich shallow reef on wind-
ward side, 0 to 8 ft. Corozal District, Ambergris Cay, 94010 (2), cut in Barrier
Reef, steep-sided channel, eroded reef rock, little living coral, 35 to 38 ft.
BLENNTIDAE
(52) Hypleurochilus aequipinnis (Gunther, 1861)
Additional material includes 48 specimens from seven collections. Remarkable
among these new specimens was a series of 11 (37.0 to 53.6 mm SL) taken on the
Belize mainland at a point ca. 1 mile west of Little Rocky Point (Belize District).
This series included the largest-bodied specimens yet taken in Belize or Hon-
duras, including a number of mature males with one enormously prolonged
orbital cirrus on each side (fig. 11). Randall (1966, p. 58) had reported prolonga-
tion of one of the orbital cirri on each side in mature males of Hypleurochilus
fissicornis (Quoy & Gaimard, 1824) and H. springeri Randall, 1966. The condition
in H. fissicornis was said to be remarkable among species of Hypleurochilus in that
the length of the prolonged cirrus was contained about 1.6 times in the head
length. In two of the largest male H. aequipinnis from the mainland series, 50.0
and 53.6 mm SL, the orbital cirrus is contained 1.3 and 1.4 times, respectively, in
the head length. Counts and selected mensural characters for Belize and Hon-
duras material (N = 10, 22.9 to 53.6 mm SL) follow: dorsal-fin elements, XII (all),
100
GREENFIELD & JOHNSON: BLENNIOID FISHES 101
13 (3), 14 (7); anal-fin elements, II (all), 15 (5), 16 (5); pectoral-fin rays, 14 (all);
pelvic-fin elements, I, 4 (all); vertebrae, 31 (8), 32 (2); number of orbital cirri (left
side only), 5 to 16, all specimens exceeding 37 mm SL (8 of 10 counted) with 10 or
more; dorsal fin, length of last spine, 33% to 57% of length of first ray of soft
dorsal fin; pectoral fin length contained 3.7 to 4.7 times in SL.
All of the additional material was taken in very shallow water (less than 10 ft).
Of the seven additional localities, five were at mangrove-covered cays inside the
Barrier Reef, two were mainland sites.
Additional material: 87952 (1), Toledo District, Frenchman's Cay, coral rubble
and algal turf, 0 to 5 ft; 87953 (1), Toledo District, "Unnamed Cay," ca. 2.5 miles
WSW of Wilson's Cay, mangrove channel with an extensive offshore Thalassia
bed, 0 to 4 ft; 87954 (1), Belize District, cay at Robinson Point, mangrove channel,
0 to 10 ft; 87955, Toledo District, "Kulu" Cay, ca. 2 miles WNW of Moho Cay,
mangrove channel with extensive offshore Thalassia bed, 0 to 4 ft; 93833, Belize
District, mainland, point ca. 1 mile west of Little Rocky Point, in and around
breakwater composed of large cobbles and surrounded by an extensive bed of
Thalassia, 0 to 3 ft. Corozal District, Chetumal Bay, 93955 (1), Cayo Negro,
eroded reef rock inshore, Thalassia offshore, 0 to 3 ft; 93956 (31); mainland, Punta
Allegre, eroded and undercut reef rock along shore, sand and Thalassia offshore,
0 to 3 ft.
Comparative material examined. H. aequipinnis: USNM 88988 (1), Dominican
Republic; USNM 120341 (1), Puerto Rico; USNM 192242 (2), Ascension Bay,
Yucatan; USNM 192400 (2), Cozumel, Quintana Roo; USNM uncat. (4), Bahia
Limon, Panama. H. bermudensis: FMNH 48709 (1), Bermuda; FMNH 48817 (1),
Bermuda; FMNH 49401 (2), Bermuda; FMNH 62261 (2), Florida, USNM 116805
(8), Tortugas. H. cf. geminatus: FMNH 40295 (1), Texas; 50734 (10), Florida;
USNM 208444 (4), Texas; USNM 217324 (3), Louisiana. H. fissicornis: BMNH
1868.9.22.9 (14), Brazil; BMNH 1891.9.30.1 (1), Uruguay. H. springeri: USNM
219141 (1), Puerto Rico.
(54) Lupinoblennius dispar Herre, 1942
Additional material includes nine specimens from four collections: Belize Dis-
trict, 90470 (4), mainland, ridge of dredge spoil at entrance to Salt Creek, mud
and cobbles, 0 to 2 ft; 90471 (1), mainland, point ca. 1 mile west of Little Rocky
Point, in and around breakwater composed of large cobbles and surrounded by
an extensive bed of Thalassia, 0 to 3 ft; 90472 (1), Moho Cay, opposite St. John's
College, Belize City, mangrove shoreline, steep mud bank, 0 to 3 ft. Corozal
District, Chetumal Bay, 93958 (3), Punta Allegre, eroded and undercut reef rock
along shore, sand and extensive Thalassia offshore, 0 to 3 ft. At each of the above
stations (except the one at Moho Cay), L. dispar was taken with (18) Paraclinus
fasciatus and (52) Hypleurochilus aequipinnis. At the station near Little Rocky Point,
(8) Labrisomus nuchipinnis and (11) Malacoctenus delalandei were also taken.
(56) Parablennius marmoreus (Poey, 1875)
Additional material includes five specimens from two collections, both from
mangrove-covered cays inside the Barrier Reef: 87948 (1), Toledo District,
"Kulu" Cay, ca. 2 miles WNW of Moho Cay, mangrove channel with an exten-
sive offshore Thalassia bed, 0 to 4 ft; 97949 (4), Toledo District, "Unnamed Cay,"
ca. 2.5 miles WSW of Wilson's Cay, mangrove channel with an extensive
offshore Thalassia bed, 0 to 4 ft.
102 FIELDIANA: ZOOLOGY
LITERATURE CITED
Bath, H. 1977. Revision der Blenniini (Pisces: Blenniidae). Senckenbergiana Biol., 57 (4/6):
167-234.
Birdsong, R. S. and A. R. Emery. 1968. New records of fishes from the western Carib-
bean. Quart. J. Fla. Acad. Sri., 30 (3): 187-196.
Bohlke, J. E. 1957. A review of the blenny genus Chaenopsis and the description of a
related new genus from the Bahamas. Proc. Acad. Nat. Sri. Phila., 109, 81-103.
1960. Two new Bahaman species of the clinid fish genus Paraclinus. Not.
Nat., Acad. Nat. Sci. Phila., no. 337, 8 pp.
Bohlke, J. E. and C. C. G. Chaplin. 1968. Fishes of the Bahamas and adjacent tropical
waters. Acad. Nat. Sri. Phila., Livingston Publ. Co., Wynnewood, Pa., xxii + 771 pp.
Bohlke, J. E. and C. H. Robins. 1974. Description of a new genus and species of clinid fish
from the western Caribbean, with comments on the families of Blennioidea. Proc. Acad.
Nat. Sri. Phila., 126 (1): 1-8.
Bohlke, J. E. and V. G. Springer. 1961. A review of the Atlantic species of the clinid fish
genus Starksia. Proc. Acad. Nat. Sri. Phila., 113 (3): 29-60.
1975. A new genus and species of fish (Nemaclinus atelestos) from the west-
ern Atlantic (Perriformes: Clinidae). Proc. Acad. Nat. Sci. Phila., 127(7): 57-61.
Breder, C. M., Jr. 1925. Notes on fishes from three Panama localities: Gatun Spillway, Rio
Tapia and Caledonia Bay. Zoologica, 4 (4): 137-158.
1927. Scientific Results of the First Oceanographic Expedition of the "Paw-
nee" 1925. Fishes. Bull. Bingham Oceanogr. Coll., 1 (art. 1): 1-90.
Briggs, J. C. 1974. Marine Zoogeography. New York, McGraw-Hill, Inc., 475 pp.
Burgess, G. H. 1978. Zoogeography and depth analysis of the fishes of the Isla de Pro-
videncia and Grand Cayman Island. Unpubl. M. Sri. Thesis, University of Florida,
Gainesville (not seen).
Caldwell, D. K. 1963. Marine shore fishes from near Puerto Limon, Caribbean Costa
Rica. Contr. Sci. L. A. Co. Mus., 67: 1-11.
Caldwell, D. K. and M. C. Caldwell. 1964. Fishes from the southern Caribbean col-
lected by Velero III in 1939. Allan Hancock Atlantic Expedition Report, no. 10, pp. 1-61.
Caldwell, D. K., L. H. Ogren and L. Giovannoli. 1959. Systematic and ecological notes
on some fishes collected in the vicinity of Torruguero, Caribbean Costa Rica. Rev. Biol.
Trop., 7(1): 7-33.
Cashman, C. W. 1973. Contributions to the ichthyofaunas of the West Flower Garden
Reef and other sites in the Gulf of Mexico and western Caribbean. Unpubl. Ph.D.
dissertation, Texas A & M Univ., College Station, 248 pp.
Cervigon, F. M. 1966. Los peces marinos de Venezuela. Estacion de Investigariones
Marinas de Margarita, Fundacion La Salle de Ciencias Naturales, Monogr., no. 11, pp.
1-951.
1968. Los peces marinas de Venezuela. Complemento I. Mem. Soc. Cienc.
Nat. La Salle, XXVIII: 177-218.
Clarke, R. D. 1977. Habitat distribution and species diversity of chaetodontid and
pomacentrid fishes near Bimini, Bahamas. Marine Biol., 40: 277-289.
Colin, P. L. 1974. Observation and collection of deep-reef fishes off the coasts of Jamaica
and British Honduras (Belize). Marine Biol., 24: 29-38.
Colin, P. L. and M. F. Gomon. 1973. Notes on the behavior, biology and distribution of
Lucayablennius zingaro (Pisces: Clinidae). Caribb. J. Sri., 13: 59-61.
Dawson, C. E. 1970. The Caribbean Atlantic blenny Lupinoblennius dispar (Tribe: Blenniini),
with observations on a Pacific population. Proc. Biol. Soc. Wash., 83 (26): 273-286.
Downie, N. M. and R. W. Heath. 1959. Basic statistical methods. New York, Harper and
Brothers, 289 pp.
Ebeling, A. W., R. M. Ibara, R. J. Lavenberg and F. J. Rohlf. 1970. Ecological groups of
deep-sea animals off Southern California. Bull. Los Angeles County Mus. Nat. Hist.,
no. 6, 58 pp.
GREENFIELD & JOHNSON: BLENNIOID FISHES 103
Echelle, A. and G. D. Schnell. 1976. Factor analysis of species associations among
fishes of the Kiamichi River, Oklahoma. Trans. Am. Fish. Soc, 105 (1): 17-31.
Fager, E. W. 1957. Determination and analysis of recurrent groups. Ecology, 38 (4): 586-
595.
Fager, E. W. and A. R. Longhurst. 1968. Recurrent group analysis of species as-
semblages of demersal fish in the Gulf of Guinea. J. Fish. Res. Bd. Canada, 25 (7):
1405-1421.
Fager, E. W. and J. A. McGowan. 1963. Zooplankton species groups in the North Pacific.
Science, 140 (3566): 453-460.
Fowler, H. W. 1903. Notes on a few fishes from the Mosquito Coast of Nicaragua. Proc.
Acad. Nat. Sci. Phila., 55: 346-350.
1916. Cold-blooded vertebrates from Costa Rica and the Canal Zone. Proc.
Acad. Nat. Sci. Phila., 68: 389-414.
1917. A second collection of fishes from the Panama Canal Zone. Proc.
Acad. Nat. Sci. Phila., 69: 127-136.
1923. Fishes from Nicaragua. Proc. Acad. Nat. Sci. Phila., 75: 23-32.
1930. Notes on tropical American fishes. Proc. Biol. Soc. Wash., 43: 145-148.
1944. Results of the Fifth George Vanderbilt Expedition (1941) (Bahamas,
Caribbean Sea, Panama, Galapagos Archipelago and Mexican Pacific Islands). The
Fishes. Acad. Nat. Sci. Phila., Monogr., no. 6, 583 pp.
1950. Results of the Catherwood-Chaplin West Indies Expedition, 1948.
Part III. The Fishes. Proc. Acad. Nat. Sci. Phila., 102: 69-93.
George A., and V. G. Springer. 1980. Revision of the clinid fish tribe Ophiclinini, in-
cluding five new species, and definition of the family Clinidae. Smithson. Contr. Zool.,
No. 307, 31 pp.
Gilbert, C. R. 1965. Starksia y-lineata, a new clinid fish from Grand Cayman Id., British
West Indies. Not. Nat., 379: 1-6.
1971. Two new Atlantic clinid fishes of the genus Starksia. Quart. J. Fla.
Acad. Sci. (1970), 33 (3): 193-206.
1973. Characteristics of the Western Atlantic reef-fish fauna. Quart. J. Fla.
Acad. Sci., 35 (2+3): 130-144.
Gilbert, C. R. and D. P. Kelso. 1971. Fishes of the Tortuguero area, Caribbean Costa
Rica. Bull. Fla. State Mus. Biol. Sci., 16(1): 1-54.
Greenfield, D. W. 1972. Notes on the biology of the arrow blenny, Lucayablennius zingaro
(Bohlke) from British Honduras. Copeia, 1972 (3): 590-592.
1975. Emblemariopsis pricei, a new species of chaenopsid blenny from Belize.
Copeia, 1975 (4): 713-715.
1978. A review of the western Atlantic Starksia ocellata — complex (Pisces:
Clinidae) with the description of two new species and proposal of superspecies status.
Fieldiana: Zoology, 73 (2): 9-48.
Greenfield, D. W. and T. A. Greenfield. Habitat and resource partitioning between two
species of Acanthemblemaria (Pisces: Chaenopsidae), with comments on the chaos
hypothesis. In Atlantic Barrier Reef Ecosystems. Carrie-Bow Cay, Belize, Sci. Rep. No.
1, Smithson. Inst. Contr. Marine Sci., in press.
Hagmeier, E. M. and C. D. Stults. 1964. A numerical analysis of the distributional
patterns of North American mammals. Syst. Zool., 13: 125-155.
Hayes, W. B. 1978. Some sampling properties of the Fager index for recurrent species
groups. Ecology, 59 (1): 194-196.
Hubbs, C. 1953. Revision of the Eastern Pacific fishes of the clinid genus Labrisomus.
Zoologica, 38 (3): 113-136.
Inger, R. F. and P. K. Chin. 1962. The fresh-water fishes of North Borneo. Fieldiana:
Zoology, 45: 1-268.
Johnson, R. K. and D. W. Greenfield. 1976. A new chaenopsid fish, Etnblemaria hyltoni,
from Isla Roatan, Honduras. Fieldiana: Zoology, 70 (2): 13-28.
104 FIELDIANA: ZOOLOGY
Jordan, D. W. and B. W. Evermann. 1898. The fishes of North and Middle America: a
descriptive catalogue of the species of fishlike vertebrates found in the waters of North
America north of the Isthmus of Panama. Smith. Inst., T. F. H. Fund Reprint 1963, 3:
1937-2860.
Mago, F. L. 1970. Lista de los Peces de Venezuela, incluyendo un estudio preliminar sobre
la ictiogeografia del pais. Ministerio de Agricultura y Cria, Oficina Nacional de Pesca,
Caracas, Venezuela, pp. 1-283.
Meek, S. E. and S. F. Hildebrand. 1923-1928. The marine fishes of Panama. Field Mus.
Nat. Hist., Zool. Ser., 15 (1-3): 1045 pp.
1928. The marine fishes of Panama. Field Mus. Nat. Hist., Zool. Ser., 15
(part III): 709-1045.
Miller, G. C. 1968. A revision of zoogeographical regions in the warm water area of the
Western Atlantic. FAO Fish. Rept., 71 (1): 141.
Miller, R. R. 1966. Geographical distribution of Central American freshwater fishes.
Copeia, 1966 (4): 773-802.
1976. Geographic distribution of Central American freshwater fishes. In T.
B. Thorsen (ed.), Investigations of the ichthyofauna of Nicaraguan lakes. University of
Nebraska (Lincoln), pp. 125-156.
Nagelkerken, W. P. 1974. On the occurrence of fishes in relation to corals in Curacao.
Stud. Fauna Curacao and Other Carib. Is., XLV (147): 118-141.
Norman, J. R. 1943. Notes on the blennioid fishes. I. A provisional synopsis of the genera
of the family Blenniidae. Ann. Mag. Nat. Hist., ser. 11, 10: 793-812.
Nursall, J. R. 1977. Territoriality in redlip blennies (Ophioblennius atlanticus — Pisces:
Blenniidae). J. Zool. (London), 182: 205-223.
Palacio, F. J. 1974. Peces colectados en el Caribe Colombiano por la Universidad de
Miami. Bol. Museo del Mar, Bogota Bol., no. 6, 137 pp.
Peters, J. A. 1968. A computer program for calculating degree of biogeographical re-
semblance between areas. Syst. Zool., 17 (1): 64-69.
Pinto, S. Y. 1954. Fauna do Distrito Federal — X Redescricao de Blennius cristatus Linnaeus
1758 (Perciformes-Blenniidae). Bol. Inst. Oceanogr., Univ. Sao Paulo, 5 (1 and 2): 213-
232.
Prentice, S. A. and J. M. Kain. 1976. Numerical analysis of subtidal communities on
rocky shores. Estuarine and Coastal Mar. Sci., 4: 65-70.
Preston, F. W. 1962. The canonical distribution of commonness and rarity, Part II. Ecol-
ogy, 43 (3): 410-432.
Randall, J. E. 1966. The West-Indian blenniid fishes of the genus Hypleurochilus, with the
description of a new species. Proc. Biol. Soc. Wash., 79: 57-71.
1968. Caribbean reef fishes. T. F. H. Publ., 318 pp.
Robins, C. R. 1971a. Comments on Chaenopsis stephensi and Chaenopsis resh, two Caribbean
blennioid fishes. Carib. J. Sci., 11 (3-4): 179-180.
1971b. Distributional patterns of fishes from coastal and shelf waters of the
tropical western Atlantic. Symposium on investigations and resources of the Caribbean
Sea and adjacent regions. Papers on Fishery Resources. FAO, Rome, pp. 249-255.
.. 1972. The state of knowledge of the coastal fish fauna of the Panamic Region
prior to the construction of an interoceanic sea-level canal. Bull. Biol. Soc. Wash., 1972
(2): 159-166.
Robins, C. R. and J. E. Randall. 1965. Three new western Atlantic fishes of the blennioid
genus Chaenopsis with notes on the related Lucayablennius zingaro. Proc. Acad. Nat. Sci.
Phila., 117(6): 213-234.
Rosen, D. E. 1975. A vicariance model of Caribbean biogeography. Syst. Zool., 24 (4):
431-464.
Rosenblatt, R. H. 1959. A revisionary study of the blennioid fish family Tripterygiidae.
Unpubl. Ph.D. dissertation, Univ. California, Los Angeles, 376 pp.
GREENFIELD & JOHNSON: BLENNIOID FISHES 105
1960. The Atlantic spedes of the blennioid fish genus Enneanectes. Proc.
Acad. Nat. Sci. Phila., 112 (1): 1-23.
.. 1963. Some aspects of speciation in marine shore fishes. Syst. Assoc. Publ.
no. 5, Speciation in the Sea, pp. 171-180.
Rosenblatt, R. H. and T. D. Parr. 1969. The Pacific species of the clinid fish genus
Paraclinus. Copeia, 1969 (1): 1-20.
Rosenblatt, R. H. and L. R. Taylor, Jr. 1971. The Pacific species of the clinid fish tribe
Starksiini. Pacific Sci., 25 (3): 436-463.
Rubinoff, I. and R. W. Rubinoff. 1962. New records of inshore fishes from the Atlantic
coast of Panama. Breviora, no. 169, pp. 1-7.
Schmitt, W. L. and L. P. Schultz. 1940. List of the fishes taken on the Presidential Cruise
of 1938. Smith. Misc. Coll., 98 (25): 1-10.
Shipp, R. L. 1975. Pirates in the northern Gulf of Mexico. The Marine Aquarist, 6 (7): 16-20.
Smith, C. L. and C. R. Powell. 1971. The summer fish communities of Brier Creek,
Marshall County, Oklahoma. Amer. Mus. Novit., No. 2458, pp. 1-50.
Smith, G. B. 1976. Ecology and distribution of eastern Gulf of Mexico reef fishes. Fla.
Marine Res. Publ., Fla. Dept. Nat. Res., Publ. no. 19, pp. 1-78.
Smith, G. B., H. M. Austin, S. A. Bortone, R. W. Hastings and L. H. Ogren. 1975.
Fishes of the Florida Middle Ground with comments on ecology and zoogeography. Fla.
Marine Res. Publ. 9, pp. 1-14.
Smith-Vaniz, W. F. and F. J. Palacio. 1974. Atlantic fishes of the genus Acanthemblemaria
with description of three new species and comments on the Pacific species (Clinidae;
Chaenopsidae). Proc. Acad. Nat. Sci. Phila., 125 (11): 197-224.
Sonnier, F., J. Teerling and H. D. Hoese. 1976. Observations on the offshore reef and
platform fish fauna of Louisiana. Copeia, 1976 (1): 105-111.
Springer, V. G. 1955a. Western Atlantic fishes of the genus Paraclinus. Texas J. Sci., 6 (4):
422-441.
1955b. The taxonomic status of the fishes of the genus Stathmonotus includ-
ing a review of the Atlantic species. Bull. Mar. Sci. Gulf & Caribbean, 5 (1): 66-80.
1959a. (1958 on paper). Systematics and zoogeography of the clinid fishes of
the subtribe Labrisomini Hubbs. Publ. Inst. Mar. Sci., Univ. Texas, 5: 417-492.
1959b. A new species of Labrisomus from the Caribbean Sea, with notes on
other fishes of the subtribe Labrisomini. Copeia, 1959 (4): 289-292.
1962. A review of the blenniid fishes of the genus Ophioblennius Gill.
Copeia, 1962 (2): 426-433.
1967. Revision of the circumtropical shorefish genus Entomacrodus (Blen-
niidae: Salariinae). Proc. U.S. Natl. Mus., 122 (3582): 1-150.
1968. Osteology and classification of the fishes of the family Blenniidae.
Bull. U.S. Nat. Mus., 284, 85 pp.
1972. Additions to revisions of the blenniid fish genera Ecsenius and En-
tomacrodus, with descriptions of three new species of Ecsenius. Smith. Contr. Zool., 134:
1-13.
Springer, V. G. and M. F. Gomon. 1975. Variation in the Western Atlantic clinid fish
Maiacoctenus triangulatus with a revised key to the Atlantic species of Malacoctenus.
Smith. Contr. Zool., no. 200, pp. 1-11.
Springer, V. G. and R. H. Rosenblatt. 1965. A new blennioid fish of the genus Lab-
risomus from Ecuador, with notes on the Caribbean species L. filamentosus. Copeia, 1965
(1): 25-27.
Springer, V. G. and W. F. Smith-Vaniz. 1970. Blennius antholops, new deep-water fish,
from the Gulf of Guinea, with comments on the bathymetric distribution of the family
Blenniidae. Proc. Biol. Soc. Wash., 83 (2): 215-220.
Stephens, J. S., Jr. 1963. A revised classification of the blennioid fishes of the American
family Chaenopsidae. U. C. Publ. Zool., 68, 165 pp.
106 FIELDIANA: ZOOLOGY
.. 1970. Seven new chaenopsid blennies from the western Atlantic. Copeia,
1970 (2): 280-309.
Stephens, J. S., Jr., E. S. Hobson and R. K. Johnson. 1966. Notes on distribution,
behavior, and morphological variation in some chaenopsid fishes from the tropical east-
ern Pacific with descriptions of two new species, Acanthemblemaria castroi and Coral-
liozetus springeri. Copeia, 1966 (3): 424-438.
Tate, M. W. and R. C. Clelland. 1957. Non parametric and shortcut statistics. Interstate
Printers and Publishers, Danville, 111., 171 pp.
Tavolga, W. N. 1954. A new species of fish of the genus Blennius from Florida. Copeia,
1954 (2): 135-139.
Wickler, W. 1967. Specializations of organs having a signal function in some marine fish.
Stud. Trop. Oceanogr., 5: 539-548.
Field Museum of Natural History
Roosevelt Road at Lake Shore Drive
Chicago, Illinois 60605
Telephone: (312) 922-9410
Live Music Archive
Librivox Free Audio
Metropolitan Museum
Cleveland Museum of Art
Internet Arcade
Console Living Room
Books to Borrow
Open Library
TV News
Understanding 9/11