JAW MORPHO-FUNCTIONAL DIVERSITY, TROPHIC ECOLOGY, AND
HISTORICAL BIOGEOGRAPHY OF THE NEOTROPICAL
SUCKERMOUTH ARMORED CATFISHES
(SILURIFORMES, LORICARIIDAE)
Except where reference is made to the work of others, the work described in this
dissertation is my own or was done in collaboration with my advisory committee. This
dissertation does not include proprietary or classified information.
_____________________________
Nathan K. Lujan
Certificate of Approval:
__________________________________________________________
Dennis R. DeVriesJonathan W. Armbruster, Chair
Professor Associate Professor
Fisheries and Allied Aquacultures Biological Sciences
__________________________________________________________
Jack W. FeminellaCraig Guyer
Professor Professor
Biological SciencesBiological Science
_____________________________
George T. Flowers
Dean
Graduate School
JAW MORPHO-FUNCTIONAL DIVERSITY, TROPHIC ECOLOGY, AND
HISTORICAL BIOGEOGRAPHY OF THE NEOTROPICAL
SUCKERMOUTH ARMORED CATFISHES
(SILURIFORMES, LORICARIIDAE)
Nathan K. Lujan
A Dissertation
Submitted to
the Graduate Faculty of
Auburn University
in Partial Fulfilment of the
Requirements for the
Degree of
Doctor of Philosophy
Auburn, Alabama
May 9, 2009
iii
JAW MORPHO-FUNCTIONAL DIVERSITY, TROPHIC ECOLOGY, AND
HISTORICAL BIOGEOGRAPHY OF THE NEOTROPICAL
SUCKERMOUTH ARMORED CATFISHES
(SILURIFORMES, LORICARIIDAE)
Nathan K. Lujan
Permission is granted to Auburn University to make copies of this dissertation at its
discretion, upon request of individuals or institutions and at their expense. The author
reserves all publication rights.
_____________________________
Signature of Author
_____________________________
Date of Graduation
iv
VITA
Nathan K. Lujan, son of Leo R. Lujan and Jeanne Lujan, brother of Stephanie L.
Rickerman and Tracy M. Lujan, was born on April 29, 1976, in Nashville, Tennessee. He
graduated from Montgomery Bell Academy in Nashville, TN in 1995, then moved to
Grand Rapids, MI where he earned a Bachelor of Science in Biology from Calvin
College in 2000. Following two years with the Tennessee Valley Authority, he entered
the Department of Biological Sciences under the guidance of Dr. Jonathan W.
Armbruster in 2002 and earned a Doctor of Philosophy in 2009.
v
DISSERTATION ABSTRACT
JAW MORPHO-FUNCTIONAL DIVERSITY, TROPHIC ECOLOGY, AND
HISTORICAL BIOGEOGRAPHY OF THE NEOTROPICAL
SUCKERMOUTH ARMORED CATFISHES
(SILURIFORMES, LORICARIIDAE)
Nathan K. Lujan
Ph. D., Auburn University, May 9, 2009
(B.S. Calvin College, 2000)
250 Typed Pages
Directed by Jonathan W. Armbruster
Tropical South America is renowned for its unparalleled biodiversity. Among
vertebrates, fish diversity exceeds that of any other group. Estimates of total Neotropical
fish richness range between 5000 and 8000 species, of which approximately 3600
(45?72%) are currently described; and of described species, nearly 20% (>700 spp.) are
catfishes in the Neotropical-endemic family Loricariidae. Loricariid catfishes, popularly
known as plecos in the aquarium trade, are distinguished by their armor plating, ventral
oral disk, and highly derived jaw structure and function. Loricariids have likely existed in
South American rivers in close to their modern form since at least the Late Cretaceous,
and they are part of a superfamilial lineage (Loricarioidea) that is sister to all other
vi
catfishes, and has likely inhabited South American rivers since well into the Early
Cretaceous. Today, loricariids have radiated to consume a variety of basal food resources
including algae, detritus, seeds, sponges, insects, and even wood, the surface layers of
which are gouged into by specialized taxa having hypertrophied jaw muscles and teeth
shaped like adzes.
In this dissertation, I attempt to fill major gaps in the knowledge of loricariid
taxonomy, jaw morphological and functional diversity, trophic ecological structure, and
historical biogeography. My taxonomic research has resulted in the discovery and/or
description of dozens of new loricariid species and at least three new genera. Published
results of this work are summarized in Appendix I. In chapter two of this dissertation, I
describe jaw morpho-functional diversity across a diverse assemblage of 25 species, 12
genera, five tribes, and two subfamilies of loricariids from the upper Amazon Basin in
Northern Peru. In chapter three, I describe gross aspects of loricariid assemblage trophic
structure as revealed by carbon and nitrogen isotope data from 19 loricariid assemblages
ranging in species richness from two to 16 species, and geographically broadly
distributed across northern South America. In chapter 4, I review the geological and
hydrological history of the Guiana Shield, a highly biodiverse and geologically ancient
region of northern South America which shelters a broad range of basal and derived
loricariid lineages. From these studies, it is clear that tremendous loricariid diversity
accumulated in the Neotropics gradually over tens of millions of years and across a broad
geographic range, and that their novel oral morphology has likely been key to their
diversification across a variety of basal resources consumed almost exclusively by
invertebrate faunas at more temperate latitudes.
vii
ACKNOWLEDGEMENTS
Comparative investigations of loricariid ecology, jaw morphology and function
face many hurdles. Beyond incomplete and rapidly changing taxonomies, weakly
supported phylogenies, and paucity of museum specimens, are the financial, logistical
and political challenges of reaching parts of South America where loricariid catfishes are
most diverse. I surmounted none of these obstacles alone. Indeed, the research described
in this dissertation would not have been possible without the generous assistance of many
people, in matters academic, logistic and domestic. For ensuring that I remained clothed,
sheltered, fed, and mentally and physically healthy while in the United States, I would
like to gratefully acknowledge the indispensable support of my parents Jeanne and Leo
Lujan, and sisters Tracy Lujan and Stephanie Rickerman. For enthusiastically funding my
international travels, while maintaining near limitless patience with my frequently
digressionary research, slow dissertation development, and half-cocked first drafts, I
thank my advisor Jon Armbruster. For funding of research and domestic travel and for
feeding my earliest flames of interest in tropical fishes, I thank Chris Beggin, owner of
the Aquatic Critter in Nashville, TN. For their friendship and indefatigable assistance in
the field and in the museum, I thank Mark Sabaj P?rez, Donald Taphorn and David
Werneke. For being open to discussions of my research at a moment?s notice, I thank my
committee: Dennis Devries, Jack Feminella, Craig Guyer, and long-distance advisor Kirk
Winemiller. For equally impromptu discussions of my hair-brained hypotheses and novel
viii
data applications, I thank my fellow graduate students Jeff Stratford, Richard Mitchell,
Steve Herrington, Michael Lowe, Brian Helms; and my labmates Marcelo Melo, Keith
Ray, Ricardo Betancur, Lesley deSouza, and Shobnom Ferdous. For the frequent loans of
specimens and access to collections, I thank collection managers Mark Sabaj P?rez
(ANSP), Sandra Raredon (NMNH), Karsten Hartel (MCZ), and Mary Anne Rogers
(FMNH). Finally, I would like to thank those people who frequently went above and
beyond their individual responsibilities to facilitate fieldwork in several South American
countries. For field and museum work in Venezuela, I thank Oscar Leon Mata, Augusto
Luna, Octaviano Santaella, Mariangeles Arce, Mayme Grant, Erin Richmond, Tim
Wesley, David Brooks, Juan Valadez, Raphael Pajua, Luiz Camico, and Franci Britto.
For field and museum work in Peru, I thank Hernan Ortega, Blanca Rengifo, Darwin
Osorio, Max Hidalgo, and Vanessa Meza. For field and museum work in Brazil, I thank
Osvaldo Oyakawa, Jose Birindelli, Leandro Sousa, and Andr? Netto-Ferreira. For field
and museum work in Guyana, I thank Calvin Bernard, Elford Livingston, and Stacy Lord.
This research was funded by the Planetary Biodiversity Inventory: All Catfish Species
(Siluriformes), a five year grant through the US National Science Foundation (NSF DEB-
0315963) and NSF grant DEB-0107751 to Jonathan W. Armbruster.
ix
Style manual or journal used: Ecology (Chapters 1, 3), Zoology (Chapter 2), Journal of
Fish Biology (Chapter 4).
Computer software used: Microsoft Office X software suite (v. X), JMP statistical
software (SAS Institute, v. 5.0.1), tpsDIG2 software (F. James Rohlf, v. 2.12), Oriana
circular statistics software (Rockware Inc.), Adobe Photoshop (v. 8.0).
x
TABLE OF CONTENTS
LIST OF TABLES...................................................xii
LIST OF FIGURES..................................................xiii
1.INTRODUCTION................................................1
1.1GENERAL INTRODUCTION...............................1
1.2INTRODUCTION TO THE LORICARIIDAE...................3
1.3INTRODUCTION TO DISSERTATION RESEARCH.............5
1.4REFERENCES..........................................10
2. METHODOLOGICAL APPROACHES TO THE QUANTIFICATION AND
COMPARISON OF LOWER JAW MORPHO-FUNCTIONAL DIVERSITY
AMONG SUCKERMOUTH ARMORED CATFISHES (SILURIFORMES,
LORICARIIDAE)................................................23
2.1ABSTRACT............................................23
2.2INTRODUCTION........................................25
2.3METHODS.............................................32
2.3.1 SPECIMEN COLLECTION, PREPARATION, AND IMAGING
................................................32
2.3.2MORPHOMETRICS................................34
2.3.3JAW MECHANICS.................................35
2.3.4JAW MECHANICS: FORCE INTENSITY...............36
2.3.5JAW MECHANICS: FORCE GEOMETRY..............37
2.4RESULTS AND DISCUSSION.............................40
2.4.1 GROSS MORPHOLOGICAL DIFFERENTIATION.......40
2.4.2JAW MECHANICS: FORCE INTENSITY...............43
2.4.3 JAW MECHANICS: FORCE GEOMETRY-
TRADITIONAL METRICS..........................48
2.4.4 JAW MECHANICS: FORCE GEOMETRY-
NOVEL METRICS.................................50
2.5CONCLUSIONS.........................................54
2.6REFERENCES..........................................56
3. STABLE ISOTOPES REVEAL TROPHIC STRUCTURE WITHIN
SYMPATRICALLY DIVERSE ASSEMBLAGES OF NEOTROPICAL
DETRITIVOROUS FISHES........................................82
3.1 INTRODUCTION........................................82
3.2METHODS.............................................89
3.3RESULTS..............................................92
3.4DISCUSSION...........................................93
3.5REFERENCES.........................................100
xi
4.GEOLOGICAL AND HYDROLOGICAL HISTORY OF THE GUIANA SHIELD
AND HISTORICAL BIOGEOGRAPHY OF ITS FISHES................132
4.1SUMMARY...........................................132
4.2INTRODUCTION.......................................135
4.3GEOLOGY AND HYDROLOGY..........................137
4.3.1OVERVIEW.....................................137
4.3.2TOPOGRAPHIC EVOLUTION......................139
4.3.3PROTO-BERBICE (CENTRAL SHIELD)..............142
4.3.4PROTO-ORINOCO (WESTERN SHIELD).............146
4.3.5EASTERN VENEZUELA BASIN (NORTHERN SHIELD).147
4.3.6PROTO-AMAZON AND EASTERN ATLANTIC DRAINAGES
(SOUTHERN AND EASTERN SHIELD)...............149
4.3.7ARIDITY AND MARINE INCURSIONS...............152
4.3.8LIMNOLOGY AND GEOCHEMISTRY OF GUIANA SHIELD
RIVERS........................................155
4.4BIOGEOGRAPHY OF GUIANA SHIELD FISHES.............156
4.4.1 MODERN CORRIDORS: THE PRONE-8..............156
4.4.2CARONI (ORINOCO) TO CUYUNI/MAZARUNI
CORRIDORS....................................158
4.4.3CASIQUIARE PORTAL............................160
4.4.4SOUTHERN GUIANA SHIELD AND NORTHERN
BRAZILIAN SHIELD CORRIDORS..................163
4.4.5RUPUNUNI PORTAL.............................164
4.4.6ATLANTIC COASTAL CORRIDORS.................165
4.4.7RELICTUAL FAUNAS............................168
4.5CONCLUSIONS........................................173
4.6REFERENCES.........................................175
APPENDIX I: CITATIONS AND ABSTRACTS OF PUBLISHED OR SUBMITTED
MANUSCRIPTS...................................................197
APPENDIX II: VECTORS FROM LORICARIID ASSEMBLAGE ?
13
C AND ?
15
N
CENTROIDS TO SPECIES MEANS. DATA FROM 19 LOCALITIES AND 83
SPECIES POOLED AND SORTED ACCORDING TO TAXON. WHEN DATA
PERMIT, DIRECTION OF MEAN VECTOR FOR EACH TAXON SHOWN AS
DASHED RADIUS, AND 95% CIRCULAR CONFIDENCE INTERVALS ARE
SHOWN AS EITHER DASHED (SIGNIFICANT, P<0.05) OR DOTTED (NOT
SIGNIFICANT, P?0.05) ARCS. VECTORS FOR SPECIES REPRESENT
INDIVIDUAL SAMPLES. VECTORS FOR ALL HIGHER TAXA REPRESENT
SPECIES MEANS..................................................205
xii
LIST OF TABLES
Chapter 1
Table 1:Summary of published descriptions of loricariid diets. Only dominant
constituents are reported where gut contents were reported as extended lists of
taxa. Species identity updated to reflect current taxonomy. References: 1=Alvim
and Peret, 2004; 2=Armbruster, 2002; 3=Armbruster, 2003; 4=Armbruster,
2004; 5=Cardone et al., 2006; 6=DeLariva and Agostinho, 2001; 7=Ferreira,
2007; 8=Forsberg et al., 1993; 9=Fugi et al., 1996; 10=Hamilton et al., 1992;
11=Hood et al., 2005; 12=Jepsen and Winemiller, 2002; 13=Kramer and
Bryant, 1995; 14=Melo et al., 2004; 15=M?rigoux and Ponton, 1998;
16=M?rona et al., 2008; 17=Nonogaki et al., 2007; 18=Peretti and Andrian,
2004; 19=Power, 1981, including several subsequent papers by this author;
20=Saul, 1975; 21=Schaefer and Stewart, 1993; 22=Vaz et al., 1999;
23=Winemiller, 1990; 24=Yossa and Araujo-Lima, 1998; 25=Zaret and Rand,
1971 (nr = not recorded)........................................17
Chapter 2
Table 1:Taxa and sample sizes examined in respective jaw and muscle parts of this
study. All specimens collected from middle reaches of the Mara?on River, a
tributary of the upper Amazon in northern Peru.......................65
Chapter 3
Table 1:19 loricariid assemblages sampled in this study, their taxonomic diversity, and
results of community-wide stable isotope analyses proposed by Layman et al.
(2008).....................................................117
Chapter 4
Table 1:Planation surfaces, their age, elevation, and name in each country of the Guiana
Shield (after Schubert et al., 1986, Briceno and Schubert, 1990, Gibbs and
Barron, 1993)...............................................189
xiii
LIST OF FIGURES
Chapter 1
Figure 1.Representative diversity of loricariid upper and lower jaw morphologies: A.
Leporacanthicus (carnivore), B. Panaque (wood-eater), C. Chaetostoma
(periphytivore)...............................................22
Chapter 2
Figure 1. Medial, sagittal sections through the snout and jaws of (A) Chaetostoma cf.
milesi, (B) Leporacanthicus joselimai, and (C) Panaque nigrolineatus. Labels:
ap = ascending process of the premaxilla, dn = dentary, me = mesethmoid, mec
= mesethmoid condyle, pm = premaxilla, rpf = retractor premaxillae fossa. CT
(computer aided tomography) reconstructions courtesy of the Digimorph
lab.........................................................66
Figure 2. Ventral views of the snout and jaws of (A) Chaetostoma cf. milesi, (B)
Leporacanthicus joselimai, and (C) Panaque nigrolineatus. Labels: dn =
dentary, pm = premaxillae, aac = anguloarticular condyle. CT (computer aided
tomography) reconstructions courtesy of the Digimorph lab.............67
Figure 3. Representative sample of lower jaws from upper Amazon Loricariidae
examined in this study: HYPOSTOMINAE: ANCISTRINI: (A) Ancistrus sp.
?longjaw?, (B) Ancistrus sp. ?shortjaw?, (C) Chaetostoma sp. 1, (D) Panaque
albomaculatus, (E) Panaque gnomus, (F) Panaque n. sp. ?Mara?on?, (G)
Panaque nocturnus, (H) Peckoltia bachi, HYPOSTOMINAE:
HYPOSTOMINI: (I) Hypostomus emarginatus, (J) Hypostomus niceforoi, (K)
Hypostomus pyrineusi in the H. cochliodon-group, (L) Hypostomus unicolor,
LORICARIINAE: HARTTIINI: (M) Farlowella amazona, (N) Lamontichthys
filamentosus, (O) Rineloricaria lanceolata, LORICARIINAE: LORICARIINI:
(P) Spatuloricaria puganensis....................................68
Figure 4. Examples of loricariid oral disks: (A) Scobinancistrus sp. and (B)
Leporacanthicus cf. galaxias. Photos by M. Sabaj P?rez................69
Figure 5.The lower jaw ramus of Chaetostoma cf. milesi (A, B, C) and Panaque
nigrolineatus (D, E, F) in positions illustrating the horizontal orientation, ventral
perspective (A, D), vertical orientation (B, E), and horizontal orientation, dorsal
perspective (C, F). Functional parameters are labled on each perspective and
xiv
view from which they were measured: AM
area
 = adductor mandibulae insertion
area, In = in-lever from center of adductor mandibulae area of insertion to
anguloarticular condyle, Out
dist
 = out-lever from anguloarticular condyle to
distalmost tooth, Out
prox
 = out-lever from angularticular condyle to
proximalmost tooth, TRL = tooth row length. CT (computer aided tomography)
reconstructions courtesy of the Digimorph lab and Kyle Luckinbill, ANSP..70
Figure 6. The lower jaw ramus of (A) Chaetostoma cf. milesi and (B) Panaque
nigrolineatus in horizontal orientation, dorsal perspective, illustrating distances
measured as the parameters Out
dist
 and H1. CT (computer aided tomography)
reconstructions courtesy of the Digimorph lab and Kyle Luckinbill, ANSP..71
Figure 7.Relationship between combined adductor mandibulae area of insertion (AM
area
)
scaled to standard length (SL), and adductor mandibulae volume (AM
vol
) scaled
to standard length. Area of adductor mandibulae insertion and volume of
adductor mandibulae were measured from separate subsets of specimens, so the
correlation is of means of each scaled separately to standard length. See Table 1
for species sample sizes........................................72
Figure 8.Three-dimensional rotating cone model linking loricariid lower jaw morphology
and function. Cones modeled after the right lower jaw ramus of Chaetostoma cf.
milesi (Cone 1) and Panaque nigrolineatus (Cone 2) illustrate near opposite
ends of a spectrum of both measured and qualitatively described jaw
morphology discussed in this study. Perspective is from the posterior. Red lines
are potential axes of rotation, with Axis 1 representing rotation in the sagittal
plane, Axis 2 representing rotation in the horizontal plane, and Axis 3
representing rotation in the transverse plane. Capitalized parameters and
parameters represented by solid green, black, and gray lines were measured or
quantified by proxy in this study (Fig. 5). See text for discussion of output plane
represented by gray triangle. Tooth row angle (TR
angle
) represented by ?....73
Figure 9.Principle component (PC) analysis of six parameters hypothesized to be
functionally relevant to the loricariid lower jaw (see Figs. 5, 6). Data from 379
individuals and 25 species of Loricariidae collected from the upper Amazon in
northern Peru (see Table 1). Parameters listed along each PC axis from left to
right or from bottom to top in order of eigenvector magnitude. PCs one through
four describe 85.7%, 11.9%, 2.0%, and 0.3% of variation in the dataset,
respectively.................................................74
Figure 10. Principle component (PC) analysis of six parameters hypothesized to be
functionally relevant to the loricariid lower jaw (see Figs. 5, 6). Data from 379
individuals and 25 species of Loricariidae collected from the upper Amazon in
northern Peru (see Table 1). Parameters listed along each PC axis from left to
right or from bottom to top in order of eigenvector magnitude. PCs one through
xv
four describe 85.7%, 11.9%, 2.0%, and 0.3% of variation in the dataset,
respectively.................................................75
Figure 11. Adductor mandibulae area of insertion (AM
area
) over tooth row length (TRL),
hypothesized to predict force intensity, with high values indicating force
concentration and low values indicating force distribution. See Table 1 for
species sample sizes...........................................76
Figure 12. Examples of convergent trends in herbivorous jaw morphologies from (A)
Insecta, Trichoptera (labrum in dorsal view modified from Satija & Satija,
1959), (B) fossil Sauropoda, Nigersaurus taqueti (lower jaw in dorsal view
modified from Sereno et al., 2007), (C) Cypriniformes, Gastromyzontinae
(snout in ventral view modified from Roberts, 1989), and (D) extinct
Mammalia, Giraffidae (premaxillae in ventral view modified from Solounias &
Moelleken, 1993). Series in (C) hypothesized to be that of a carnivore
(narrowest, bottom) to herbivore (broadest, top). Series in (D) hypothesized to
represent a foraging spectrum from browser (narrowest, bottom) to mixed
grazer (middle) to grazer (broadest, top), with support provided by fecal data
(Solounias & Moelleken, 1993)...................................77
Figure 13. The novel metric of force intensity AM
area
/TRL (see Fig. 11) plotted against
traditional metrics of mechanical advantage: distance from anguloarticular
condyle to center of adductor mandibulae insertion area (In) over respective
distances from anguloarticular condyle to distalmost (Out
dist
) and proximalmost
(Out
prox
) tooth insertions. Capital letters indicate jaws illustrated in Fig. 3. See
Table 1 for species sample sizes..................................78
Figure 14. Three novel metrics of lower jaw morphological diversity hypothesized to be
linked to function: (1) Adductor mandibulae area of insertion (AM
area
) over
tooth row length (TRL) ? hypothesized to be a measure of force distribution
versus force concentration; (2) distance parameter H1 (see Fig. 7) over TRL ?
hypothesized to be a combined measure of mechanical advantage (force vs.
speed optimization) and force distribution; and (3) the distal out-lever arm
(Out
dist
, see Fig. 5) over H1 ? a measure of the gross length (Out
dist
) versu height
(H1) dimensions of the lower jaw and hypothesized to be an indicator of the
predominant plane of torque through the lower jaw. Capital letters indicate jaws
illustrated in Fig. 3. See Table 1 for species sample sizes...............79
Figure 15. Mean angles and angular standard deviations between the tooth row and
distalmost output lever for species examined in this study. See Table 1 for
sample sizes.................................................80
Figure 16. The metric Out
dist
-Out
prox
/In, a measure of torque differential across the tooth
arcade. See Table 1 for sample sizes...............................81
xvi
Chapter 3.
Figure 1. Drainage map of northern South America with regions sampled in this study
marked by dotted ovals. Region 1 (Venezuela) included the Casiquiare
Assemblage (7 spp.), the Orinoco-1 Assemblage (11 spp.), the Orinoco-2
Assemblage (11 spp.), and the Ventuari Assemblage (14 spp.). Region 2
(Guyana) included the Bununi Assemblage (5 spp.), Rupununi Assemblage (12
spp.), and the Takutu Assemblage (6 spp.). Region 3 (Peru) included the
Almendro Assemblage (5 spp.), the Huancabamba Assemblage (3 spp.), the
Mara?on Assemblage (16 spp.), the Nieva Assemblage (7 spp.), the Siasme-lwr
Assemblage (3 spp.), the Siasme-upr Assemblage (3 spp.), and the Utcubamba
Assemblage (5 spp.). Region 4 (Brazil) included the Peixoto Assemblage (3
spp.), the Jamanxim Assemblage (11 spp.), and the Curu? Assemblage (14
spp.)......................................................118
Figure 2. Representative diversity of loricariid upper and lower jaw morphologies: A.
Leporacanthicus, B. Panaque, C. Chaetostoma.....................119
Figure 3. Relationships between loricariid assemblage species richness and six metrics of
trophic structure proposed by Layman et al. (2008; see Introduction): (A) NR,
(B) CR, and (C) TA are indicators of whole assemblage trophic niche breadth,
whereas (D) CD, (E) NND, and (F) SDNND are indicators of species niche
breadth and spacing within each assemblage. Squares represent clearwater
shield habitats and circles represent whitewater Andean habitats.........120
Figure 4. Relationships between loricariid assemblage species richness and the absolute
values of (A) CD and (B) NND residuals. Squares represent clearwater shield
habitats and circles represent whitewater Andean habitats..............121
Figure 5. Relationships between loricariid assemblage tribe richness and (A) CD
calculated as the mean distance from the species centroid to each tribe mean
(trCD); and (B) NND calculated as the mean distance from each species to its
nearest neighbor within its own tribe (trNND). Squares represent clearwater
shield habitats and circles represent whitewater Andean habitats.........122
Figure 6. Comparison of species versus tribe NND for each assemblage sampled in this
study. Assemblages in ascending order of species richness from left to
right......................................................123
Figure 7. ?
15
N and ?
13
C relationships of loricariid plus astroblepid consumers in (A)
Shaapan creek, Mara?on (Amazon) drainage, northern Peru; (B) Peixoto river,
Tapaj?s (Amazon) drainage, Brazil; (C) upper Siasme creek, Mara?on
(Amazon) drainage, northern Peru; and (D) Bununi creek, Rupununi
(Essequibo) drainage, Guyana. Sample sizes in parentheses............124
xvii
Figure 8. ?
15
N and ?
13
C relationships of loricariid consumers in (A) the Utcubamba river,
Mara?on (Amazon) drainage, northern Peru; (B) Casiquiare canal, Negro
(Amazon) drainage, southern Venezuela; and (C) Takutu river, Branco
(Amazon) drainage, Guyana. Sample sizes in parentheses..............125
Figure 9. ?
15
N and ?
13
C relationships of loricariid plus astroblepid consumers in (A) the
Huancabamba river, and (B) lower Siasme creek, both Mara?on (Amazon)
drainages in northern Peru. Sample sizes in parentheses...............126
Figure 10. ?
15
N and ?
13
C relationships of loricariid plus astroblepid consumers in (A)
Almendro creek, and (B) the Nieva river, both Mara?on (Amazon) drainages in
northern Peru. Sample sizes in parentheses.........................127
Figure 11. ?
15
N and ?
13
C relationships of loricariid consumers in the (A) Essequibo river,
Guyana, and (B) the Jamanxim river, Tapajos (Amazon) drainage, Brazil.
Sample sizes in parentheses....................................128
Figure 12. ?
15
N and ?
13
C relationships of loricariid consumers in the Orinoco river main
channel, (A) below and (B) above the mouth of the Ventuari river. Sample sizes
in parentheses...............................................129
Figure 13. ?
15
N and ?
13
C relationships of loricariid consumers in (A) the Rupununi river,
Essequibo drainage, Guyana; and (B) the Curu? river, Xingu drainage, Brazil.
Sample sizes in parentheses....................................130
Figure 14. ?
15
N and ?
13
C relationships of loricariid consumers in (A) the Ventuari river,
Orinoco drainage, southern Venezuela; and (B) the Mara?on river, Amazon
drainage, northern Peru. Sample sizes in parentheses.................131
Chapter 4.
Figure 1. Major rivers and drainage basins of the Guiana Shield: 1. Orinoco River, 2.
Caroni River with Paragua River as its eastern tributary, 3. Caura River, 4.
Ventuari River, 5. Orinoco headwater rivers, from north to south: Padamo,
Matacuni, Ocamo, Orinoco, Mavaca, 6. Casiquiare Canal, 7. Siapa River, 8.
Negro River, 9. Demini River, 10. Branco River, 11. Uatuma River, 12.
Trombetas River, 13. Paru do Oeste River, 14. Paru River, 15. Jari River, 16.
Oyapok River, 17. Marone River, 18. Coppename River to the west and
Surinam River to the east, 19. Corentyne River, 20. Essequibo River, 21. Potaro
River, 22. Cuyuni River, 23. Uraricoera River, 24. Rupununi Savanna bordered
on the west by the Takutu River and on the east by the Rupununi........191
Figure 2. Schematic showing relationships among planation surfaces in Guyana, their
historical contiguity (dashed lines) and their modern remnants (solid lines).
xviii
Elevation of each surface relative to contemporary sea level in meters on the left
and feet on the right (from McConnell, 1968).......................192
Figure 3. The Prone-8: hypothesized areas of movement between basins of the Guiana
Shield. Area of some connections are approximate...................193
Figure 4. Phylogenetic hypothesis for the Chaetostoma-group (from Armbruster,
2008)................................................194
Figure 5. Null hypothesis of areal relationships among Guiana Shield fishes based only
upon hydrologic history. Basal node represents the historical continental divide
between eastern and western drainages at the Purus Arch. Terminal nodes
represent modern river drainages with text color indicating major modern
drainage basin: green = Amazon River, blue = Orinoco River. Three major
clades of modern river drainages (proto-Orinoco, proto-Berbice, and NE
Atlantic Coast) represent historical contiguity and regional affinities. Historical
geologic and hydrologic events at internal nodes labeled accordingly.....195
Figure 6. Three different biogeographic hypotheses based on differential use of
connections in the Prone-8. A. Hypothesis based on current drainage patterns.
B. Hypothesis if the Mazaruni-Caroni, Cuyuni- Caroni or Western Atlantic
Coastal Connections were used. C. Hypothesis considering the Casiquiare to be
Orinoco in origin.............................................196
Appendix II.
Figure1. Hypostominae: Ancistrini, Hypostomini, Pterygoplichthyini; and
Hypoptopomatinae: Hypoptopomatini.............................205
Figure 2. Loricariinae: Loricariini, Harttiini, Farlowellini.....................206
Figure 3. Hypostominae, Ancistrini: Ancistrus, Chaetostoma, Lasiancistrus, and
Dekeyseria.................................................207
Figure 4. Hypostominae, Ancistrini: Baryancistrus, Hemiancistrus, Hopliancistrus,
Lithoxus, Oligancistrus, and Pseudancistrus........................208
Figure 5. Hypostominae, Ancistrini: Peckoltia, Pseudacanthicus, Pseudolithoxus, and
New Genus 3...............................................209
Figure 6. Hypostominae, Ancistrini: Hypancistrus, Leporacanthicus, and
Scobinancistrus.............................................210
Figure 7. Loricariinae, Hartiini: Harttia, Lamontichthys, and Sturisoma; Farlowellini,
Farlowella; and Hypoptopomatinae, Hypoptopomatini, Hypoptopoma....211
xix
Figure 8. Loricariinae, Loricariini: Loricaria, Pseudoloricaria, Limatulichthys,
Rineloricaria, Loricariichthys, and Spatuloricaria...................212
Figure 9. Hypostominae, Ancistrini: Ancistrus macrophthalmus, A. temminckii, Ancistrus
sp. ?longjaw?, Ancistrus sp. ?shortjaw?, Ancistrus sp. ?wormline?, and Ancistrus
sp........................................................213
Figure 10. Hypostominae, Ancistrini: Baryancistrus beggini, B. demantoides,
Baryancistrus sp. ?B&W?, Baryancisrus sp. ?gold nugget?, and Baryancistrus
sp. ?green nugget?............................................214
Figure 11. Hypostominae, Ancistrini: Chaetostoma microps, C. cf. milesi, and
Chaetostoma sp..............................................215
Figure 12. Hypostominae, Ancistrini: Hemiancistrus guahiborum, H. sabaji, H.
snethlageae, H. subviridis, and Hemiancistrus sp. ?gold spot?...........216
Figure 13. Hypostominae, Ancistrini: Hypancistrus contradens, H. furunculus, H.
inspector, and H. lunaorum....................................217
Figure 14. Hypostominae, Ancistrini: Hopliancistrus tricornis, New Genus 3,
Lasiancistrus schomburgkii, L. tentaculatus, and Dekeyseria
scaphirhyncha..............................................218
Figure 15. Hypostominae, Ancistrini: Panaque albomaculatus, P. gnomus, P. nocturnus,
P. cf. maccus, Panaque n.sp. ?Mara?on?, and P. cf. nigrolineatus........219
Figure 16. Hypostominae, Ancistrini: Peckoltia braueri, P. cavatica, P. vermiculata, and
Peckoltia sp. ?big spot?........................................220
Figure 17. Hypostominae, Ancistrini: Pseudancistrus nigrescens, Pseudolithoxus
anthrax, Pseudancistrus pectegenitor, Pseudolithoxus dumus, Pseudancistrus
sidereus, and Pseudolithoxus tigris...............................221
Figure 18. Hypostominae, Ancistrini: Leporacanthicus cf. galaxias, L. triactis.
Oligancistrus puctatissimus, Lithoxus lithoides, Pseudacanthicus leopardus, and
Scobinancistrus sp...........................................222
Figure 19. Hypostominae, Hypostomini (Hypostomus cochliodon-group): Hypostomus
macushi, H. pyrineusi, H. taphorni, Hypostomus sp. ?dirty?, and Hypostomus sp.
?spots?.....................................................223
Figure 20. Hypostominae, Hypostomini: Hypostomus hemiurus, H. niceforoi, H. rhantos,
and Hypostomus sp...........................................224
xx
Figure 21. Hypostominae, Hypostomini (Hypostomus emarginatus-group): Hypostomus
emarginatus, H. cf. emarginatus, H. squalinus, and H. unicolor.........225
Figure 22. Loricariinae, Harttiini: Harttia platystoma, Harttia sp., Sturisoma monopelte,
Sturisoma nigrirostrum, and Lamontichthys filamentosus..............226
Figure 23. Loricariinae, Loricariini: Limatulichthys griseus, Pseudoloricaria sp.,
Rineloricaria fallax, R. lanceolata, R. stewarti, and Rineloricaria sp.....227
Figure 24. Loricariinae, Loricariini: Loricaria clavipinna, Loricaria sp. 1, Loricaria sp.,
and Loricariichthys brunneus...................................228
Figure 25. Loricariinae, Loricariini: Spatuloricara puganensis, Spatuloricaria sp. 1, and
Spatuloricaria sp. 2...........................................229
Figure 26. Loricariinae, Farlowellini: Farlowella acus, F. amazona; Hypoptopomatinae,
Hypoptopomatini, Hypoptopoma guianense; and Hypostominae,
Pterygoplichthyini, Pterygoplichthys gibbiceps......................230
1
CHAPTER 1. INTRODUCTION
1.1 GENERAL INTRODUCTION
?The most curious fact is the perfect gradation in the size of the beaks in the different
species of Geospiza, from one as large as that of a hawfinch to that of a chaffinch??
[Darwin, 1845: 379]
?As might be expected in the greatest river in the world, there is a corresponding
abundance and variety of fish. They supply the Indians with the greater part of their
animal food, and are at all times more plentiful, and easier to be obtained, than birds or
game from the forest. ? From the number of new fishes constantly found in every
locality and in every fisherman?s basket, we may estimate that at least five hundred
species exist in the Rio Negro and its tributary streams. The number in the whole valley
of the Amazon it is impossible to estimate with any approach to accuracy.?
[Wallace, 1889: 324-325]
Since variation in beak sizes across finches from the Galapagos Islands first
piqued Darwin?s curiosity, ranges of trophic morphologies across closely related groups
of organisms have intrigued evolutionary biologists and inspired myriad contributions to
the fields of ecology, functional morphology, and ontogenetics. Darwin?s empirical
2
observations of the birds that have come to bear his name began a research arc that
continues to this day (e.g. Lack, 1947, Grant, 1999, Schluter, 2000, Abzhanov et al.,
2006), and has been broadly successful in providing detailed explanations of major
ecological, functional, and morphogenetic mechanisms driving diversification in at least
the 14 species radiation of Geospizinae finches. My doctoral dissertation differs in
taxonomic scope, focusing instead on a diverse family of catfishes (Loricariidae), but like
Darwin?s early observations of finches, provides empirical observations of
biogeographical patterns and jaw morphological diversity that lay the foundation for
more mechanistically detailed explanations of an incredible New World trophic radiation.
Like Darwin, Alfred Russell Wallace was inspired by the diversity he encountered
in the Neotropics. He was especially taken with the diversity of fishes, over 200 species
of which he illustrated with care and precision despite the near constant torment of biting
insects (Wallace, 1889). Wallace?s estimates of fish diversity in the Negro River were
surprisingly accurate, coming close to the 450 species that have been described to date
from the Negro?s main channel (Goulding et al., 1988). And indeed, fish diversity across
the Neotropics exceeds that of any other vertebrate group. Modern estimates of total
Neotropical fish richness range between 5000 and 8000 species (Lundberg et al., 2000),
of which approximately 3600 (45?72%) are currently described (Reis et al., 2003). Of
described species, a disproportionate number (nearly 20%, or over 700 species) are
catfishes in the Neotropical-endemic family Loricariidae, which is also one of the fastest
growing of all fish families in terms of new species described, having increased by 35
species in 2008 alone.
3
Biogeographical patterns and taxonomic, morphological, and ecological diversity
of this group have been the foci of my research for the last seven years, including eight
expeditions to Brazil, Guyana, Peru, and Venezuela. My taxonomic research, frequently
conducted in collaboration with others, has resulted in description of 15 new species and
3 new genera (see Appendix I). The remainder of my research into loricariid
biogeography, and jaw morphological and trophic ecological diversity is the subject of
my dissertation.
1.2 INTRODUCTION TO THE LORICARIIDAE
Popularly known as plecos in the aquarium trade, where some rare and boldly
patterned species fetch upwards of several hundred dollars each, loricariid catfishes are
externally distinguished by having a ventral oral disk with jaws that are ventrally rotated
and highly modified to allow for efficient scraping of surfaces (Adriaens et al., 2009)
and/or winnowing of loosely aggregated food particles. They are further distinguished by
having bodies covered with ossified dermal plates and bristle-like odontodes
histologically similar to teeth (Sire & Huysseune, 1996). These distinctive morphological
characteristics earn loricariids the English common name of suckermouth armored
catfishes or, in Latin America, ?chupa piedras? (rock suckers). Other regional names
include cuchas (Colombia, Venezuela), corronchos (Colombia, Venezuela), carachamas
(Ecuador, Per?), and cascudos (Brazil).
Loricariidae is further diagnosed by several internal characters of the jaw,
including insertion of a novel division of the adductor mandibulae directly onto the
premaxilla, and absence of 3 plesiomorphic jaw linkages: 1) the interoperculo-
4
mandibular ligament, 2) the medial symphysis between left and right lower jaw rami, and
3) the tight ligamentous connection of premaxillae to the mesethmoid (Schaefer &
Lauder, 1986). The biomechanical result of these decouplings is a highly dynamic upper
jaw that can be adducted independently of left and right lower jaw rami, which are
themselves bilaterally independent (Adriaens et al., 2009). Sequential evolutionary loss
of these linkages has also been correlated with a trend toward increased minimum ranges
of jaw morphological diversity within the superfamilial clade Loricarioidea, of which
Loricariidae is the most derived family (Schaefer & Lauder, 1996). This correlation
supports the proposition (termed the ?decoupling hypothesis?, Schaefer & Lauder, 1986,
1996) that these linkages may function as evolutionary constraints in basal clades where
they remain present, whereas loss or decoupling of these linkages in derived clades
permits greater evolutionary freedom and, therefore, increased rates of morphological and
functional diversification.
The relative contribution of intrinsic morphological innovations versus extrinsic
ecological opportunity in shaping morphological and/or functional diversification is a
fundamental question in loricariid evolution, as it is in organismal evolution as a whole
(Lauder et al., 1980). Heterogeneity in rates of morphological change appears likely
within Loricariidae, but no studies have addressed how changes in the rate or pattern of
loricariid jaw morphological diversification might be correlated with extrinsic factors
such as trophic ecology or biogeography. Previous research correlating rates of
morphological diversification with losses of morphological constraints in Loricarioidea
(Schaefer & Lauder, 1996) focused on interfamilial comparisons, and was unable to
reveal intrafamilial patterns because of its inclusion of relatively few taxa (8 species) per
5
family. More comprehensive taxon sampling has been needed to both reveal intrafamilial
patterns and determine the validity of previous interfamilial generalizations based on
small sample sizes. Correlation of intrafamilial patterns of jaw morphological diversity
with ecological data would also provide support for alternative, albeit not necessarily
mutually exclusive, hypotheses of extrinsic mechanisms for morphological change.
1.3 INTRODUCTION TO DISSERTATION RESEARCH
Evidence suggesting correlative form and trophic ecological function has been
observed within a few sympatric, paraphyletic assemblages of loricariids (e.g. Delariva &
Agostinho, 2001), but these studies have been universally qualitative and without
standardized methods. This limitation has complicated broad synthesis and comparison
across diverse regions and taxa. Most descriptions of morphological variation within
Loricariidae are taxonomic or phylogenetic in focus. These studies provide an important
indication of the range of variation, but they typically only examine external aspects of
jaw morphology (e.g. tooth row angle and length, tooth number, size, and cusp shape).
These commonly examined characters are among the most highly variable morphological
traits below the rank of subfamily, which is where most variation in loricariid gross body
morphology occurs. Externally visible variation in jaw morphology at the rank of genus
and species has given these characters taxonomic significance as a means for diagnosing
lower level taxa throughout the family (e.g. Lujan et al., 2007, Armbruster, 2008,
Appendix I). Phylogenetic patterns of variation in just those characters examined for their
taxonomic value reveals the frequent convergence and repeated evolution of both long,
straight-jawed morphologies with many small teeth (e.g. Fig. 1A) and relatively short,
6
more-highly angled jaws with fewer teeth (e.g. Fig. 1B). Two notable examples of the
latter for which distinctive diets are documented are the wood-eating lineages Panaque
and the Hypostomus cochliodon-group, which both have relatively large adz-shaped teeth
set in short, highly angled tooth rows (Schaefer & Stewart, 1993, Armbruster, 2003).
Such examples support an important role for trophic ecology in allowing the
diversification of loricariid jaws and/or canalizing evolutionary outcomes of jaw
decoupling.
Standardizing morphometric methods so that they quantify homologous axes of
evolutionary divergence in shape is a central challenge to investigations of the
considerable variation in jaw morphologies across Loricariidae. Loricariid jaw elements
are highly 3-dimensional and curvilinear in outline, with few anatomically homologous
landmarks by which measurements can be standardized. In dynamic, functional systems
such as the loricariid jaw, defining landmarks or distances based on their functional role
(e.g. fulcrum, lever arm) provides an alternative to strictly anatomical homology. Such
functional homology has the potential advantage of directly linking quantification of
variation in form with variation in function, but the reliability of such a link depends on
accurate biomechanical models of system function. Recent investigations of loricariid
ontogeny and anatomy have provided detailed descriptions of jaw osteology and myology
of a few representative taxa (Geerinckx, 2006; Geerinckx et al., 2007), and a preliminary
study of kinematics has revealed gross aspects of loricariid jaw function (Adriaens et al.,
2009). The nomenclature of muscles involved with jaw function and the static
relationships between jaw muscles and bones are well understood, but the dynamic
motion and mechanics of loricariid jaws have yet to be described in precise detail.
7
Modeling of loricariid jaw mechanics is complicated by their high degree of kineticism
and three-dimensional freedom of movement. In contrast to the extensively investigated
planar 4-bar linkage by which the jaws of most teleosts function (Westneat, 2004),
loricariid jaws are characterized by the absence of key linkages, allowing a degree of
independent upper and lower jaw movement that is without precedent among fishes
(Adriaens et al., 2009). Studies of fish jaw mechanics rely heavily upon 2-dimensional
cineradiographic analyses, but these standard methods are insufficient for loricariids
because of their poor ability to resolve motion in 3 dimensions, and because dermal
armor on heads of most loricariids reduces the resolution of x-ray images of the jaws.
Precise descriptions of dietary variation present a second major challenge to
investigations of correlated jaw and trophic ecological diversification in Loricariidae.
Previous studies of fish community ecology in the Neotropics have revealed little trophic
segregation among loricariids, generally classifying them, instead, as undifferentiated
algivores and detritivores (see Table 1 for summary). Historically low-resolution
descriptions of loricariid diets are at least partially attributable to frequently small sample
sizes and a heavy dependence on gut content data, which are inherently low resolution
due to the amorphous nature of partially digested algal and detrital food particles.
Loricariid gut content data are further plagued by low accuracy caused by incredibly fast
gut passage rates (as little as 40 minutes; Hood et al., 2005), and the fact that annual dry
seasons, when most specimens are collected, are frequently times of suspended feeding
(Lowe-McConnell, 1964). A broad spectrum of trophic resources, from seeds to sponge
to wood and insects, have been hypothesized in loricariid taxonomic literature,
sometimes accompanied by hypotheses of specialization on such resources, but these
8
observations and conclusions are typically drawn from few, and sometimes single,
specimens. More robust support for a nutrient axis on which loricariids might be
segregating basal resources has recently been provided by Hood et al. (2005), who
demonstrated that some loricariids appear to maintain a stoichiometrically balanced diet
by selectively consuming the most phosphorus-rich fractions of the available epilithon.
Indeed, fine-scale feeding selectivity is a capability that has been broadly demonstrated in
both aquatic vertebrate (Bakare, 1970, Bowen, 1979, Ahlgren, 1996, Yossa & Araujo-
Lima, 1998) and invertebrate detritivores (Newell, 1965, B?rlocher & Kendrick, 1973,
Arsuffi & Suberkropp, 1985).
In this dissertation, I advance a research program intended to ultimately provide a
mechanistically detailed explanation of jaw morphological and trophic ecological
diversification in Loricariidae. I begin, in Chapter 2, by proposing several metrics by
which jaw morphological variation may be quantified in both a functionally homologous
and functionally informative manner across all Loricariidae and I use these metrics to
quantify jaw morphological variation across 2 loricariid subfamilies, 4 tribes, and 25
species. In Chapter 3, I use stable isotope data to examine the trophic structure of 19
sympatric loricariid assemblages distributed across 4 countries and 3 major Neotropical
river basins (Essequibo, Orinoco, Amazon). Here, I suggest that across this range,
loricariid assemblages are trophically structured, and that structuring forces increase with
species richness. Together, these chapters provide the first standardized comparison of
trophic morphological and ecological variation across a broad taxonomic and/or
geographic spectrum of loricariids. In Chapter 4, I discuss loricariid taxonomic
diversification in light of the geologic and hydrologic history of northern South America.
9
Loricariid diversity is distributed heterogeneously across the Neotropics, and
patterns within this diversity are likely to be closely linked to the geochemistry and
geologic history of Neotropical rivers. Any comprehensive explanation of loricariid
taxonomic and/or morphological diversification must take this landscape scale variation
into account. Loricariids are basal consumers and their trophic diversity and abundance
are likely at least partially linked to biogeochemistry, as illustrated by the Hood et al.
(2005) study of phosphorus limitation in Loricariidae.
10
1.4 REFERENCES
Abzhanov, A., W. P. Kuo, C. Hartmann, B. R. Grant, P. R. Grant, and C. J. Tabin. 2006.
The calmodulin pathway and evolution of elongated beak morphology in
Darwin's finches. Nature 442:563?567.
Adriaens, D., T. Geerinckx, J. Vlassenbroeck, L. Van Hoorebeke, and A. Herrel. 2009.
Extensive jaw mobility in suckermouth armored catfishes (Loricariidae): a
morphological and kinematic analysis of substrate scraping mode of feeding.
Physiological and Biochemical Zoology 82:51?62.
Ahlgren, M. O. 1996. Selective ingestion of detritus by a north temperate omnivorous
fish, the juvenile white sucker, Catostomus commersoni. Environmental Biology
of Fishes 46:375?381.
Alvim, M. C. C., and A. C. Peret. 2004. Food resources sustaining the fish fauna in a
section of the upper S?o Francisco River in Tr?s Marias, MG, Brazil. Brazilian
Journal of Biology 64:195?202.
Armbruster, J. W. 2002. Hypancistrus inspector, a new species of suckermouth armored
catfish (Loricariidae: Ancistrinae). Copeia 2002:86?92.
Armbruster, J. W. 2003. The species of the Hypostomus cochliodon group (Siluriformes:
Loricariidae). Zootaxa 249:1?60.
11
Armbruster, J. W. 2004. Phylogenetic relationships of the suckermouth armoured
catfishes (Loricariidae) with emphasis on the Hypostominae and the Ancistrinae.
Zoological Journal of the Linnean Society 141:1?80.
Armbruster, J. W. 2008. The genus Peckoltia with the description of two new species and
a reanalysis of the phylogeny of the genera of the Hypostominae (Siluriformes:
Loricariidae). Zootaxa 1822:1?76.
Arsuffi, T. L., and K. Suberkropp. 1985. Selective feeding by stream caddisfly
(Trichoptera) detritivores on leaves with fungal-colonized patches. Oikos
45:50?58.
Bakare, O. 1970. Bottom deposits as food of inland freshwater fish. Pages 65?85 in S. A.
Visser, editor. Kainji Lake Studies. University Press, Ibadan, Nigeria.
B?rlocher, F., and B. Kendrick. 1973. Fungi and food preferences of Gammarus
pseudolimnaeus. Archiv f?r Hydrobiologie 72:501?516.
Bowen, S. H. 1979. A nutritional constraint in detritivory by fishes: the stunted
population of Saratherodon mossambicus in Lake Sibaya, South Africa.
Ecological Monographs 1979:17?31.
Cardone, I. B., S. E. Lima-Junior, and R. Goitein. 2006. Diet and capture of Hypostomus
strigaticeps (Siluriformes, Loricariidae) in a small Brazilian stream: relationship
with limnological aspects. Brazilian Journal of Biology 66:25?33.
Darwin, C. 1845. Journal of Researches into the Natural History and Geology of the
Countries Visited During the Voyage of H.M.S. Beagle Round the World, Under
the Command of Captain FitzRoy, R.N. John Murray, London.
12
Delariva, R. L., and A. A. Agostinho. 2001. Relationships between morphology and diets
of six neotropical loricariids. Journal of Fish Biology 58:832?847.
Ferreira, K. M. 2007. Biology and ecomorphology of stream fishes from the rio Mogi-
Gua?u basin, Southeastern Brazil. Neotropical Ichthyology 5:311?326.
Forsberg, B. R., C. A. R. M. Araujo-Lima, L. A. Martinelli, R. L. Victoria, and J. A.
Bonassi. 1993. Autotrophic carbon sources for fish of the Central Amazon.
Ecology 74:643?652.
Fugi, R., N. S. Hahn, and A. A. Agostinho. 1996. Feeding styles of five species of
bottom-feeding fishes of the high Parana River. Environmental Biology of Fishes
46:297-307.
Geerinckx, T. 2006. Ontogeny and functional morphology of a highly specialized trophic
apparatus: a case study of neotropical suckermouth armoured catfishes
(Loricariidae, Siluriformes). Dissertation. University of Gent, Gent, Belgium.
Geerinckx, T, M. Brunain, A. Herrel, P. Aerts, and D. Adriaens. 2007. A head with a
suckermouth: a functional-morphological study of the head of the suckermouth
catfish Ancistrus cf. triradiatus (Loricariidae, Siluriformes). Belgian Journal of
Zoology 137:47?66.
Goulding, M., M. L. Carvalho, and E. G. Ferreira. 1988. Rio Negro ? Rich Life in Poor
Water. SPB Academic Press, The Hague.
Grant, P. R. 1999. The Ecology and Evolution of Darwin's Finches. Princeton University
Press, Princeton.
13
Hamilton, S. K., W. M. Lewis, Jr., and S. J. Sippel. 1992. Energy sources for aquatic
animals in the Orinoco River floodplain: evidence from stable isotopes. Oecologia
89:324-330.
Hood, J. M., M. J. Vanni, and A. S. Flecker. 2005. Nutrient recycling by two phosphorus
rich grazing catfish: the potential for phosphorus-limitation of fish growth.
Oecologia 146:247?257.
Jepsen, D. B., and K. O. Winemiller. 2002. Structure of tropical river food webs revealed
by stable isotope ratios. Oikos 96:46?55.
Kramer, D. L., and M. J. Bryant. 1995. Intestine length in the fishes of a tropical stream:
2. Relationships to diet ? the long and short of a convoluted issue. Environmental
Biology of Fishes 42:129-141.
Lack, D. 1947. Darwin's Finches. Cambridge University Press, Cambridge, UK.
Lauder, G.V., A.W. Crompton, C. Gans, J. Hanken, K.F. Liem, W.O. Maier, A. Meyer,
R. Presley, O.C. Rieppel, G. Roth, D. Schluter, and G.A. Zweers. 1980. Group
report: how are feeding systems integrated and how have evolutionary
innovations been introduced? Pages 97?115 in D. B. Wake and G. Roth, editors.
Complex Organismal Functions: Integration and Evolution in Vertebrates. John
Wiley and Sons Ltd., NY.
Lowe-McConnell, R. H. 1964. The fishes of the Rupununi savanna district of British
Guiana, South America: Part 1. Ecological groupings of fish species and effects of
the seasonal cycle on the fish. Journal of the Linnean Soceity (Zoology)
45:103?144.
14
Lujan, N. K., J. W. Armbruster, and M. H. Sabaj. 2007. Two new species of
Pseudancistrus from southern Venezuela (Siluriformes: Loricariidae).
Ichthyological Exploration of Freshwaters 18:163?174.
Lundberg, J. G., M. Kottelat, G. R. Smith, M. L. J. Stiassny, and A. C. Gill. 2000. So
many fishes, so little time: an overview of recent ichthyological discovery in
continental waters. Annals of the Missouri Botanical Garden 87:26?62.
Melo, C. E. de, F. de Arruda Machado, and V. Pinto-Silva. 2004. Feeding habits of fish
from a stream in the savanna of Central Brazil, Araguaia Basin. Neotropical
Ichthyology 2:37?44.
M?rigoux, S., and D. Ponton. 1998. Body shape, diet and ontogenetic diet shifts in young
fish of the Sinnamary River, French Guiana, South America. Journal of Fish
Biology 52:556?569.
M?rona, B. de, B. Hugueny, and F. L. Tejerina-Garro. 2008. Diet-morphology
relationship in a fish assemblage from a medium-sized river of French Guiana: the
effect of species taxonomic proximity. Aquatic Living Resources 21:171?184.
Newell, R. 1965. The role of detritus in the nutrition of two marine deposit feeders, the
prosobranch Hydrobia ulvae and the bivalve Macoma balthica. Proceedings of
the Zoological Society of London 144:25?45.
Nonogaki, H., J. A. Nelson, and W. P. Patterson. 2007. Dietary histories of herbivorous
loricariid catfishes: evidence from ?
13
C values of otoliths. Environmental Biology
of Fishes 78:13?21.
15
Peretti, D., and I. de Fatima Andrian. 2004. Trophic structure of fish assemblages in five
permanent lagoons of the high Paran? River floodplain, Brazil. Environmental
Biology of Fishes 71:95?103.
Power, M. E. 1981. The grazing ecology of armored catfish (Loricariidae) in a
Panamanian stream. PhD dissertation. University of Washington, Seattle.
Reis, R. E., S. O. Kullander, and C. J. Ferraris, Jr. 2003. Check List of the Freshwater
Fishes of South and Central America. EDIPUCRS, Porto Alegre.
Saul, W. G. 1975. An ecological study of fishes at a site in upper Amazonian Ecuador.
Proceedings of the Academy of Natural Sciences of Philadelphia 127:93?134.
Schaefer, S. A. and G. V. Lauder. 1986. Historical transformation of functional design:
evolutionary morphology of feeding mechanisms in loricarioid catfishes.
Systematic Zoology 35:489?508.
Schaefer, S. A., and G. V. Lauder. 1996. Testing historical hypotheses of morphological
change: biomechanical decoupling in loricarioid catfishes. Evolution
50:1661?1675.
Schaefer, S. A., and D. J. Stewart. 1993. Systematics of the Panaque dentex species
group (Siluriformes: Loricariidae), wood-eating armored catfishes from tropical
South America. Ichthyological Exploration of Freshwaters 4:309?342.
Schluter, D. 2000. The Ecology of Adaptive Radiation. Oxford University Press, Oxford,
UK.
Sire, J.-Y., and A. Huysseune. 1996. Structure and development of the odontodes in an
armoured catfish, Corydoras aeneus (Siluriformes, Callichthyidae). Acta
Zoologica, Stockholm 77:51?72.
16
Vaz, M. M., M. Petrere Jr., L. A. Martinelli, and A. A. Mozeto. 1999. The dietary regime
of detritivorous fish from the river Jacar? Pepira, Brazil. Fisheries Management
and Ecology 6:121?132.
Wallace, A. R. 1889. A Narrative of Travels on the Amazon and Rio Negro: With an
Account of the Native Tribes, and Observations on the Climate, Geology and
Natural History of Amazon Valley. Ward, Lock, London, UK.
Westneat, M. W. 2004. Evolution of levers and linkages in the feeding mechanisms of
fishes. Integrative and Comparative Biology 44:378?389.
Winemiller, K. O. 1990. Spatial and temporal variation in tropical fish trophic networks.
Ecological Monographs 60:331?367.
Yossa, M. I., and C. A. R. M. Araujo-Lima. 1998. Detritivory in two Amazonian fish
species. Journal of Fish Biology 52:1141?1153.
Zaret, T. M., and A. S. Rand. 1971. Competition in tropical stream fishes: support for the
competitive exclusion principle. Ecology 52:336?342.
17
Table 1. Summary of published descriptions of loricariid diets. Only dominant constituents are reported where gut contents were
reported as extended lists of taxa. Species identity updated to reflect current taxonomy. References: 1=Alvim and Peret, 2004;
2=Armbruster, 2002; 3=Armbruster, 2003; 4=Armbruster, 2004; 5=Cardone et al., 2006; 6=DeLariva and Agostinho, 2001;
7=Ferreira, 2007; 8=Forsberg et al., 1993; 9=Fugi et al., 1996; 10=Hamilton et al., 1992; 11=Hood et al., 2005; 12=Jepsen and
Winemiller, 2002; 13=Kramer and Bryant, 1995; 14=Melo et al., 2004; 15=M?rigoux and Ponton, 1998; 16=M?rona et al., 2008;
17=Nonogaki et al., 2007; 18=Peretti and Andrian, 2004; 19=Power, 1981, including several subsequent papers by this author;
20=Saul, 1975; 21=Schaefer and Stewart, 1993; 22=Vaz et al., 1999; 23=Winemiller, 1990; 24=Yossa and Araujo-Lima, 1998;
25=Zaret and Rand, 1971 (nr = not recorded).
? ? Species (corrected)ReferenceDrainageCountryMethodn Diet
Hypoptopomatinae
Hypoptopomatini
Hypoptopoma sp. 14AraguaiaBrazil gut contents3detritus
Hypoptopoma sp. 23ApureVenezuelagut contents52nr
Otocinclus cf. mariae20Napo Ecuadorgut contents2detritus
Otocinclus sp. 23ApureVenezuelagut contents686nr
Otothyrini
Hisonotus sp. 7 Mogi-Gua?uBrazil gut contents128organic matter, vascular plants, algae
Parotocinclus sp. 14AraguaiaBrazil gut contents117detritus, other, filamentous algae
Hypostominae
Ancistrini
Ancistrus 12CinarucoVenezuelaisotopesnrdetritus
Ancistrus chagresi13Rio FrijolesPanamagut contents10detritus
Ancistrus chagresi19Rio FrijolesPanamabehaviornralgae
17
18
Ancistrus hoplogenys15SinnamaryFrench Guianagut contents24vegetative detritus, substratum
Ancistrus hoplogenys20Napo Ecuadorgut contents4
sand; detritus of undetermined
composition
Ancistrus sp. 23ApureVenezuelagut contents62nr
Ancistrus triradiatus11ApureVenezuelagut contentsnralgae
Chaetostoma dermorhynchum20Napo Ecuadorgut contents3detritus
Chaetostoma fischeri13Rio FrijolesPanamagut contents6detritus
Chaetostoma milesi11ApureVenezuelagut contentsnralgae
Dekeyseria scaphiryncha8 Amazon basinBrazil isotopes1phytoplankton
Dekeyseria scaphiryncha12CinarucoVenezuelaisotopesnrdetritus
Hypancistrus inspector2 CasiquiareVenezuelagut contents~1detritus, algae, seeds
Lasiancistrus tentaculatus12PasimoniVenezuelaisotopesnrdetritus
Lasiancistrus tentaculatus12PasimoniVenezuelaisotopes5algae, detritus, plants
Megalancistrus aculeatus6 Parana RiverBrazil gut contents10
60% sponge, 18% organic detritus,
11% sediment, 8% Bryozoa
Panaque albomaculatus21Mara?onPeru gut contents~1wood
Panaque cf. nigrolineatus17Mogi-Gua?uBrazilotolith isotopes22wood
Panaque dentex21Mara?onPeru gut contents~1wood
Panaque gnomus21Mara?onPeru gut contents~1wood
Panaque maccus21ApureVenezuelagut contents~1wood
Panaque nigrolineatus20Napo Ecuadorgut contents1plant debris
Panaque nigrolineatus21ApureVenezuelagut contents~1wood
Panaque nocturnus21Mara?onPeru gut contents~1wood
Panaque purusiensis21Purus Peru/Brazilgut contents~1wood
Hypostomini
Hemiancistrus aspidolepis13Rio FrijolesPanamagut contents2detritus
Hemiancistrus aspidolepis19Rio FrijolesPanamabehaviornralgae
Hemiancistrus aspidolepis25Pedro MiguelPanamagut contents1algae
Hypostomus ancistroides7 Mogi-Gua?uBrazil gut contents13organic matter, vascular plants, algae
Hypostomus argus23ApureVenezuelagut contents379algae, detritus, microscopic animals
Hypostomus cochliodon3 ParaguayBrazil/Paraguaygut contents~1wood
18
19
Hypostomus emarginatus14AraguaiaBrazil gut contents3detritus
Hypostomus ericae14AraguaiaBrazil gut contents2other (wood?)
Hypostomus ericius3 Mara?onPeru gut contents~1wood
Hypostomus gymnorhynchus16Mahury R.French Guianagut contents14100% detritus
Hypostomus hemicochliodon3 Amazon basinPeru/Ecuador/Brazilgut contents~1wood, detritus
Hypostomus hondae3 Maracaibo/MagdalenaColombia/Venezuelagut contents~1wood
Hypostomus levis3 Beni Boliviagut contents~1wood
Hypostomus margaritifer6 Parana RiverBrazil gut contents10
49% Bryophyta, 21% organic detritus,
18% sediment, 10% Rhodophyta
Hypostomus microstomus6 Parana RiverBrazil gut contents10
60% sponge, 16% sediment, 11%
organic detritus, 3% plant detritus
Hypostomus oculeus3 Upper AmazonPeru/Ecuadorgut contents~1wood
Hypostomus pagei3 Aroa/YaracuyVenezuelagut contents~1wood
Hypostomus plecostomoides3 OrinocoColombia/Venezuelagut contents~1wood
Hypostomus plecostomoides23ApureVenezuelagut contents6nr
Hypostomus plecostomus8 Amazon basinBrazil isotopes1detritus
Hypostomus plecostomus10OrinocoVenezuelaisotopes1-10herbivore
Hypostomus plecostomus20Napo Ecuadorgut contents5detritus, sphaerid clams
Hypostomus pyrineusi3 Upper AmazonPeru/Ecuador/Brazilgut contents~1wood
Hypostomus regani6 Parana RiverBrazil gut contents10
71% detritus, 21% sediment, 7% plant
detritus
Hypostomus regani17Mogi-Gua?uBrazilotolith isotopes14algae
Hypostomus sculpodon3 Negro/OrinocoVenezuela/Brazilgut contents~1wood, detritus
Hypostomus sp. a1 1 S?o FranciscoBrazil gut contentsnrdetritus
Hypostomus sp. a2 1 S?o FranciscoBrazil gut contentsnrdetritus
Hypostomus sp. b1 14AraguaiaBrazil gut contents33detritus, filamentous algae, other
Hypostomus sp. b2 14AraguaiaBrazil gut contents6detritus
Hypostomus sp. b3 14AraguaiaBrazil gut contents17detritus
Hypostomus sp. b4 14AraguaiaBrazil gut contents12detritus
Hypostomus spp. (seven)22
Parana River
Brazilguts & isotopesnrdetritus
19
20
Hypostomus strigaticeps5 Curumbata?Brazil gut contents938
diatoms, fungi hyphae, chlorophytes,
cyanophytes
Hypostomus taphorni3 EssequiboVenezuela/Guyanagut contents~1wood
Hypostomus ternetzi6 Parana RiverBrazil gut contents10
42% Bryozoa, 32% sediment, 23%
organic detritus
Pterygoplichthini
Pterygoplichthys multiradiatus23ApureVenezuelagut contents479nr
Pterygoplichthys pardalis24Amazon basinBrazil gut contents29detritus
Pterygoplichthys radiatus8 Amazon basinBrazil isotopes5detritus
Rhinelepini
Rhinelepis aspera6 Parana RiverBrazil gut contents1096% detritus, 4% sediment
Loricariinae
Farlowellini
Farlowella kneri20Napo Ecuadorgut contents2detritus
Farlowella sp. 14AraguaiaBrazil gut contents20detritus, filamentous algae, insects
Farlowella sp. 23ApureVenezuelagut contents1nr
Harttiini
Sturisoma nigrirostrum14AraguaiaBrazil gut contents8detritus, other, leaves and flowers
Sturisoma sp. 23ApureVenezuelagut contents1nr
Loricariini
Crossoloricariasp. 4 not citednot citedgut contents~1seeds
Loricaria cataphracta16Mahury R.French Guianagut contents21
45.7% aquatic inverts., 28.1% detritus,
21% higher plants, 13.3% terr.
inverts.,  12.9% plankton
Loricaria cataphracta20Napo Ecuadorgut contents1detritus
Loricaria filamentosa20Napo Ecuadorgut contents3detritus
Loricaria sp. 4 not citednot citedgut contents~1seeds
Loricaria sp. 14AraguaiaBrazil gut contents35
fruits and seeds, detritus, other, leaves
and flowers, aquatic insects
Loricariichthys platymetopon8 Amazon basinBrazil isotopes3nr
Loricariichthys platymetopon9
Parana RiverBrazil
gut contents116organic detritus, chironomids
20
21
Loricariichthys platymetopon18Parana RiverBrazil gut contents49
detritus, sediment, organic detritus,
diatoms
Loricariichthys typus23ApureVenezuelagut contents501
large fractions of plant and animal
material
Pseudohemiodon cf. laticeps20Napo Ecuadorgut contents2insect debris, caddisfly larvae, snail
Pseudoloricaria sp. 14AraguaiaBrazil gut contents47
filamentous algae, detritus, aquatic
insects, leaves and flowers
Rineloricara lanceolata20Napo Ecuadorgut contents2aquatic plants, detritus
Rineloricaria caracasensis12Aguaro, ApureVenezuelaisotopesnrdetritus
Rineloricaria caracasensis12Aguaro, ApureVenezuelaisotopes4algae, detritus, plants
Rineloricaria caracasensis12ApureVenezuelaisotopes5algae, detritus, plants
Rineloricaria caracasensis23ApureVenezuelagut contents628nr
Rineloricaria uracantha13Rio FrijolesPanamagut contents8detritus, aquatic invertebrates
Rineloricaria uracantha19Rio FrijolesPanamabehaviornralgae
Spatuloricaria caquetae20Napo Ecuadorgut contents1detritus
? ? Spatuloricaria sp. 14AraguaiaBrazil gut contents12
fruits and seeds, aquatic insects,
detritus, arthropod, terrestrial insects
21
22
Figure 1. Representative diversity of loricariid upper and lower jaw morphologies: A.
Leporacanthicus (carnivore), B. Panaque (wood-eater), C. Chaetostoma (periphytivore).
23
CHAPTER 2. METHODOLOGICAL APPROACHES TO THE
QUANTIFICATION AND COMPARISON OF LOWER JAW MORPHO-
FUNCTIONAL DIVERSITY AMONG SUCKERMOUTH ARMORED
CATFISHES (SILURIFORMES, LORICARIIDAE)
2.1 ABSTRACT
I examined lower jaw morphological diversity across 25 species, 12 genera, 5
tribes, and 2 subfamilies of the Neotropical riverine catfish family Loricariidae
(Siluriformes, Loricarioidea) and I interpreted this diversity in light of the unique form
and function of the loricariid jaw. The loricariid jaw is highly derived and adapted to
surface scraping: the upper jaw is kinetic but independent of the lower jaw, which is
divided into bilaterally independent rami. Lower jaws are posteromedially deflected and
rotate predominantly around their long axis, although axes of rotation are variable and not
precisely determinable. Quantification, comparison, and functional interpretation of
loricariid jaw morphological diversity is complicated by highly 3-dimensional variation
in both shape and motion. Putatively functionally relevant morphological diversity across
lower jaws of the sampled taxa was quantified by measuring the area of insertion of the
combined adductor mandibulae muscle (AM
area
), and 5 linear distances, 3 of which were
analogous to traditional teleost in- and out-levers for jaw closure. Adductor mandibulae
volume was also measured from a subset of sampled taxa and found to correlate with
24
AM
area
 (p<0.001), supporting treatment of AM
area
 as a proxy for force into the loricariid
lower jaw. Principle component analysis of the morphometric dataset demonstrated a
broad spectrum of jaw morphologies and patterns consistent with both taxon at the ranks
of tribe, genus, and species, and broad dietary categories including lignivory,
herbivory/periphytivory, and granivory/nsectivory. Loricariid jaws exhibit a broad range
of mechanical advantages as traditionally defined, but inability to define axes of loricariid
lower jaw rotation hinders the functional interpretation of traditional mechanical
advantage metrics, and several additional aspects of function were considered. A key
metric hypothesized to be a reliable means of predicting loricariid jaw function from
morphology is the ratio AM
area
/tooth row length (TRL), interpreted as a measure of force
intensity, or the magnitude of force applied to substrate per unit of tooth row length.
Hypotheses of correspondance between changes in jaw geometry and changes in jaw
function were developed using a model of three-dimensional relationships between
measured parameters, potential forces, axes of rotation, and the jaw ramus. This model, in
conjunction with known kinematics of loricariid jaw function, illustrate theoretical
problems with the application of traditional metrics of mechanical advantage to
Loricariidae, and the need for novel metrics ? several of which are proposed and
discussed. Morphometric methods used in this study provide a standardized means of
quantitative comparison of shape across lower jaws of all Loricariidae, and the
hypothetical model of jaw mechanics developed herein generates specific predictions
about how select combinations of the parameters measured herein are linked to jaw
function.
25
2.2 INTRODUCTION
With approximately 30,000 species, ray-finned fishes (Actinopterygii) contribute
just over half of all vertebrate taxa. An evolutionary progression can be observed across
this radiation from strictly rigid lower jaw movement in basal lineages such as
Polypteriformes and Lepisosteiformes, to various manifestations among more derived
lineages of integrated upper and lower jaw articulation (Westneat, 2004). An early major
step in this progression resulted in the maxilla being decoupled from the neurocranium,
and ligamentously linked to the mandible ? a character first observed in the Amiiformes
but likely homologous with the character observed in most higher actinoptergians
(Lauder, 1979, Westneat, 2004). In contrast to a hypothesized single origin of the
maxillary-mandibular linkage, evolution of a kinetic premaxilla may have occurred up to
five times in Siluriformes, Cypriniformes, Gadiformes, Stylephoriformes, and Zeiformes
(Westneat, 2004, modified with findings of Miya et al., 2007). In the most derived of
these lineages (Zeiformes, sister to Percomorpha), the premaxilla is linked to the maxilla,
and movement of both remains linked to the lower jaw. Morphological and functional
diversity and kinematics of the integrated mandible-maxilla-premaxilla mechanism have
been described in detail for many members of the percomorph suborder Labroidea (e.g.
Cichlidae: Bouton et al., 1998; Albertson et al., 2005; Hulsey et al., 2006; Labridae:
Alfaro et al., 2004, 2005). Relatively little attention has been given to morphological
variation across, and function within, Loricariidae, a highly diverse siluriform exception
to this mechanism of upper and lower jaw movement.
Among catfishes (Siluriformes), the strictly Neotropical superfamily
Loricarioidea exhibits an evolutionary progression from immobile premaxillae that are
26
plesiomorphically fused to the mesethmoid, to highly mobile premaxillae loosely
suspended from a ventrally directed mesethmoid condyle (Fig. 1), and linked neither to
the maxilla nor to the mandible (Schaefer and Lauder, 1986). In the most derived
loricarioid family, Loricariidae, the premaxillae are tightly ligamentously connected to
each other along their anterior and mesial faces (Schaefer and Stewart, 1993; Geerinckx
et al., 2007), and are adducted as a single unit via direct insertion of the retractor
premaxillae muscle, a novel division of the adductor mandibulae (Geerinckx et al., 2007;
Adriaens et al., 2009). This muscle extends dorsomedially over the dentary and inserts in
a fossa along the posterior face of the premaxilla (retractor premaxilla fossa or rpf, Fig.
1B, C). In contrast to most other teleost groups with mobile premaxillae (e.g. Cichlidae,
Cyprinidae), loricariid premaxillae have only a modest ascending process, or none at all
(ap, Fig. 1A), and are largely reduced to hollow tooth cups. Emergent teeth are born in a
single row along the anteroventral margin of the tooth cup, and batteries of unerupted
teeth fill the cup?s interior (Geerinckx et al., 2007). In some taxa (e.g. Reganella; Rapp
Py-Daniel, 1997), the premaxilla may be highly reduced and poorly ossified with few to
no teeth, and in others (e.g. Chaetostoma, Fig. 2A), the premaxilla may be broadly
expanded laterally away from the mesethmoid condyle, and supports a single row of
>100 emergent teeth.
Loricariid lower jaw rami, consisting of the fused dentary and anguloarticular, are
morphologically and functionally more complex and interspecifically variable than the
upper jaw, and are more consistently well-ossified across all taxa (Schaefer, 1987; Rapp
Py-Daniel, 1997; Armbruster, 2004). Each lower jaw ramus in Loricariidae is
functionally independent of the upper jaw and its opposite ramus ? the medial mandibular
27
symphysis being absent. The long axis of the loricariid lower jaw ramus is largely
perpendicular to the longitudinal body axis and sagittal plane (Fig. 2A). Each ramus
extends medially from a prominent, lateral, anguloarticular condyle and the lateral
condyle articulates within a shallow fossa at the anteroventral corner of the quadrate (Fig.
2). As with the premaxilla, a single tooth row variable in numbers of teeth, length, and
angle, emerges from the posteroventeral margin of the dentary tooth cup, and batteries of
replacement teeth fill the cup?s interior (Geerinckx et al., 2007). A cartilage plug extends
rostrally from the hyoid arch, just behind the lower jaw, into the posteromedial gap
between left and right dentaries (Schaefer and Lauder, 1986); and although this plug has
no direct attachment to either jaw ramus, it has been hypothesized to provide medial
support for the lower jaw during adduction (Geerinckx et al., 2007; Adriaens et al.,
2009). The variable angle and length of the dentary tooth row typically matches that of
the premaxilla (Fig. 2A, C), although divergent premaxillary and dentary tooth row
morphologies are diagnostic of a few taxa (e.g. Hypancistrus, Leporacanthicus,
Spectracanthicus; Fig. 2B). Morphology of the region between the dentary tooth cup and
the anguloarticular condyle, where the main mass of the adductor mandibulae inserts
dorsally, is also highly variable ? ranging from a largely featureless column in some
Loricariinae (e.g. Reganella, Rineloricaria; Fig. 3O; Rapp Py-Daniel, 1997), to a
flattened, broadly expanded flange and/or coronoid arch in a few Loricariinae (e.g.
Lamontichthys, Spatuloricaria; Fig. 3N, P) and most Hypostominae (e.g. Panaque n. sp.
?Mara?on?; Fig. 3F). The predominant axis of lower jaw rotation likely passes through
this region and the observed morphological variation would alter geometric relationships
28
between the axis and regions of force-in and -out, thereby contributing to variation in the
mechanical advantage produced during rotation of the jaw ramus.
Functional ramifications of the highly derived, deintegrated, and ventrally rotated
jaw morphology described above include the capacity for up to 180? of jaw abduction, so
that upper and lower tooth rows can be synchronously occluded by, and adducted across,
flat surfaces proximal to the mouth (Figs. 1, 2; Adriaens et al., 2009). Premaxillae are
adducted from anterior to posterior with rostrocaudal translation and rotation in the
sagittal plane, whereas each lower jaw ramus is adducted anteromedially with rotation
possible in at least the horizontal and sagittal planes (Adriaens et al., 2009). Loricariid
jaws function within a fleshy, papillose labial disk (Fig. 4) that likely serves several
functions, including chemosensory food detection (Ono, 1990), substrate attachment
(Geerinckx et al., 2007a), dislodgement of food particles (Geerinckx et al., 2007b), and
funneling of dislodged food particles into the buccal cavity (Arens, 1994). The maxillary
barbel is integrated with the labial disk, and inserts on the maxilla, which can be adducted
independently of all other jaw elements by direct insertion of the levator tentaculi muscle
(another novel aspect of loricariid jaw anatomy; Geerinckx et al., 2007). Although
anatomically independent of tooth-bearing jaw rami, adduction of the maxilla/maxillary
barbel system may still interact with jaw function by enhancing the labial disk?s capacity
for surface attachment (Geerinckx et al., 2007). Surface attachment increases the normal
force and, therefore, the frictional force of tooth rows against substrates during jaw
adduction, likely similar to that of leeches, lampreys, larval anurans (Wassersug and
Yamashita, 2001; Larson and Reilly, 2003), and several teleosts (e.g. Benjamin, 1986;
Roberts, 1989). Complicated force and flow dynamics of the integrated oral disk system
29
are poorly understood, but appear to allow simultaneous surface attachment, surface
scraping, ingestion of food particles, and respiration (Geerinckx et al., 2007).
Loricariidae and 5 other strictly Neotropical catfish families comprise the
Loricarioidea, a clade sister to all other catfishes (Sullivan et al., 2006). Loricariidae is
the most derived of these families, and several evolutionary stages precursor to the highly
derived loricariid jaw can be observed among loricarioid lineages. Sequential loss of
ligamentous linkages between jaw elements is particularly apparent. In addition to an
increased range of movement, sequential decoupling of jaw elements within the
loricarioid lineage has been hypothesized to have contributed to a progressive increase in
rates of functional and morphological diversification (Schaefer and Lauder, 1986). This
decoupling hypothesis states that plesiomorphic linkages function as evolutionary
constraints in basal clades where they remain present, and that greater evolutionary
freedom is allowed by the absence of these linkages in derived clades (Vermeij, 1973,
1974; Schaefer and Lauder, 1986). Previous quantification of morphological diversity
across small subsamples of the 4 most species-rich loricarioid lineages provides support
for the decoupling hypothesis by demonstrating a trend toward increased morphological
diversity correlated with extent of jaw decoupling (Schaefer and Lauder, 1996); however,
it is likely that Schaefer and Lauder (1996) greatly underestimated intrafamilial diversity
by examining only eight species from each of the four most species-rich loricarioid
families. No other study has quantitatively examined jaw morphological diversity in the
Loricarioidea.
The relative contribution of intrinsic morphological innovations, such as
decoupling, versus extrinsic ecological opportunity in controlling morphological and/or
30
functional diversification is a fundamental question in studies of both loricariid evolution
and organismal evolution as a whole (Lauder et al., 1980; Hughes and Eastwood, 2006).
As the most highly decoupled and derived of the loricarioid families, Loricariidae would
be predicted to have jaws that are the most plastic in their response to natural selection
and, therefore, most highly adaptive in their correlation of form with function. The
loricariid jaw has diversified across a species-rich radiation whose members are
ubiquitous in tropical Central and South American rivers and streams. Over 700 species
are currently described, and many species remain undescribed. Fine-scale patterns of
ecological and jaw morphological diversity across this radiation remain largely
undescribed. External features of loricariid trophic morphology (e.g. tooth row angle and
length, and tooth number, size, and cusp shape) have been associated with trophic
ecology, but only among small, paraphyletic groups, and only in a qualitative or non-
standardized manner (e.g. Delariva and Agostinho, 2001; Fugi et al., 2001; M?rona et al.,
2008). Most descriptions of morphological variation within Loricariidae have had only
taxonomic or phylogenetic foci; and although taxonomically significant external jaw
characters provide an indication of the wide potential range of jaw variation, these studies
have thus far ignored the considerable internal morphological variation described herein
(Fig. 3). In addition to holistic quantification of jaw morphological variation, there is a
need for biomechanical hypotheses linking morphological variation with variation in jaw
function and trophic ecology.
Defining homology across a wide range of taxa and highly variable morphologies
presents a central challenge to the standardized quantification of jaw morphological
variation in Loricariidae. Loricariid jaw elements are highly variable, 3-dimensional, and
31
curvilinear in outline, with few discrete, anatomically homologous landmarks by which
measurements can be standardized (e.g., Type l landmarks, or discrete juxtapositions of
tissues; sensu Bookstein, 1991). In dynamic systems such as the loricariid jaw, the
functional role of a given landmark or dimension (e.g., fulcrum, lever arm) can also be
used to standardize measurements. Such an approach has the added advantage of directly
linking variation in form with variation in function, thereby facilitating ecological
interpretations of morphological diversity. Tools for ecological inference from loricariid
jaw morphology would be particularly valuable given the scarcity of such data; however,
the reliability of any ecomorphological inference depends on accurate biomechanical
models of system function, which are also lacking for the loricariid jaw. Recent
investigations of loricariid ontogeny and anatomy have provided detailed descriptions of
jaw osteology and myology for only a few representative taxa (Geerinckx, 2006;
Geerinckx et al., 2007) and a preliminary study of kinematics has described gross aspects
of loricariid jaw motion (Adriaens et al., 2009). Dynamic function of the loricariid jaw
has yet to be described in detail, but it seems that the absence of key linkages allows 3-
dimensional freedom of independent upper and lower jaw movement that is unique
among fishes (Adriaens et al., 2009).
In most gnathostomes, the lower jaw is a single, rigid unit, allowing all parts of
the jaw to be modeled as if they exist within the same 2-dimensional plane of rotation,
and allowing jaw morphology to be matched to jaw function along a single gradient of
mechanical optimization, either for speed (short input lever/long output lever) or force
(long input lever/short output lever). The unique structure and function of the loricariid
jaw challenges this traditional approach to jaw function, and invites consideration of
32
several additional aspects of function. A central challenge to modeling loricariid lower
jaw function in a manner consistent with traditional investigations is the inability to
define a dominant axis of rotation and measure its distance from regions where force is
either entering or leaving the jaw ? the definition of respective input and output lever
arms. Adriaens et al. (2009) demonstrated quantitatively what can also be easily observed
qualitatively by watching a loricariid scrape the inside glass walls of an aquarium:
rotation of the lower jaws is predominantly in the sagittal plane, but also in the horizontal
plane, sweeping anteriorly around the anguloarticular-quadrate joint. My goals in this
study are (1) to provide empirical, quantitative data describing variation in several
aspects of loricariid jaw morphology, and (2) to make explicit predictions about how
variation in these dimensions contribute to variation in function.
2.3 METHODS
2.3.1 Specimen collection, preparation, and imaging
I examined 25 species and 379 individuals from 5 tribes and 2 subfamilies within
Loricariidae (two to 83 individuals per species; Table 1). All specimens were collected by
seining and electroshocking middle reaches of the Mara?on River, a tributary of the
upper Amazon Basin in northern Peru in August 2006. New Genus 3 refers to an
undescribed genus and species of Ancistrini (Hypostominae) whose phylogenetic
relationships to other loricariid genera are discussed by Armbruster (2008). All
dissections were made from voucher specimens cataloged at the Auburn University
Museum Fish Collection.
33
Quantitative morphological comparison of the upper jaw across Loricariidae is
complicated by extreme diminution of the premaxilla in some taxa (e.g. Reganella,
Pseudohemiodon; Rapp Py-Daniel, 1997) and by a scarcity of landmarks on the
premaxilla that might be considered homologous across all taxa. Additionally, premaxilla
morphological variation is difficult to relate to function using only disarticulated osteal
elements because a highly variable (Armbruster, 2004) cartilage meniscus separates the
premaxilla from its rotational axis in the mesethmoid condyle. Therefore, I restricted this
study to the right lower jaw ramus (fused dentary, mentomeckelium, and anguloarticular;
Geerinckx et al., 2007), which I dissected from each specimen, then cleared and stained
individually (Fig. 3). Loricariid lower jaw elements are small, ranging in this study from
three to 15 mm in greatest linear dimension. To facilitate manipulation and imaging, I
removed all soft tissue following clearing and staining and allowed the osteal elements to
air dry. I then photographed each lower jaw element in at least 2 of the 3 perspectives
illustrated in Fig. 5 using a Nikon Coolpix 990 digital camera mounted to a Leica MZ6
stereomicroscope.
Interspecific torsional variation in the relative position and orientation of distal
versus proximal jaw regions eliminates the possibility of any one perspective being
homologous relative to all jaw regions across all loricariid taxa. Therefore, I arbitrarily
standardized jaw orientations by ensuring that the broad anguloarticular-dentary coronoid
flange was orthogonal (parallel or perpendicular) to the field of view in each image.
Horizontal orientations in ventral (Fig. 5A, D) and dorsal (Fig. 5C, F) perspectives were
defined as having the flange parallel with the stage; vertical orientations (Fig. 5B, E)
were defined as having the flange perpendicular to the stage. These positions were
34
advantageous by being stable resting positions for many specimens, and by presenting at
least one, but usually two, parameters both parallel with and proximal to the stage, further
standardizing measurements. I used a small piece of wire mesh as a support should a jaw
element not rest naturally in the defined orientations.
2.3.2 Morphometrics
I measured 5 linear distances and 1 area (hereafter referred to as parameters) from
digital images of each lower jaw ramus (Fig. 5). Parameters were selected based on the
criteria that they quantify distinct components of interspecific morphological variation
observable in gross visual surveys of loricariid lower jaw rami (e.g., Fig. 3), and that they
be measurable in a standardized manner across all Loricariidae. All parameters were
measured from digital images using tpsDIG2 software (Rohlf, 2008, v. 2.12), and were
individually standardized to a millimeter scale attached to the microscope stage and
visible in each frame. Three linear distance parameters were anatomically analogous with
lever arms for lower jaw closure in the majority of actinopterygians (e.g. Westneat, 2004)
and were named accordingly. Two of these linear distances were measured from a
midpoint along the surface of the anguloarticular condyle (AAC) to respective distalmost
(distal out-lever, Out
dist
; Fig. 5B, E) and proximalmost (proximal out-lever, Out
prox
; Fig.
5B, D) tooth insertions, and the third was measured from a midpoint along the surface of
the AAC to the center of the adductor mandibulae (AM) area of insertion (in-lever, In;
Fig. 5C, F). To accommodate jaw samples that lost teeth during clearing, staining, and
soft tissue removal, and to standardize morphometric methods across loricariid taxa with
a broad diversity of dentition, I defined tooth landmarks as the tooth insertion rather than
35
the tooth cusp. I also measured the following 3 parameters (2 linear distances, 1 area) not
typically examined in previous investigations of fish jaw morphology and function: (1)
linear distance from proximalmost to distalmost tooth insertions (tooth row length, TRL;
Fig. 5A, D), (2) area across which the combined adductor mandibulae (AM; A1-OST and
A3? divisions; Geerinckx, 2006) muscle inserts (AM
area
; Fig. 5C, F), and (3) the linear
distance perpendicular to Out
dist
, from Out
dist
 to the maximum excursion of the coronoid
arch (H1; Fig. 6). I examined the cumulative extent to which measured parameters
differentiated loricariid lower jaw morphologies at various taxonomic ranks by
conducting a principle component analysis in JMP statistical software (SAS Institute, v.
5.0.1a).
2.3.3 Jaw mechanics
Little is known about loricariid lower-jaw mechanics. I synthesized traditional
studies of fish jaw levers (e.g. Westneat, 2004) with the single published study of
loricariid jaw kinematics (Adriaens et al., 2009), personal qualitative observations of
loricariid jaw function in the field and in aquaria, and detailed studies of loricariid jaw
anatomy (e.g. Geerinckx et al., 2007) to develop several hypotheses of how interspecific
variation in the above morphometric parameters might relate to interspecific variation in
loricariid lower jaw function. Aspects of the putative functional variation predicted by
these hypotheses are presented as a series of 6 ratios and 1 angle (hereafter referred to as
metrics). Ratios and angles serve the dual purposes of adding functional significance
(e.g., lever-arm ratios) and compensating for allometric variation. The first ratio is a
novel metric hypothesized to predict force intensity and will be discussed separately from
36
the remaining 5 ratios (2 traditional, 3 novel) and the angle, which describe aspects of
force geometry (i.e. spatial relationships between areas of force-in, force-out, and axes of
rotation).
2.3.4 Jaw mechanics: force intensity
The combined adductor mandibulae (AM) of loricariids is approximately
columnar is gross dimensions and inserts broadly on the dorsal surface of the
anguloarticular-dentary coronoid flange (Fig. 5), which is hypertrophied in loricariids
known to be durophages (e.g., Fig. 3F). Given this morphology, I hypothesized that AM
insertion area (AM
area
) correlates with AM volume (AM
vol
) and variation in
anguloarticular-dentary coronoid flange area might thereby provide an indicator of
interspecific variation in magnitude of force into the lower jaw system. I examined the
relationship between AM
area
 and AM
vol
 after all jaw osteological data had been collected
by selecting a subset of approximately six specimens from each of 12 species (Table 1)
spanning subfamilies Hypostominae and Loricariinae, and dissecting the AM from the
left side of each specimen. Muscle volume was measured to the nearest 0.01 mL via
displacement of water in a 5 mL graduated cylinder. In order to examine whether
interspecific variation in AM
area
 is correlated with variation in AM
vol
 after collecting
these data from separate specimens, I standardized AM
vol
 and AM
area
 to standard body
lengths (SL) of the respective specimens from which they were dissected, and regressed
mean values of AM
vol
/SL for each species against mean values of AM
area
/SL for each
species (Fig. 7).
37
Loricariids exhibit considerable variation in tooth number, shaft morphology,
length, rigidity, and cusp dimensions (e.g. Fig. 4; Delariva and Agostinho, 2001; Muller
and Weber, 1992; Geerinckx et al., 2007), but all loricariids have the ability to
synchronously apply their entire dentary tooth row to substrates, and most loricariids
have teeth and tooth cusps distributed in a single contiguous row from proximalmost to
distalmost tooth. As a single standardized measure of variation in dentition, I measured
tooth row length (TRL; distance from proximalmost to distalmost tooth) and
hypothesized that interspecific variation in TRL correlates with interspecific variation in
the area across which force through the lower jaw is distributed. I related TRL to AM
area
in the metric AM
area
/TRL, which I hypothesize predicts interspecific variation in the
magnitude of force applied to substrates per unit of tooth row length, assuming no
variation in mechanical advantage.
2.3.5 Jaw mechanics: force geometry
Mechanical advantage is the factor by which a rotating system, or lever,
multiplies force; it can be predicted by dividing the system?s in-lever (distance from
region of force-in to axis of rotation) by its out-lever (distance from axis of rotation to
region of force out), with high values indicating optimization for force and low values
indicating optimization for speed. In-levers for fish jaw closure are commonly defined as
distances from anguloarticular condyle (AAC) to area of adductor mandibulae (AM)
insertion, and out-levers as distances from AAC to tooth cusps (e.g. Westneat, 2004). I
calculated 2 mechanical advantage metrics using similar parameters (In/Out
dist
 and
38
In/Out
prox
), but relevance of these metrics to Loricariidae is reduced by the highly derived
structure and function of loricariid jaws.
In contrast to most fish mandibles, which close by rotating around the AAC-
quadrate joint in a manner similar to flexion of human forearms around the ulna-humerus
joint, the loricariid lower jaw rotates around its long axis, in a manner more similar to
human forearm torsion, or rotation of the radius around the ulna. Although Adriaens et al.
(2009) described gross jaw kinematics in the single loricariid species Pterygoplichthys
disjunctivus (Hypostominae, Pterygoplichthyini), intra- and interspecific variation in
loricariid lower jaw kinematics are poorly understood. Given the high degree of freedom
of loricariid lower jaws, it seems likely that loricariid lower jaw kinematics are inherently
highly variable and largely dependent on variation in substrate friction coefficients. To
compensate for the absence of kinematic data specific to loricariid species in this study
and the kinematic flexibility likely inherent to the loricariid jaw system, I related 1-
dimensional parameters measured herein to each other, to potential axes of rotation, and
back to the loricariid lower jaw with a 3-dimensional rotating cone model of loricariid
lower jaw function (Fig. 8). This jaw model, in conjunction with the gross kinematic
observations of Adriaens et al. (2009), provided a means for interpreting the probable
effects of interspecific jaw morphological variation on variation in jaw function, and
provided the foundation upon which the following 4 novel metrics (3 ratios and 1 angle)
are discussed:
H1/TRL: treated as a combined measure of force intensity and novel calculation
of mechanical advantage.
39
Out
dist
/H1: treated as a measure of the predominant plane of torque through the
jaw ramus.
TR
angle
 (calculated trigonometrically by treating Out
dist
, Out
prox
, and TRL as three
sides of a scalene triangle): treated as a measure of torque differential
across the tooth row, correlated with Out
dist
-Out
prox
/In.
Out
dist
-Out
prox
/In: treated as a measure of torque differential across the tooth row.
A central component of the rotating cone model is the output plane, defined by a
triangle with sides formed by Out
prox
, Out
dist
, and TRL (shaded gray, Fig. 8). The
posterolateralmost vertex of the triangle is the anguloarticular condyle and the other
vertices are the distalmost and proximalmost teeth (Fig. 8). Extending dorsomedially
away from the anguloarticular condyle, oblique with respect to the output plane, is the in-
lever arm (In, Fig. 8). Rotation around the long axis of the jaw (Axis 1) is illustrated as a
rotating cone with the anguloarticular condyle at its apex, and excursion of the distalmost
tooth tracing an arc around the base of the cone. A portion of the morphological variation
observed across loricariid jaws is represented by cones of 2 different dimensions: a
relatively long cone with a small angle between in-lever and output plane and a relatively
shorter arc describing excursion of the distalmost tooth (Cone 1, e.g. Chaetostoma cf.
milesi, Fig. 8), and a shorter cone with a greater angle between in-lever and output plane
and a longer arc describing excursion of the distalmost tooth (Cone 2, e.g. Panaque
nigrolineatus, Fig. 8).
Variation in TRL and TR
angle
 (?, Fig. 8) within the output plane are also
illustrated, with Cone 1 having a relatively long TRL and a low TR
angle
, and Cone 2
having a shorter TRL and a greater TR
angle
. Angle between left and right dentary tooth-
40
rows is a variable and important taxonomic character diagnostic for several genera and
species. The genera Peckoltia and Hemiancistrus, for example, are currently
differentiated based solely on whether angle between left and right dentary tooth-rows is
<90? (Peckoltia) or >100? (Hemiancistrus; Armbruster, 2008); however, angle between
left and right dentary tooth-rows is variable within an individual because of the
independent articulation of left and right lower jaw rami, and is difficult to quantify
without reference to a point or line that is fixed relative to the tooth-row. I calculated
angle of the tooth-row (TR
angle
, ? in Fig. 8) with respect to the distal out-lever arm
trigonometrically, by treating Out
dist
, Out
prox
, and TRL as 3 sides of a scalene triangle
(shaded gray in Fig. 8).
Interpretation and discussion of the jaw model and novel metrics are largely
contingent upon four assumptions: (1) that rotation is predominantly in the sagittal plane
(around Axis 1; Fig. 8), (2) that rotation in the horizontal plane (around Axis 2; Fig. 8) is
secondary, and (3) that rotation in the transverse plane (Axis 3; Fig. 8) is negligible. All
of these assumptions are supported by kinematic data from the loricariid species
Pterygoplichthys disjunctivus (Adriaens et al., 2009). The final assumption is (4) that the
predominant rotational axis (Axis 1) crosses through a midpoint along the long axis of the
ramus, approximately equidistant from the input lever arm and the output plane. The
importance of this assumption is discussed further under Results and Discussion: Force
Geometry-Novel Metrics.
2.4 RESULTS AND DISCUSSION
2.4.1 Gross morphological differentiation
41
Principle component (PC) analysis of the loricariid jaw parameters measured in
this study (Figs. 9, 10) revealed morphological segregation of the jaws of most taxa at the
ranks of genus and species, and jaw morphological segregation corresponding with the
following three broad trophic guilds: lignivores (wood-eating),
periphytivores/detritivores, and insectivores/granivores. The subfamilies Hypostominae
(Ancistrini, Hypostomini; Fig. 9A, B) and Loricariinae (Loricariini, Farlowellini,
Harttiini; Fig. 9A, B) broadly overlap in jaw morphologies, but jaw morphologies
become differentiable at the rank of tribe. Loricariini are largely differentiable from
Harttiini along PC2 (Fig. 9A) and are differentiable from all other tribes along PC3 and
PC4 (Fig. 9B), which also allow differentiate Harttiini from Farlowellini. Ancistrini
exhibits the broadest range in jaw morphologies across all PCs, but is also the tribe with
the greatest taxonomic diversity and largest sample sizes represented in the dataset.
Jaw morphological diversity and differentiation is greatest among genera (Fig.
9C, D), which is likely at least partially attributable to the frequent reliance upon jaw
characters for the diagnosis of loricariid genera. Jaw morphologies of most of the 6
Ancistrini genera in the dataset can be distinguished along PC2, although Peckoltia
occurs within the distribution of Panaque, and Lasiancistrus is overlapped slightly by
both Panaque and Ancistrus. All Ancistrini genera exhibit broad overlap along PC3 and
PC4 (not shown).
Jaw morphological patterns at the rank of species vary among genera. Of the three
genera with the greatest number of species in the dataset, Chaetostoma exhibits broad
overlap among several species (Fig. 10A), whereas most Panaque species have
distinctive jaw morphologies described by PC3 and PC4 (Fig. 10B) and most
42
Hypostomus species have distinctive jaw morphologies described by PC2 (Fig. 10C).
Within Chaetostoma, jaw morphologies of C. microps, C. cf. milesi, and Chaetostoma sp.
1, are largely differentiable from each other along PC2, as is Chaetostoma sp. 4 from
Chaetostoma sp. 1 (Fig. 10C). Jaw morphologies of Chaetostoma sp. 2 and Chaetostoma
sp. 3 broadly overlap both each other and the other Chaetostoma species along PC2, and
jaw morphologies of Chaetostoma sp. 4 broadly overlap those of both C. microps and C.
cf. milesi (Fig. 10C). Jaw morphologies of all Chaetostoma exhibit broad overlap along
PC3 and PC4 (not shown).
Distributions of jaw morphologies across species in the genera Panaque and
Hypostomus deserve special attention because of the specialization within some species
in these genera on diets consisting largely of wood (Schaefer and Stewart, 1993; Nelson
et al., 1999; Armbruster, 2003; Nonogaki et al., 2007; German, 2008). Submerged, solid
wood in the form of tree boles and branches (coarse woody debris or CWD) is a food
resource consumed by few other aquatic metazoans. Besides loricariids, only the
caddisfly Heteroplectron californicum (Calamoceratidae, Anderson et al., 1978), the
beetle Lara avara (Elmidae, Anderson et al., 1978), and the beaver (Mammalia, Castor
spp.) are known to be specialist consumers of CWD in stream systems, and all of these
are restricted to temperate latitudes. Recent studies of the nutritional physiology of
lignivorous loricariids indicate that these trophic specialists have an unspecialized
alimentary canal and that, like many aquatic insect detritivores (Cummins, 1973), the
energy and nutritive value of the wood they consume is largely derived from associated
fungi and free monosaccharides and amino acids released during fungal predigestion of
the wood (German, 2008).
43
Loricariid specializations for consumption of wood, therefore, appear to be
restricted to the jaw system and, indeed, all lignivorous loricariids have specialized
spoon-shaped teeth (Fig. 3D, E, F, G, K; Armbruster, 2003), and occupy a region of jaw
morphospace along the PC2 axis also occupied by only one other non-lignivorous species
(Peckoltia bachi; Figs 9C, 10C). Peckoltia bachi, which is phylogenetically close to
Panaque (Armbruster, 2008), lacks the spoon-shaped teeth of other lignivores, but
overlaps Panaque along PC2 (Fig. 9C), PC3, and PC4 (not shown). Diet of Peckoltia
bachi has not been described. Within Hypostomus, only the single lignivorous species H.
pyrineusi (Fig. 10C) has a jaw morphology overlapping that of the largely lignivorous
genus Panaque along PC2 (cf. Fig. 9C). Hypostomus pyrineusi is a member of the largely
wood-eating Hypostomus cochliodon-group defined by Armbruster (2003). Segregation
of Panaque species along PC3 and PC4 (Fig. 10B) suggests the possibility that loricariids
are further partitioning trophic resources within the highly specialized lignivorous niche.
Additional support for trophic segregation within sympatric assemblages of lignivorous
Panaque species is provided by partially differentiable patterns of carbon and nitrogen
consumer stable isotope data (Appendix II). The most distinctive of the Panaque species
jaw morphologies is that of Panaque n. sp. ?Mara?on?, which is segregated along PC3
from most other taxa in the dataset, including all other Panaque species.
2.4.2 Jaw mechanics: force intensity
Differentiation of Panaque n. sp. ?Mara?on? along the PC3 axis is derived mostly
from its increased area of adductor mandibulae (AM) insertion (Fig. 10B), which
correlates with AM hypertrophy (Fig. 7), and is consistent with its membership of the
44
durophagous lignivorous guild. Additional differentiation of lignivorous taxa along PC2
is driven mostly by their relatively short tooth row lengths (TRL; Figs. 9C, 10C), which
can be interpreted functionally as a means of concentrating jaw forces onto smaller areas
of substrate. As a group, lignivorous taxa have jaw morphologies consistent with
increases in force generation, and decreases in force distribution ? aspects of loricariid
jaw function that are related in the single metric AM
area
/TRL. The aspect of jaw function
that this metric is hypothesized to predict can be referred to as force intensity, or the
degree of force per unit of tooth row length. This metric ranges from high values
indicative of forces being concentrated, as with wood-gouging lignivorous taxa, to low
values indicative of forces being distributed, as with the periphytivorous genus
Chaetostoma and the detritivorous/insectivorous species Rineloricaria lanceolata (Fig.
11).
Relative distribution or concentration of jaw force per unit of substrate area is an
aspect of jaw function not typically examined in teleosts, but one which is important for
both aquatic and terrestrial herbivores that use teeth for both food gathering and for the
grinding or trituration of plant cell walls necessary to enhance digestion of these low
quality (i.e. N-poor, C-rich) foods. Most loricariids are described as detritivores or
herbivores based on gut contents (Melo et al., 2004) and stable isotope analyses (Jepsen
and Winemiller, 2002). Unlike detritivorous/herbivorous fishes in the Perciformes,
Characiformes, and Cypriniformes, which typically have well-developed pharyngeal
jaws, a highly muscular stomach, and/or highly acidic stomachs for the trituration of cell
walls (Bowen, 1976, 1981, 1983), loricariids have relatively unspecialized pharyngeal
jaws (Armbruster, 2004), a thin-walled anterior alimentary canal (Delariva and
45
Agostinho, 2001), and intestinal pHs near neutral (German, 2008). Loricariid herbivores
may, therefore, benefit from the grinding of food particles against solid substrates during
jaw adduction and prior to ingestion.
Alternatively, nutritive deficiencies of low quality foods can be compensated for
by increasing consumption rate and food volume (Horn and Messer, 1992). This appears
to be the case in at least the species Ancistrus triradiatus and Panaque nigrolineatus,
which have gut passage times measured at 40 minutes (Hood et al., 2005) and four hours
(German, 2008), respectively. Given that loricariids have intestinal physiologies
optimized for the rapid assimilation of free monosaccharides and amino acids associated
with microbial degradation of detritus (German, 2008), it is likely that they ingest as
much food as they are able, given the need to first separate it from substrates. For
generalist consumers of flocculent detritus or loosely attached periphyton, low values of
the metric AM
area
/TRL indicative of force distribution would be expected. Indeed, species
of Chaetostoma have some of the lowest AM
area
/TRL values (Fig. 11) and some of the
longest/ broadest tooth rows of any loricariid (Fig. 3C), and gut content studies have
revelaed them to be detritivores in piedmont streams of Ecuador (Saul, 1975), Venezuela
(Hood et al., 2005), and Panama (Kramer and Bryant, 1995).
Broad, truncate tooth rows, or their invertebrate equivalent, are characteristic of a
wide variety of herbivorous/detritivorous grazers outside of Loricariidae. Caddisfly
larvae (Trichoptera), which are among the most ubiquitous metazoan grazers in temperate
streams, also frequently have a broad, bristly labrum with which they scrape algae and
gather detritus from stream surfaces (Fig. 12A; see also Arens, 1989, 1990, 1994). Sereno
et al. (2007) described a Cretaceous sauropod with an extremely broad muzzle
46
(Nigersaurus taqueti, Fig. 12B), and used the orientation of its otic canals and aspects of
its cervical vertebrae to support the hypothesis that it was specialized for a low, grazing
head posture, and the cropping of vegetation close to the ground. Solounias and
Moelleken (1993) examined the jaws and feces of a wide range of modern and extinct
ruminant mammals and were able to classify them into one of three feeding modes
(browsers, mixed feeders, and grazers) based solely on premaxilla width (narrowest to
widest, Fig. 12D; see also Janis and Erhardt, 1988). In Asian tropical streams, the
Gastromyzontinae (Cypriniformes) exhibit a range of jaw widths (Fig. 12C) that Roberts
(1989) associates with a diet spectrum from carnivory (narrow) to herbivory (wide). And
in the tropical rivers of Africa, the surface-scraping catfish tribe Atopochilini
(Mochokidae; Vigliotta, 2008) is remarkably convergent on members of the Loricariidae:
all atopochilins have very wide jaws that function within a fleshy labial disk and the tribe
is at least partially diagnosed by having a ventral mesethmoid condyle for articulation
with premaxillae, although neither atopochilin jaw function nor trophic ecology has been
described in detail.
In marine systems, Blenniidae (Springer, 1988) and the squamipinnes group (an
assemblage of nine morphologically and taxonomically diverse fish families including
the surgeonfishes, Acanthuridae, and the butterflyfishes, Chaetodontidae; Konow et al.,
2008) are abundant on reefs and include many surface-scraping species with relatively
broad, truncate muzzles. The squamipinnes group is also notable for having evolved a
novel intramandibular joint at least three and possibly five times, in each case associated
with a shift from suspended to surface-attached prey items (Konow et al., 2008). The
increased jaw flexion allowed by this joint, in all cases between the dentary and the
47
anguloarticular, appears to be an adaptation for surface scraping by allowing the tooth
row to remain in more continuous contact with surfaces during adduction. Similar
articulations between the dentary and the articular have been described in surface-
scraping parrotfishes (Scaridae; Bellwood and Choate, 1990). Likewise, Yamaoka (1982)
observed a weakening of the intramandibular symphysis between left and right
mandibular rami in the Petrochromis genus (Cichlidae) of surface-scraping, epilithic
algivores. Together, these comparative examples support the hypothesis that decoupling
of linkages as observed in the lineage leading to Loricariidae may have at least partially
been an adaptive response to a surface-scraping feeding mode, and the associated need to
independently accommodate tooth rows to surface irregularities during jaw adduction.
Surface scraping in Loricarioidea is likely to have evolved in close association
with surface attachment in high gradient hillstream habitats. Schaefer and Provenzano
(2008) describe several features of the pelvic girdle in members of the Lithogeninae that
they hypothesize are adaptations for surface attachment and, in conjunction with the
labial disk, locomotion via surface attachment (i.e. climbing, even up vertical surfaces
above the water line; Johnson, 1912). Lithogeninae is sister to either the Loricariidae
(Schaefer, 2003) or Astroblepidae (Armbruster, 2004; Hardman, 2005); and the
Astroblepidae, which are sister to the Loricariidae (Sullivan et al., 2006), are restricted to
high-elevation Andean hillstream habitats (Armbruster, 2004). Likewise, the basal
loricariid subfamily Delturinae is restricted to high gradient headwaters (Reis et al.,
2006). These morphological features, and the high gradient habitats characteristic of the
basal lineages Lithogeninae and Astroblepidae support the hypothesis that Loricariidae
evolved in hillstream habitats. Convergent modifications of the pelvic fin and oral disk
48
are common among several lineages of hillstream ichthyofaunas in tropical Asia and
Africa (Hora, 1922, Annandale and Hora, 1922, Hora, 1930, Saxena and Chandy, 1966,
Singh and Agarwal, 1993); in several of these lineages, jaw modifications for surface
scraping modes of feeding have evolved in association with development of labial disks
(Roberts and Stewart, 1976, Benjamen, 1986, Roberts, 1989).
2.4.3 Jaw mechanics: force geometry-traditional metrics
Relative distances between areas of force-in or force-out and axes of rotation
determine a rotating system?s mechanical advantage, or the degree to which the system is
optimized for force or speed. Previous investigations of teleost jaw function have
revealed a wide range of mechanical advantage ratios, calculated by dividing in-lever
distance by out-lever distance. In general, piscivorous fishes feeding on elusive prey
items have jaws with long out-levers, short in-levers, and mechanical advantages for jaw
closure as low as 0.04 (e.g. needlefishes, Belonidae; Westneat, 2004), indicating
optimization for speed. Conversely, fishes that feed on sessile, slow-moving, and/or hard
prey items have jaws with short out-levers, long in-levers, and mechanical advantages for
jaw closure as high as 0.68 (e.g. coralivorous parrotfishes, Scaridae; Westneat, 2004),
indicating optimization for force. Loricariid diets are entirely benthic, non-living
(detritus, wood), sessile (algae, diatoms, sponge), or slow-moving (e.g. aquatic
Trichoptera and Lepidoptera larvae, snails). Loricariids might, therefore, be predicted to
have relatively slow, force-optimized jaws; however, calculations of jaw closing
mechanical advantage using traditionally defined parameters (Fig. 5) describe a wide
range of mechanical advantages, from 0.24 at the distalmost tooth of the insectivorous
49
species Rineloricaria lanceolata, to 1.26 at the proximalmost tooth of the detritivorous
species Chaetostoma sp. 2 (Fig. 13).
If only the distalmost tooth is considered (as in most investigations of teleost jaw
closure), the wood-eating species Panaque n. sp. ?Mara?on? is the most force-optimized,
with a mean mechanical advantage of 0.41 (Fig. 13). Given the ability of loricariids to
synchronously apply distalmost and proximalmost teeth to substrates, mechanical
advantage at the proximalmost tooth may be equally functionally relevant; and Panaque
n. sp. ?Mara?on? has a proximalmost mechanical advantage of 0.57, also among the
highest of the sampled taxa (Fig. 13). Several species of algivorous/detritivorous taxa
have proximalmost tooth values much greater than Panaque n. sp. ?Mara?on?, including
all members of the algivorous/detritivorous genera Ancistrus and Chaetostoma (Fig. 13).
In consideration of the functional relevance of these major mechanical advantage
gradients between distalmost and proximalmost teeth, the force distributed per unit of
tooth row length, as predicted by AM
area
/TRL, should be factored into analyses of
mechanical advantage (Figs. 11, 13).
Some taxa like Panaque n. sp. ?Mara?on? have morphologies consistent with both
force concentration (Y-axis, Fig. 13) and force optimization at distalmost and
proximalmost teeth (X-axes, Fig. 13). Other taxa (e.g. Rineloricaria lanceolata,
Hypostomus unicolor) have morphologies consistent with force distribution and speed
optimization at distalmost and proximalmost teeth. Novel force intensity and traditional
force geometry metrics for these taxa converge upon descriptions of morphologies at
opposite ends of both functional spectra. In contrast, taxa such as Ancistrus and
Chaetostoma, are predicted to be force-distributers, and have a range of mechanical
50
advantage values from relatively speed optimized distalmost teeth, to force optimized
proximalmost teeth (e.g. Ancistrus sp. ?longjaw?; Fig. 13). This wide range of mechanical
advantage values across the tooth row challenges both the functional comparison of taxa
within Loricariidae, and the comparison of loricariid jaw mechanical advantage with that
of other teleosts. Indeed, several aspects of the relatively novel form and function of
loricariid jaws provide caveats to uncorrected interpretations of mechanical advantage
metrics computed using traditional anatomical definitions of in-levers and out-levers.
2.4.4 Jaw mechanics: force geometry-novel metrics
Variation in mechanical advantage is likely an important consequence of jaw
morphological variation across the Loricariidae; however, accurate prediction of
mechanical advantage is dependent upon precisely locating predominant axes of rotation.
The high degree of freedom of loricariid lower jaw rotation makes it unlikely that a
single, fixed axis of rotation exists, and the three-dimensional range across which the
rotational axis might shift is likely subject to interspecific variation. Detailed kinematic
data is generally lacking for Loricariidae, but the single cineradiographic study of
Pterygoplichthys gibbiceps (Hypostominae) by Adriaens et al. (2009) observed rotation
predominantly within the sagittal plane, represented by Axis 1 in Fig. 8. Viewed in this
light, relevance of the in-lever and out-lever parameters to predictions of mechanical
advantage is reduced by the nearly perpendicular orientation of these parameters with
respect to the sagittal plane (Fig. 8). Dimensions of shape variation most relevant to
loricariid jaw mechanical advantage are illustrated in the rotating cone model as radius 1
(r
1
; Fig. 8), the distance perpendicular to the predominant axis of rotation, from the axis
51
to the region of force-in, and radius 2 (r
2
; Fig. 8), the distance perpendicular to the
predominant axis of rotation, from the axis to the region of force-out. These are sagittal
components of the respective in-lever and distalmost out-lever, and should, therefore,
provide more accurate predictions of mechanical advantage generated during sagittal
rotation than measured distances from the AAC to respective regions of force-in and
force-out.
Direct measurement of r
1
 and r
2
 is not possible because of the inability to
precisely locate a dominant rotational axis in three dimensions. Instead, the parameter H1
(Fig. 6) was measured and treated as a correlate of combined variation in r
1
 and r
2
. H1
increases relative to each of the unmeasurable r
1
 and r
2
 parameters. Absent the ability to
compute ratios of r
1
 to r
2
, I hypothesize that major morphological trends observed among
the lower jaws of loricariid lignivores toward a shortened long axis, toward hypertrophy
of the coronoid arch, and toward increased distance between the coronoid arch and the
tooth row (Fig. 3), are functionally correlated, and independently contribute to more
forceful jaw geometries. Longer jaws with low or absent coronoid arches, and tooth rows
closer to the jaw?s long axis are hypothesized to be optimized for speed. Representing
near opposite ends of these extremes are the jaws of Chaetostoma cf. milesi (Cone 1, Fig.
8), and Panaque nigrolineatus (Cone 2, Fig. 8). By measuring relative height of the
coronoid arch, and its distance from the tooth row in the sagittal plane, H1 is
hypothesized to be a direct indicator of mechanical advantage, with high values
indicating optimization for force, and low values indicating optimization for speed.
Two metrics were calculated with the novel H1 parameter: H1/TRL and
Out
dist
/H1 (Fig. 14). The spectrum of H1/TRL values is hypothesized to provide a
52
combined measure of force optimization and force concentration, versus combined speed
optimization and force distribution. High values are observed among the wood-eating
taxa Panaque spp., and Hypostomus pyrineusi, and low values are observed among both
the algivorous long-tooth-row genus Chaetostoma and the short-tooth-row insectivorous
or detritivorous genera Loricaria and Rineloricaria (Fig. 14). In these latter taxa, speed
optimization may be adaptive for either the rapid consumption of large volumes of low
quality food, or for the capture of prey items that are mobile. Zuanon (1999) examined
the diet of Ancistrus ranunculus, which was not examined herein, but which has very
elongate jaw rami and relatively short tooth rows convergent upon jaw morphologies of
some Loricariini (per. obs.), and hypothesized that it is a filter-feeder. Likewise, K. O.
Winemiller (pers. comm.) has hypothesized that many loricariines may feed by
winnowing loosely aggregated benthic food items in a manner similar to that of sturgeon
(Carroll and Wainwright, 2003). Fast adduction of speed-optimized jaws that are
relatively long and gracile with low coronoid arches may serve a hydrodynamic function
in filter-feeders and substrate winnowers by enhancing current flows into the buccal
chamber.
Length to height ratios of loricariid jaws are described by the metric Out
dist
/H1,
with high values describing relatively long and gracile jaws (e.g. Loricariini), and
intermediate values describing moderately long and moderately tall jaws (e.g.
Chaetostoma, Fig. 14; Cone 1, Fig. 8). Low values of Out
dist
/H1 describe relatively
narrow and tall jaws (e.g. most Hypostominae, Fig. 14; Cone 2, Fig. 8). By describing
gross dimensions, this metric is also hypothesized to predict the potential for torque
magnification through the jaw ramus in sagittal (low values of Out
dist
/H1) versus
53
horizontal (high value of Out
dist
/H1) planes. Relatively long and low jaws (e.g. Loricaria,
Rineloricaria) would be expected to be subject to greater torque along the horizontal or
transverse planes (Axis 2 and 3, Fig. 8), but given the correspondingly low AM
area
/TRL
and H1/TRL values for these taxa, they are likely subject to lower overall torsional
forces. On the other hand, relatively narrow and tall jaws (e.g. Panaque spp., Hypostomus
pyrineusi; Cone 2, Fig. 8) with low Out
dist
/H1 values concentrate their mass more within
the sagittal plane, and would be predicted to be subject to more intense torsional forces
by their higher AM
area
/TRL values.
The final two metrics TR
angle
 and Out
dist
-Out
prox
/In are correlated descriptors of
the length and orientation of the tooth row, and are hypothesized to predict torque
differentials across loricariid tooth rows during jaw adduction. Rotating systems that, like
loricariid lower jaws, distribute forces across an extended surface with variable resistance
are likely subject to torque differentials, or ranges of torque values from regions of force-
out closest to the axis of rotation to regions of force-out furthest from the axis of rotation.
One way to reduce such torque differentials is to reduce the area across which force is
distributed (lower TRL). A second way is to change the orientation, or angle, of the area
across which force is distributed. Increasing the angle of the loricariid dentary tooth row
with respect to the distalmost out-lever arm, for example, shifts the relative positions of
distalmost and proximalmost teeth so that the difference between their distances to the
anguloarticular condyle is reduced. This result of increased tooth row angles is
summarized by the metric Out
dist
-Out
prox
/In (i.e., the difference between out-lever arms
standardized to the in-lever arm). Taxa with high TR
angle
 values (Fig. 15) are
hypothesized to have a relatively low torque differential as predicted by the correlated
54
metric Out
dist
-Out
prox
/In (Fig. 16). Selection for reduction of the torque differential via
increased TR
angle
 is likely to be greatest among those taxa with jaws subject to more
intense torsional forces as predicted by AM
area
/TRL (e.g. Panaque, Hypostomus
pyrineusi; Fig. 11). Relatively weak-jawed taxa in the Loricariinae also exhibit high
TR
angle
 values; however, these morphologies may result not from selection for decreased
torque differential but for increased tooth row protrusion associated with browsing or
selective feeding ? similar to increases in acuteness of premaxillary tooth row angles that
can be observed in the muzzles of selectively browsing ruminants (Fig. 12).
2.5 CONCLUSIONS
I proposed methods by which morphological diversity across lower jaws of the
Loricariidae may be quantified and compared in a homologous and functionally
informative manner. These methods were used to describe jaw morpho-functional
diversity across an assemblage of loricariids in the upper Amazon Basin of Northern Peru
consisting of 5 tribes, 12 genera, and 25 species. Five of these species, comprised of
species of two different lineages, have been previously described as specialized wood-
eaters based on gut-content data; all had distinctive jaw morphologies consistent with
convergence upon a durophagous jaw morphology and within this guild at least four
species were distinguishable from each other by their distinctive jaw morpho-functional
attributes alone. In addition to describing taxonomic patterns of jaw morphological
diversity, several explicit hypotheses for how jaw morphometric parameters might be
used to predict function were proposed. Loricariidae has been hypothesized to have
higher rates of jaw morphological and functional diversification than loricarioid clades
55
with plesiomorphic jaw linkages, or constraints, missing in Loricariidae (Schaefer and
Lauder, 1986). Methods proposed herein provide several means by which jaw diversity in
Loricariidae may be described and correlated with extrinsic ecological and
biogeographical factors. They also provide several discrete, functionally related
morphometrics that, in conjunction with morphology-independent phylogenies, can be
used to investigate rates and patterns of loricariid jaw morphological, functional, and
ecological diversification. Given the intrinsic capacity for morphological and functional
diversification putatively enhanced by decoupling in the Loricarioidea, and the trophic
ecological diversity available to loricariids as basal consumers in the world?s largest
tropical freshwater systems, there is great potential for considerable undescribed trophic
and ecological diversity in Loricariidae. Highly derived structure and function of
loricariid jaws confound attempts to place their morphological and functional radiation in
the context of previous investigation of jaw functional diversity in the Actinopterygii.
Indeed, the structure and function of loricariid jaws are perhaps best compared to other
surface-attaching, surface-scraping feeders from tropical rivers in Asia and Africa (e.g.
Cyprinidae, Balitoridae, and Gyrinocheilidae (Cypriniformes), and Mochokidae and
Sisoridae (Siluriformes)) ? all of which are still awaiting detailed investigations of jaw
morphological and functional diversity.
56
2.6 REFERENCES
Adriaens, D., Geerinckx, T., Vlassenbroeck, J., Van Hoorebeke, L., Herrel, A., 2009.
Extensive jaw mobility in suckermouth armored catfishes (Loricariidae): a
morphological and kinematic analysis of substrate scraping mode of feeding.
Physiological and Biochemical Zoology 82, 51?62.
Albertson, C.R., Streelman, J. T., Kocher, T.D., Yelick, P.C., 2005. Integration and
evolution of the cichlid mandible: the molecular basis of alternate feeding
strategies. Proceedings of the National Academy of Science 102, 16287?16292.
Alfaro, M., Bolnick, D.I., Wainwright, P.C., 2005. Evolutionary consequences of many-
to-one mapping of jaw morphology in labrid fishes. The American Naturalist 165,
E140?E154.
Alfaro, M.E., Bolnick, D.I., Wainwright, P.C., 2004. Evolutionary dynamics of complex
biomechanical systems: an example using the four-bar mechanism. Evolution 58,
495?503.
Anderson, P.S.L., Westneat, M.W., 2006. Feeding mechanics and bite force modelling of
the skull of Dunkleosteus terrelli, an ancient apex predator. Biological Letters 3,
76?79.
Annandale, N., Hora, S.L., 1922. Parallel evolution in the fish and tadpoles of mountain
57
torrents. Records of the Indian Museum 24, 505?509.
Arens, W., 1989. Comparative functional morphology of the mouthparts of stream
animals feeding on epilithic algae. Archiv f?r Hydrobiologie 83 (Supplement),
253?354.
Arens, W., 1990. Wear and tear on mouthparts: a critical problem in stream animals
feeding on epilithic algae. Canadian Journal of Zoology 68, 1896?1914.
Arens, W. 1994. Striking convergence in the mouthpart evolution of stream-living algae
grazers. Journal of Zoological Systematics & Evolutionary Research 32, 319?343.
Armbruster, J.W., 2004. Phylogenetic relationships of the suckermouth armoured
catfishes (Loricariidae) with emphasis on the Hypostominae and the Ancistrinae.
Zoological Journal of the Linnean Society 141, 1?80.
Armbruster, J.W., 2008. The genus Peckoltia with the description of two new species and
a reanalysis of the phylogeny of the genera of the Hypostominae (Siluriformes:
Loricariidae). Zootaxa 1822, 1?76.
Benjamin, M., 1986. The oral sucker of Gyrinocheilus aymonieri (Teleostei:
Cypriniformes). Journal of Zoology, London, B 1, 211?254.
Bookstein, F.L., 1991. Morphometric Tools for Landmark Data. Cambridge University
58
Press, New York.
Bouton, N., van Os, N., Witte, F., 1998. Feeding performance of Lake Victoria rock
cichlids: testing predictions from morphology. Journal of Fish Biology 53 (Suppl.
A), 118?127.
Delariva, R.L., Agostinho, A.A., 2001. Relationships between morphology and diets of
six neotropical loricariids. Journal of Fish Biology 58, 832?847.
Ebeling, A.W., 1957. The dentition of eastern Pacific mullets, with special reference to
adaptation and taxonomy. Copeia 1957, 173?185.
Fugi, R., Agostinho, A.A., Hahn, N.S., 2001. Trophic morphology of five benthic-
feeding fish species of a tropical floodplain. Revista Brasileira de Biologia 61,
27?33.
Geerinckx, T., 2006. Ontogeny and functional morphology of a highly specialized trophic
apparatus: a case study of neotropical suckermouth armoured catfishes
(Loricariidae, Siluriformes). Dissertation. University of Gent, Gent.
Geerinckx, T, De Poorter, J., Adriaens, D., 2007. Morphology and development of teeth
and epidermal brushes in loricariid catfishes. Journal of Morphology 268,
805?814.
Geerinckx, T., Brunain, M., Adriaens, D., 2007. Development of the osteocranium in the
suckermouth armored catfish Ancistrus cf. triradiatus (Loricariidae,
Siluriformes). Journal of Morphology 268, 254?274.
59
Geerinckx, T., Brunain, M., Herrel, A., Aerts, P., Adriaens, D., 2007. A head with a
suckermouth: a functional-morphological study of the head of the suckermouth
catfish Ancistrus cf. triradiatus (Loricariidae, Siluriformes). Belgian Journal of
Zoology 137, 47?66.
Hardman, M., 2005. The phylogenetic relationships among non-diplomystid catfishes as
inferred from mitochondrial cytochrome b sequences; the search for the ictalurid
sister taxon (Otophysi: Siluriformes). Molecular Phylogenetics and Evolution 37,
700?720.
Hora, S.L., 1922. Structural modifications in the fish of mountain torrents. Records of the
Indian Museum 24, 31?61.
Hora, S.L., 1930. Ecology, bionomics and evolution of the torrential fauna, with special
reference to the organs of attachment. Philosophical Transactions of the Royal
Society of London, Biological Sciences 218, 171?282.
Horn, M.H., Messer, K.S., 1992. Fish guts as chemical reactors: a model of the
alimentary canals of marine herbivorous fishes. Marine Biology 113, 527?535.
Hughes, C., Eastwood, R., 2006. Island radiation on a continental scale: Exceptional rates
of plant diversification after uplift of the Andes. Proceedings of the National
Academy of Sciences 103, 10334?10339.
Hulsey, C.D., Garc?a de Le?n, F.J., Rodiles-Hern?ndez, R., 2006. Micro- and
macroevolutionary decoupling of cichlid jaws: a test of Liem's key innovation
60
hypothesis. Evolution 60, 2096?2109.
Janis, C.M., Ehrhardt, D., 1988. Correlation of relative muzzle width and relative incisor
width with dietary preference in ungulates. Zoological Journal of the Linnean
Society 92, 267?284.
Jepsen, D.B., Winemiller, K.O., 2002. Structure of tropical river food webs revealed by
stable isotope ratios. Oikos 96, 46?55.
Johnson, R.D.O., 1912. Notes on the habits of a climbing catfish (Arges marmoratus)
from the Republic of Colombia. Annals New York Academy of Sciences 22,
327?333.
Konow, N., Bellwood, D.R., Wainwright, P.C., Kerr, A.M., 2008. Evolution of novel jaw
joints promote trophic diversity in coral reef fishes. Biological Journal of the
Linnean Society 93, 545?555.
Larson, P.M., Reilly, S.M., 2003. Functional morphology of feeding and gill irrigation in
the anuran tadpole: electromyography and muscle function in larval Rana
catesbeiana. Journal of Morphology 255, 202?214.
Lauder, G.V., 1979. Feeding mechanics in primitive teleosts and in the halecomorph fish
Amia calva. Journal of Zoology, London 187, 543?578.
Lauder, G.V., Crompton, A.W., Gans, C., Hanken, J., Liem, K.F., Maier, W.O., Meyer,
A., Presley, R., Rieppel, O.C., Roth, G., Schluter, D., Zweers, G.A., 1980. Group
report: how are feeding systems integrated and how have evolutionary
61
innovations been introduced? In: Wake, D.B., Roth, G. (Ed.), Complex
Organismal Functions: Integration and Evolution in Vertebrates. John Wiley &
Sons Ltd., pp. 97?115.
Melo, C.E. de, de Arruda Machado, F., Pinto-Silva, V., 2004. Feeding habits of fish from
a stream in the savanna of Central Brazil, Araguaia Basin. Neotropical
Ichthyology 2, 37?44.
Miya, M., Holcroft, N.I., Satoh, T. P., Yamaguchi, M., Nishida, M., Wiley, E.O., 2007.
Mitochondrial genome and a nuclear gene indicate a novel phylogenetic position
of deep-sea tube-eye fish (Stylephoridae). Ichthyological Research 54, 323?332.
Mochizuki, K., Fukui, S., 1983. Development and replacement of upper jaw teeth in
gobiid fish, Sicyopterus japonicus. Japanese Journal of Ichthyology 30, 27?36.
Muller, S., Weber, C., 1992. Les dents des sous-familles Hypostominae et Ancistrinae
(Pisces, Siluriformes, Loricariidae) et leur valeur taxonomique. Revue suisse de
Zoologie 99, 747?754.
M?rona, B. de, Hugueny, B., Tejerina-Garro, F.L., 2008. Diet-morphology relationship in
a fish assemblage from a medium-sized river of French Guiana: the effect of
species taxonomic proximity. Aquatic Living Resources 21, 171?184.
Norris, K.S., Prescott, J.H., 1959. Jaw structure and tooth replacement in the opaleye,
Girella nigricans (Ayres) with notes on other species. Copeia 1959, 275?283.
Ono, R.D., 1980. Fine structure and distribution of epidermal projections associated with
62
taste buds on the oral papillae in some loricariid catfishes (Siluroidei:
Loricariidae). Journal of Morphology 164, 139?159.
Rapp Py-Daniel, L., 1997. Phylogeny of the neotropical armored catfishes of the
subfamily Loricariinae (Siluriformes: Loricariidae). PhD dissertation. The
University of Arizona.
Reis, R.E., Pereira, E.H.L., Armbruster, J.W., 2006. Delturinae, a new loricariid catfish
subfamily (Teleostei, Siluriformes), with revisions of Delturus and
Hemipsilichthys. Zoological Journal of the Linnean Society 147, 277?299.
Roberts, T.R., 1989. The freshwater fishes of Western Borneo (Kalimantan Barat,
Indonesia). Memoirs of the California Academy of Sciences 14, 1?210.
Roberts, T.R., Stewart, D.J., 1976. An ecological and systematic survey of fishes in the
rapids of the Lower Zaire of Congo River. Bulletin of the Museum of
Comparative Zoology 147, 239?317.
Saxena, S. C., Chandy, M., 1966. Adhesive apparatus in certain Indian hill stream fishes.
Journal of Zoology 148, 315?340.
Schaefer, S.A., 1987. Osteology of Hypostomus plecostomus (Linnaeus), with a
phylogenetic analysis of the loricariid subfamilies (Pisces: Siluriformes).
Contributions in Science, Natural History Museum of Los Angeles County 394,
1?31.
Schaefer, S.A., 1988. Homology and evolution of the opercular series in the loricarioid
63
catfishes (Pisces: Siluroidei). Journal of Zoology, London 214, 81?93.
Schaefer, S.A., 2003. Relationships of Lithogenes villosus Eigenmann, 1909
(Siluriformes, Loricariidae): evidence from high-resolution computed
microtomography. American Museum Novitates 3401, 1?26.
Schaefer, S.A., Stewart, D.J., 1993. Systematics of the Panaque dentex species group
(Siluriformes: Loricariidae), wood-eating armored catfishes from tropical South
America. Ichthyological Exploration of Freshwaters 4, 309?342.
Schaefer, S.A., Provenzano, F., 2008. The Lithogininae (Siluriformes, Loricariidae):
anatomy, interrelationships, and description of a new species. American Museum
Novitates 3637, 1?49.
Schaefer, S.A., Lauder, G.V., 1986. Historical transformation of functional design:
evolutionary morphology of feeding mechanisms in loricarioid catfishes.
Systematic Zoology 35, 489?508.
Schaefer, S.A., Lauder, G.V., 1996. Testing historical hypotheses of morphological
change: biomechanical decoupling in loricarioid catfishes. Evolution 50,
1661?1675.
Singh, N., Agarwal, N.K., 1993. Organs of adhesion in four hillstream fishes, a
comparative morphological study. In: Singh, H.R. (Ed.), Advances in Limnology.
Narendra Publishing House, New Delhi, pp. 311?316.
Solounias, N., Moelleken, S.M.C., 1993. Dietary adaptation of some extinct ruminants
64
determined by premaxillary shape. Journal of Mammalogy 74, 1059?1071.
Springer, V.G., 1988. The Indo-Pacific blenniid fish genus Ecsenius. Smithsonian
Contributions to Zoology 465, 1?134.
Sullivan, J.P., Lundberg, J.G., Hardman, M., 2006. A phylogenetic analysis of the major
groups of catfishes (Teleostei: Siluriforms) using rag1 and rag2 nuclear gene
sequences. Molecular Phylogenetics and Evolution 41, 636?662.
Vermeij, G.J., 1973. Biological versatility and Earth history. Proceedings of the National
Academy of Sciences USA 70, 1936?1938.
Vermeij, G.J., 1974. Adaptation, versatility, and evolution. Systematic Zoology 22,
466?477.
Vigliotta, T.R., 2008. A phylogenetic study of the African catfish family Mochokidae
(Osteichthyes, Ostariophysi, Siluriformes), with a key to genera. Proceedings of
the Academy of Natural Sciences of Philadelphia 157, 73?136.
Wassersug, R.J., Yamashita, M., 2001. Plasticity and constraints on feeding kinematics in
anuran larvae. Comparative Biochemistry and Physiology, Part A 131, 183?195.
Westneat, M.W., 2004. Evolution of levers and linkages in the feeding mechanisms of
fishes. Integrative and Comparative Biology 44, 378?389.
Yamaoka, K., 1982. Morphology and feeding behaviour of five species of genus
Petrochromis (Teleostei, Cichlidae). Physiology and Ecology Japan 19, 57?75.
65
Table 1. Taxa and sample sizes examined in respective jaw and muscle parts of this
study. All specimens collected from middle reaches of the Mara?on River, a tributary of
the upper Amazon in northern Peru.
  
Species
N
jaw
N
muscle
SUBFAMILY: Hypostominae
TRIBE: Ancistrini
Ancistrus sp. 'longjaw'5
Ancistrus sp. 'shortjaw'3
Ancistrus sp. 'wormline'2
Chaetostoma cf. milesi4
Chaetostoma microps276
Chaetostoma sp. 127
Chaetostoma sp. 2166
Chaetostoma sp. 3336
Chaetostoma sp. 415
Lasiancistrus schomburgkii226
New genus 3186
Panaque albomaculatus134
Panaque gnomus306
Panaque n. sp. 'Mara?on'135
Panaque nocturnus836
Peckoltia bachi2
TRIBE: Hypostomini
Hypostomus emarginatus3
Hypostomus niceforoi4
Hypostomus pyrineusi196
Hypostomus unicolor4
SUBFAMILY: Loricariinae
TRIBE: Farlowellini
Farlowella amazonum3
TRIBE: Harttiini
Lamontichthys filamentosus186
TRIBE: Loricariini
Loricaria clavipinna3
Rineloricaria lanceolata3
  Spatuloricaria puganensis9 4
66
Figure 1. Medial, sagittal sections through the snout and jaws of (A) Chaetostoma cf. milesi, (B) Leporacanthicus joselimai, and
(C) Panaque nigrolineatus. Labels: ap = ascending process of the premaxilla, dn = dentary, me = mesethmoid, mec = mesethmoid
condyle, pm = premaxilla, rpf = retractor premaxillae fossa. CT (computer aided tomography) reconstructions courtesy of the
Digimorph lab.
66
67
Figure 2. Ventral views of the snout and jaws of (A) Chaetostoma cf. milesi, (B)
Leporacanthicus joselimai, and (C) Panaque nigrolineatus. Labels: dn = dentary, pm =
premaxillae, aac = anguloarticular condyle. CT (computer aided tomography)
reconstructions courtesy of the Digimorph lab.
68
Figure 3. Representative sample of lower jaws from upper Amazon Loricariidae examined in this study: HYPOSTOMINAE:
ANCISTRINI: (A) Ancistrus sp. ?longjaw?, (B) Ancistrus sp. ?shortjaw?, (C) Chaetostoma sp. 1, (D) Panaque albomaculatus, (E)
Panaque gnomus, (F) Panaque n. sp. ?Mara?on?, (G) Panaque nocturnus, (H) Peckoltia bachi; HYPOSTOMINAE: HYPOSTOMINI:
(I) Hypostomus emarginatus, (J) Hypostomus niceforoi, (K) Hypostomus pyrineusi in the H. cochliodon-group, (L) Hypostomus
unicolor; LORICARIINAE: HARTTIINI: (M) Farlowella amazona, (N) Lamontichthys filamentosus, (O) Rineloricaria lanceolata;
LORICARIINAE: LORICARIINI: (P) Spatuloricaria puganensis.
68
69
Figure 4. Examples of loricariid oral disks: (A) Scobinancistrus sp. and (B) Leporacanthicus cf. galaxias. Photos by M. Sabaj
P?rez.
69
70
Figure 5. The lower jaw ramus of Chaetostoma cf. milesi (A, B, C) and Panaque nigrolineatus (D, E, F) in positions illustrating
the horizontal orientation, ventral perspective (A, D), vertical orientation (B, E), and horizontal orientation, dorsal perspective (C,
F). Functional parameters are labled on each perspective and view from which they were measured: AM
area
 = adductor mandibulae
insertion area, In = in-lever from center of adductor mandibulae area of insertion to anguloarticular condyle, Out
dist
 = out-lever
from anguloarticular condyle to distalmost tooth, Out
prox
 = out-lever from angularticular condyle to proximalmost tooth, TRL =
tooth row length. CT (computer aided tomography) reconstructions courtesy of the Digimorph lab and Kyle Luckinbill, ANSP.
70
71
Figure 6. The lower jaw ramus of (A) Chaetostoma cf. milesi and (B) Panaque
nigrolineatus in horizontal orientation, dorsal perspective, illustrating distances measured
as the distalmost out-lever (Out
dist
) and coronoid height (H1) parameters. CT (computer
aided tomography) reconstructions courtesy of the Digimorph lab and Kyle Luckinbill,
ANSP.
72
Figure 7. Relationship between combined adductor mandibulae area of insertion (AM
area
) scaled to standard length (SL), and
adductor mandibulae volume (AM
vol
) scaled to standard length. Area of adductor mandibulae insertion and volume of adductor
mandibulae were measured from separate subsets of specimens, so the correlation is of means of each scaled separately to standard
length. Plotted values are means ? standard deviation. See Table 1 for species sample sizes.
72
73
Figure 8. Three-dimensional rotating cone model linking loricariid lower jaw morphology and function. Cones modeled after the
right lower jaw ramus of Chaetostoma cf. milesi (Cone 1) and Panaque nigrolineatus (Cone 2) illustrate near opposite ends of a
spectrum of both measured and qualitatively described jaw morphology discussed in this study. Perspective is from the posterior.
Red lines are potential axes of rotation, with Axis 1 representing rotation in the sagittal plane, Axis 2 representing rotation in the
horizontal plane, and Axis 3 representing rotation in the transverse plane. Capitalized parameters and parameters represented by
solid green, black, and gray lines were measured or quantified by proxy in this study (Fig. 5). See text for discussion of output
plane represented by gray triangle. Tooth row angle (TR
angle
) represented by ?.
73
74
Figure 9. Principle component (PC) analysis of six parameters hypothesized to be
functionally relevant to the loricariid lower jaw (see Figs. 5, 6). Data from 379
individuals and 25 species of Loricariidae collected from the upper Amazon in northern
Peru (see Table 1). Parameters listed along each PC axis from left to right or from bottom
to top in order of eigenvector magnitude. PCs one through four describe 85.7%, 11.9%,
2.0%, and 0.3% of variation in the dataset, respectively.
75
Figure 10. Principle component (PC) analysis of six parameters hypothesized to be
functionally relevant to the loricariid lower jaw (see Figs. 5, 6). Data from 379
individuals and 25 species of Loricariidae collected from the upper Amazon in northern
Peru (see Table 1). Parameters listed along each PC axis from left to right or from bottom
to top in order of eigenvector magnitude. PCs one through four describe 85.7%, 11.9%,
2.0%, and 0.3% of variation in the dataset, respectively.
76
Figure 11. Adductor mandibulae area of insertion (AM
area
) over tooth row length (TRL), hypothesized to predict force intensity,
with high values indicating force concentration and low values indicating force distribution. Plotted values are means ? standard
deviation. See Table 1 for species sample sizes.
76
77
Figure 12. Examples of convergent trends in herbivorous jaw morphologies from (A)
Insecta, Trichoptera (labrum in dorsal view modified from Satija & Satija, 1959), (B)
fossil Sauropoda, Nigersaurus taqueti (lower jaw in dorsal view modified from Sereno et
al., 2007), (C) Cypriniformes, Gastromyzontinae (snout in ventral view modified from
Roberts, 1989), and (D) extinct Mammalia, Giraffidae (premaxillae in ventral view
modified from Solounias & Moelleken, 1993). Series in (C) hypothesized to be that of a
carnivore (narrowest, bottom) to herbivore (broadest, top). Series in (D) hypothesized to
represent a foraging spectrum from browser (narrowest, bottom) to mixed grazer (middle)
to grazer (broadest, top), with support provided by fecal data (Solounias & Moelleken,
1993). See Fig. 2 for similar spectrum of jaw morphologies in Loricariidae.
78
Figure 13. The novel metric of force intensity AM
area
/TRL (see Fig. 11) plotted against traditional metrics of mechanical
advantage: distance from anguloarticular condyle to center of adductor mandibulae insertion area (In) over respective distances
from anguloarticular condyle to distalmost (Out
dist
) and proximalmost (Out
prox
) tooth insertions. Capital letters indicate jaws
illustrated in Fig. 3. Plotted values are means ? standard deviation. See Table 1 for species sample sizes.
78
79
Figure 14. Three novel metrics of lower jaw morphological diversity hypothesized to be linked to function: (1) Adductor
mandibulae area of insertion (AM
area
) over tooth row length (TRL) ? hypothesized to be a measure of force distribution versus
force concentration; (2) distance parameter H1 (see Fig. 6) over TRL ? hypothesized to be a combined measure of mechanical
advantage (force vs. speed optimization) and force distribution; and (3) the distal out-lever arm (Out
dist
, see Fig. 5) over H1 ? a
measure of the gross length (Out
dist
) versu height (H1) dimensions of the lower jaw and hypothesized to be an indicator of the
predominant plane of torque through the lower jaw. Capital letters indicate jaws illustrated in Fig. 3. Plotted values are means ?
standard deviation. See Table 1 for species sample sizes.
79
80
Figure 15. Mean angles and angular standard deviations between the tooth row and distalmost output lever for species examined
in this study. Plotted values are means ? standard deviation. See Table 1 for sample sizes.
80
81
Figure 16. The metric Out
dist
-Out
prox
/In, a measure of torque differential across the tooth arcade. Plotted values are means ?
standard deviation. See Table 1 for sample sizes.
81
82
CHAPTER 3. STABLE ISOTOPES REVEAL TROPHIC STRUCTURE WITHIN
SYMPATRICALLY DIVERSE ASSEMBLAGES OF NEOTROPICAL
DETRIVOROUS FISHES
3.1 INTRODUCTION
Detritus is broadly recognized as the foundation of most aquatic food webs
(Mann, 1972, 1988, Wetzel et al., 1972, Cummins, 1973), yet the fine-scale use and
partitioning of detrital resources, and variation in the contribution of detritus to the
biodiversity and productivity of food webs, remain poorly understood (Moore et al.,
2004). Prior to robust quantitative descriptions of the trophic structure of tropical
freshwater fish communities (e.g. Flecker, 1992, Winemiller, 1990, Lewis et al., 2001,
Layman et al., 2005), fishes were described as primarily carnivorous, and largely
dependent on invertebrates for their trophic connections to either detritus or primary
production (Odum, 1970, Eggers et al., 1978). This pattern is not the case for lowland
tropical freshwater fish communities in general, and Neotropical riverine fish
communities especially. Indeed, the detritivore niche, which is almost entirely associated
with invertebrates at temperate latitudes (Anderson and Sedell, 1979, Cummins, 1979,
Maltby, 1994), is filled largely by fishes in lowland tropical systems and detritivores
compose most of fish biomass in most Neotropical rivers (Bonetto et al., 1969, Quir?s
and Baig?n, 1985, Araujo-Lima et al., 1986, Bayley, 1989, Benedito-Cecilio et al., 2000,
83
Alvim and Peret, 2004). Neotropical freshwater fish communities are among the most
diverse in the world, with total fish richness estimated to be between 5000 and 8000
species (Lundberg et al., 2000). Describing major patterns of variation within the trophic
structure of these highly diverse communities is a challenging but pressing goal for fish
ecologists. Highly detritus-dependent Neotropical fish food webs are facing severe
perturbation from overfishing (Bayley and Petrere, 1986, Galvis and Mojica, 2007,
Rodr?guez et al., 2007) and habitat alteration (Winemiller et al., 1996, Mol and Ouboter,
2004), just as light is beginning to be shed on the astounding diversity, distinctiveness,
and complexity of these systems, particularly compared with well-studied temperate
counterparts (Benedito-Cecilio and Araujo-Lima, 2002, Flecker et al., 2002, Jepsen and
Winemiller, 2007).
My study focused on the Loricariidae, a diverse radiation of mostly detritivorous
catfishes that are endemic to the freshwaters of tropical Central and South America.
Carbon and nitrogen stable isotope data and several recently proposed metrics for the
analysis and comparison of community-wide consumer isotope data (Layman et al.,
2008) were used to reveal the potential for trophic differentiation within Loricariidae,
upon a diet that has been largely treated as homogenous. Detrital resources are
heterogeneous on a fine scale, and determination of variation in detrital composition and
quality is frequently dependent not on gross taxonomic identification of component
species, but on its chemical composition and nutrient ratio (Frost et al., 2002, Cross et al.,
2003, Moe et al., 2005). Stable isotope techniques for food-web analysis (reviewed by
Fry and Sherr, 1984, Peterson and Fry, 1987, Thompson et al., 2005, Newsome et al.,
2007) have greatly improved the description of the fate and food value of detritus in
84
aquatic ecosystems (Roman and Tenore, 1984); however, studies inferring the structure
of detritus- or algae-based food webs from isotope data have traditionally struggled with
both the differentiation of basal resource components and the identification of a spectrum
of discrete, consistent ?
13
C and ?
15
N basal resource values needed for accurate
association with consumers (Hecky and Hesslein, 1995, Post, 2002, Hamilton et al.,
2005).
Layman et al. (2008) proposed that consumer isotopic signatures could be used to
compare trophic structure among communities in the absence of either finely partitioned
basal resource data or a well-supported linkage model of trophic relationships within
communities. They proposed the following 6 metrics that each quantify, or correct for, a
different potential aspect of variation across food webs that might be separated by space,
time, and/or taxonomic composition:
1) ?
15
N range (NR): Measured as the distance between mean ?
15
N values of the
most ?
15
N-enriched and most ?
15
N-depleted member of the community. NR is
treated as a measure of vertical structure within the community, and a potential
indicator of the number of trophic levels present.
2) ?
13
C range (CR): Measured as the distance between mean ?
13
C values of the
most ?
13
C-enriched and most ?
13
C-depleted member of the community. CR is
treated as a measure of the diversity of a community?s basal resources.
3) Total area (TA): Measured as the area of a convex polygon, of which each
vertex is the mean ?
15
N and ?
13
C value of a species at the perimeter of the
community in ?
13
C-?
15
N biplot space. TA is treated as a measure of the total
amount of trophic niche space occupied by the community.
85
4) Mean distance to centroid (CD): Calculated as the mean Euclidean distance
from the community centroid (mean) to the mean ?
15
N/?
13
C value of each
species in the community. CD is treated primarily as a measure of trophic
diversity within the community, and, to a lesser extent, the trophic distance
among species within the community.
5) Mean nearest neighbor distance (NND): Calculated as the mean Euclidean
distance between mean ?
13
C-?
15
N values of each species and the species to
which it is closest in ?
13
C-?
15
N bi-plot space. NND is treated primarily as a
measure of species packing within community trophic niche space and, to a
lesser extent, trophic diversity within the community.
6) Standard deviation of nearest neighbor distance (SDNND): Calculated as the
standard deviation of NND, and treated as a measure of the evenness of species
packing in community trophic niche space.
The 6 Layman et al. (2008) metrics were critiqued by Hoeinghaus and Zueg (2008),
who asserted that foodweb ecologists should persist in the struggle to identify and
quantify patterns of isotopic variation across basal resources and that in most ecosystems,
failing to carefully consider the effect of variation in the isotopic baseline on variation in
consumers largely negates any inferences that can be made about foodweb structure.
Layman and Post (2008), in a reply to Hoeinghaus and Zueg (2008), questioned the
latter?s generalization to most ecosystems. Layman and Post (2008) acknowledged that
finely partitioned gut content data and a nuanced understanding of variation in the
isotopic baseline are valuable; but, they state that in most systems the resource-based
standardization methods proposed by Hoeinghaus and Zueg (2008; also Newsome et al.,
86
2007) are not possible because of the high diversity of basal resources. Layman and Post
(2008) also concluded that one of the ways to resolve problems raised by Hoeinghaus and
Zueg (2008) would be to include a covariate such as species richness.
I applied the community-wide metrics of Layman et al. (2008) to ?
13
C and ?
15
N
data from 19 sympatric assemblages of mostly detritivorous Neotropical catfishes (Table
1), varying in richness of species (2?16), genera (1?11), tribes (2?6), subfamilies (1?3),
and families (1-2). Isotope data from a total of 79 species of Loricariidae and 3 species of
Astroblepidae (sister to Loricariidae) were examined. Localities of sampled assemblages
were broadly distributed across northern South America, including 2 sites in the
Essequibo River Basin (Guyana), three sites in the Orinoco River Basin (Venezuela), and
12 sites in the Amazon River Basin (Brazil, Guyana, Venezuela, and Peru; Fig. 1). South
American rivers have historically been divided into 3 major limnological categories
according to their water chemistry and underlying geology (Sioli, 1975, 1984). My study
was restricted to two of these water types: clearwater rivers draining the geologically
ancient Brazilian and Guiana Shields in central and northern South America, and
whitewater rivers draining the geologically younger Andes Mountains in western South
America. The third major type, blackwater rivers, which typically drain low lying areas
of the shields or Amazon Basin, are highly acidic with little primary productivity (Jepsen
and Winemiller, 2007), and possess a depauperate fauna of loricariid catfishes, were not
examined in this study.
Loricariidae is a taxonomically highly diverse family, containing over 700
described species, or 20% of the approximately 3600 fish species currently described
from Neotropical freshwaters (Reis et al., 2003). Popularly known as plecos in the
87
aquarium trade, or suckermouth armored catfishes, loricariids are distinguished by having
armored plating and a ventral oral disk. Throughout their native range of tropical Central
and South America, loricariids have undergone a remarkable diversification of jaw, tooth,
and oral disk morphologies (Fig. 2). Trophic ecological data are generally lacking for
most loricariid species and, where available, are typically qualitative. A review of the
taxonomic and ecological literature yields diet descriptions for 82 species (Chapter 1,
Table 1). Most of these (39 species) are described primarily as detritivores (e.g. Saul,
1975, Kramer and Bryant, 1995, Vaz et al., 1999, Jepsen and Winemiller, 2002, Alvim
and Peret, 2004, Melo et al., 2004, Ferreira, 2007). Other food items reported from
various numbers of loricariid species include wood (19 spp.; Schaefer and Stewart, 1993,
Armbruster, 2003, Nonagaki et al., 2003), algae (11 spp.; Zaret and Rand, 1971, Hood et
al. 2005, Nonogaki et al., 2007, Jepsen and Winemiller, 2007), benthic
macroinvertebrates (8 spp.; M?rona et al., 2008, Forsberg et al., 1993, Winemiller, 1990,
Saul, 1975, Kramer and Bryant, 1995, Melo et al., 2004), mosses and plants (3 spp.; Saul,
1975, Delariva and Agostinho, 2001), fruits and seeds (3 spp.; Armbruster, 2002, 2004,
Melo et al., 2004), and sponges (2 spp.; DeLariva and Agostinho, 2001). The consensus
is that loricariid diets have remained largely restricted to basal resources, a set of trophic
niches frequently associated with derived morphologies (Kramar and Bryant, 1995,
Bellwood, 2003, Konow et al., 2008) and physiologies (Buddington et al., 1987, Choat
and Clements, 1998). Numerous oral and jaw specializations described for Loricariidae
are shared by their sister lineage, Astroblepidae, which has a geographic range restricted
to high elevation hillstream habitats in the Andes. Together, the derived trophic
morphologies shared only by Loricariidae and Astroblepidae support the distinctiveness
88
of their ecological niche, and the assumption that any trophic competition they might face
is greatest from other members of these same families.
Precise descriptions of trophic structure within diverse assemblages of basal
consumers face numerous challenges. Detritivores and herbivores, for example, are
characterized by having fast gut passage rates (Cummins, 1979, Horn, 1992), which limit
the relevance of any given gut sample to a brief period prior to capture. Gut passage
times measured for loricariids range from 40 minutes to four hours, through intestines
that can be >10 times standard body length (Delariva and Agostinho, 2001, Hood et al.,
2005, German, 2008). Most material ingested by detritivores, including substantial
inorganic material, is not assimilated (Yossa and Araujo-Lima, 1998, Bowen et al.,
2006), further limiting the relevance of any given gut content sample, and suggesting that
large sample sizes would be needed for dietary studies to reveal consumer nutrition
(Hyslop, 1980, Bowen, 1983). These problems are exacerbated when patterns are sought
across a large geographic and taxonomic range of consumers in the Neotropics, where
vertebrate taxonomies and that of their algal and microbial food resources are poorly
resolved (Lundberg et al., 2000); and where food availability undergoes strong seasonal
shifts (Lowe-McConnell, 1964, Winemiller, 1990). The dry season, when water levels are
lowest, is the only time when many species can be collected; however, this is also when
food resources are extremely limited, and when many species switch from preferred to
low-quality, detrital foods and/or begin to starve (Lowe-McConnell, 1964, Prejs and
Prejs, 1987, Winemiller, 1990). For these reasons, and for the purpose of quantifying
trophic structure based on the energy or nutritive value of resources consumed, gut
89
content data from loricariids collected in the dry season can be highly imprecise and
potentially misleading.
Attempts to link specific detritivores to distinct subsets of detrital resources using
?
13
C signatures face similar challenges to the differentiation of basal resources (Benedito-
Cecilio et al., 2000, Leite et al., 2002, Jepsen and Winemiller, 2007). In addition, ?
13
C
signatures of detritus can vary seasonally (McCutchan and Lewis, 2001) and variations in
the detrital qualities that are most biologically significant to consumers may not
correspond with variation in their ?
13
C signatures. These challenges largely preclude
robust linkage of detritivorous consumers to subsets of the detrital resource base as
recommended by Hoeinghaus and Zueg (2008), yet isotopes remain a powerful tool for
investigation of trophic structure across detritivore assemblages because of their unique
capacity to quantify time-integrated patterns of resource assimilation (Perga and
Gerdeaux, 2005). The 6 metrics proposed by Layman et al. (2008) offer a means by
which variation in ?
13
C and ?
15
N signatures of sympatric assemblages of detritivores like
the Loricariidae may be analyzed and compared across broad limnological, geographical,
temporal, and taxonomic ranges.
3.2 METHODS
Nineteen geographically, hydrologically, and taxonomically distinct loricariid
assemblages (Table 1) were sampled during expeditions to Venezuela (March 2004,
March 2005), Guyana (October 2005), Peru (August 2006), and Brazil (October 2007;
Fig. 1). Specimens were collected by hand or with seines, gillnets, castnets, or a backpack
electroshocker. Small (<1 g) samples of postdorsal-fin epaxial or postanal-fin hypaxial
90
muscle were excised from specimens in the field and preserved in salt according to
Arrington and Winemiller (2002). Most whole specimens from which tissues were
removed were fixed in 10% formalin and have been deposited in museums in South
America (Brazil: MZUSP; Guyana: UG; Peru: MUSM; Venezuela: MCNG) and the
United States (ANSP, AUM). Many loricariid species collected during these expeditions
were undescribed, and have either recently been described (e.g. Lujan et al., 2009), or are
being described. As a means of differentiation, general descriptors are provided for those
species whose taxonomy is uncertain (e.g. Ancistrus sp. ?wormline?). Generic
assignments within Loricariidae are also often poorly resolved (e.g. Lujan et al., 2009),
and/or poorly correlated with phylogeny (e.g. Armbruster, 2008). Tribes were therefore
used as a proxy for higher phylogenetic relations among species, following the
taxonomies of Armbruster (2004) for subfamily Hypostominae, and Rapp Py-Daniel
(1997) for subfamily Loricariinae.
Salt-preserved tissue samples were rinsed, soaked, dried, and ground according to
the methods of Arrington and Winemiller (2002). Subsamples of each sample were
weighed to 10
?5
g, pressed into Ultra-Pure tin capsules (Costech), and sent for analysis of
stable isotope ratios (
13
C/
12
C and 
15
N/
14
N) to the Analytical Chemistry Laboratory,
Institute of Ecology, University of Georgia, Athens, USA. Encapsulated samples were
dry-combusted (micro Dumas technique) with a Carlo Erba CHN elemental analyzer.
Purified gases (CO
2
 and N
2
) were introduced into a Finnigan Delta C mass spectrometer,
and the isotopic composition was quantified relative to a standard reference material.
Standards were carbon in the PeeDee Belemnite and molecular nitrogen gas in the air.
Results were reported as parts per mille (?) differences from the corresponding standard:
91
?X= [(R
sample
/R
standard
) ? 1] x 10
3
where R = 
15
N/
14
N or 
13
C/
12
C.
Assemblages were defined as all members of the Loricariidae plus Astroblepidae
(the sister family to Loricariidae; only present in three Andean assemblages) sampled
during the same expedition and from the same stream or river channel. Samples from
multiple localities in the same river channel were combined, except in the upper Orinoco
River main channel in southern Venezuela, and in the Siasme River in northern Peru.
Sampling localities in the upper Orinoco main channel were divided into two separate
reaches, respectively dowstream (Orinoco-1) and upstream (Orinoco-2) of the mouth of
the Ventuari River, a major tributary of the upper Orinoco known to have relatively
enriched particulate ?
13
C values relative to the Orinoco River (Tan and Edmund, 1993).
Loricariid/astroblepid assemblages in the Siasme River were also separated into
respective headwater (Siasme-upr) and lower reaches (Siasme-lwr) because sampling
localities were taxonomically and limnologically distinct and separated by several
kilometers.
?
13
C and ?
15
N signatures for all specimens (if >1) of a given species in a given
assemblage were pooled, and mean ?
13
C and ?
15
N values for each species were used to
compute metrics described by Layman et al. (2008). Instead of calculating total area (TA,
metric 3) according to Cornwall et al. (2006), areas were measured empirically from
scatterplots with standardized X and Y-axes. In addition to determining the metrics of
Layman et al. (2008) for species in each assemblage, mean Euclidean distances to
92
centroid (CD) and mean nearest neighbor distances (NND) were computed for the
taxonomic rank of tribe (trCD and trNND, respectively). To calculate trCD, assemblage
centroid remained defined as the species mean, and Euclidean distances from this
centroid to each tribe mean were determined. To calculate trNND, mean Euclidean
distances from each species to its nearest neighbor within its own tribe were calculated.
Differences between tribe and species values were interpreted as a measure of how
phylogenetic rank might influence trophic diversity within loricariid assemblages (cf.
niche conservatism; Peterson et al., 1999, Wiens and Graham, 2005).
3.3 RESULTS
Relationships between loricariid/astroblepid assemblage species richness and the 6
metrics of trophic structure proposed by Layman et al. (2008) are presented in Table 1
and Fig. 3. Metrics 1-3 (NR, CR, and TA; Fig. 3A-C) showed a significant (p<0.001)
positive correlation with richness. As richness increased, trophic niche space of the entire
assemblage increased both vertically, indicating trophic level expansion, and
horizontally, indicating basal resource expansion. Correlation between richness and
centroid distance was also positive (Fig. 3D), albeit less significantly (p=0.09),
suggesting that as species are added to an assemblage, they occupy trophic positions
further from the most common, shared resource. Correlations between richness and NND
and SDNND (Fig. 3E, F) were both negative, although only significantly so (p=0.05) in
the first case. These patterns suggest that niche packing is occurring, and that as species
are added to the assemblage, basal resources are increasingly shared among multiple
species.
93
Plots showing relationships between species richness and CD, NND, and SDNND
(Fig. 3D-F) show considerable variation among species-poor assemblages that appear to
be inversely related to richness. Correlation between richness and variation from
predicted relationships was examined explicitly by regressing absolute values of the
residuals of CD and NND against richness (Fig. 4A, B). Significance of these
relationships (CD: p=0.06; NND: p=0.007) suggest that rules potentially governing the
trophic structure of loricariid assemblages become increasingly strict as richness
increases, and that trophic structure within species-poor assemblages is less predictable.
Relationships between assemblage tribe richness and trCD and trNND likewise
showed considerable diminution in variability with increasing species richness (Fig. 5A,
B), but otherwise showed no significant correlation (p>0.60). When compared with
species metric NND (Fig. 3E), tribe NND (Fig. 5B) was higher in most assemblages (Fig.
6), suggesting that higher level taxa (clades) are more divergent from one another than
lower level taxa.
No divergent patterns were observed between whitewater and clearwater
limnological categories in any of the analyses.
3.4 DISCUSSION
Ricklefs and Travis (1980) first compared communities using metrics similar to
those proposed by Layman et al. (2008; e.g. total niche volume, NND, SDNND), except
the former applied their metrics to ordinations of 8 ecomorphological correlates of niche
space, rather than to direct descriptions of the realized trophic niche (as provided herein
by isotopes; Newsome et al., 2007). Regardless, Ricklefs and Travis (1980) observed
94
similar relationships between richness and niche structure as that observed among
loricariid assemblages in my study. They found that across 11 passerine bird assemblages
in temperate-zone scrub habitats of North and South America, total assemblage
hypervolume increased, and NND and SDNND decreased in direct proportion to number
of species. They interpreted these observations as providing no support for the hypothesis
that density or regularity of species spacing are caused by species interactions (i.e.
competition).
Ricklefs and Travis (1980) remained confident that species are added to
communities in a decidedly non-random manner and they attributed the lack of support
for interspecific interactions provided by their data to two sources of sampling error: (1)
the problem of low sample sizes when data from each whole community contribute only
a single data point, and (2) artifacts of treating all members within a given spatially
defined habitat as equivalent members of the community. In communities of highly
mobile species like birds, the species encountered in any given habitat likely vary in the
degree to which they are optimized for the specific habitat being sampled and, therefore,
the degree to which they regularly interact with other species in that habitat. My study at
least partially corrects for these potential sources of error in the Ricklefs and Travis
(1980) study by increasing overall sample size from 11 to 19 communities, and by
examining taxa that are relatively non-vagile (e.g. Power, 1984). The presence of any
given loricariid species in one of the assemblages defined herein is an indicator that it is a
permanent member of the assemblage, and regularly interacts with other sampled species.
Unfortunately, divergent geochemical and watershed characteristics of the different
habitats sampled in this study contributed new potential sources of variability that
95
Ricklefs and Travis (1980) were able to minimize by restricting their study to similar
(scrub) habitats. Sioli?s (1964, 1967, 1975) classic typology of South American rivers as
black-, clear-, or whitewater according to watershed and geochemical characteristics has
provided a common context for comparative examination of trophic dynamics in
Neotropical rivers (e.g. Fisher, 1979, Bayley, 1981, Rodr?guez and Lewis, 1990,
Henderson and Crampton, 1997, Putz and Junk, 1997, Saint-Paul et al., 2000, Melack and
Forsberg, 2001, Jepsen and Winemiller, 2007). Whitewater rivers are highly turbid, have
relatively high conductance and alkalinity, slight acidity, and carry high suspended and
dissolved loads. In contrast, both black- and clearwaters have very low turbidity and little
to no suspended or dissolved solids. Blackwaters are distinguished from clear by having
high levels of dissolved organic matter, little to no alkalinity, and high acidity. These
limnological characteristics are closely linked to each river?s watershed, with whitewater
rivers draining the geologically younger and rapidly eroding Andean uplands of western
South America; and black- and clearwater rivers draining the highly weathered and
geologically ancient shield areas to the north (Guiana Shield) and south (Brazilian Shield)
of the Amazon Basin. High acidity, alkalinity and DOM of blackwater rivers is due to
their origination in poorly drained, low lying shield areas that promote extended
residence times and the leaching of humic acids. In contrast, most clearwater rivers
originate in well-forested shield uplands. Concentrations of major ions and nutrients in
both black- and clearwater rivers draining shield uplands fall near or even below those
expected in precipitation, indicating minimal contributions from geology and significant
nutrient sequestration by forests (Lewis and Weibezahn, 1981).
Most studies comparing trophic structure within these water types have focused on
96
blackwater versus whitewater river channels or floodplains since these are the dominant
habitat types in the central Amazon Basin. Whitewater habitats have generally been
demonstrated to be more productive, but some studies (e.g. Henderson and Crampton,
1997) have observed either consistently greater fish abundance in blackwater habitats, or
seasonal shifts with respect to which is more productive. In one of the few studies to
examine differences in trophic structure across all three water types using stable isotopes,
Jepsen and Winemiller (2002) concluded that food chain length was higher in nutrient-
poor rivers than in more productive whitewater rivers. These results are questionable
however, because their methods for inferring trophic position and food chain length
assumed a ?
15
N enrichment factor (2.8?) nearly half the rate (5.08?) recently
determined experimentally for the loricariid Pterygoplichthys disjunctivus (German,
2008). Indeed, the avoidance of potentially inter- and intraspecifically variable
enrichment factors is one advantage of using total ?
15
N range (NR) as a measure of
vertical structure rather than attempting to define distances from a standardized baseline.
Regardless, these divergent interpretations of whitewater versus black- or
clearwater systems with respect to productivity and resource diversity make the spectrum
of white- and clearwater habitats sampled in this study a potentially confounding
variable. Niche expansion in response to increases in species richness (Fig. 3A-C) could
be a response to interspecific competition (i.e. niche diversification), or it could simply
reflect a bottom-up response to habitats (whitewater or clearwater) that contain more
resources. A prediction consistent with the latter hypothesis is that assemblages would
assort to trophic space above or below the regression line according to their limnology. If
we were to assume, for example, that whitewater habitats are more productive, and
97
feature a broader diversity of resources, and that an increase in resource diversity or
abundance is the true driver of observed relationships, then we would predict that at any
given species richness, whitewater assemblages would occupy greater niche space than
clearwater habitats. Instead of this pattern (or the opposite pattern in which clearwater
habitats are greater than predicted), assemblages appear to be evenly distributed with
respect to the regression line in all metrics of trophic niche breadth (Fig. 3A-C).
Potential increases in basal resource diversity could also be inferred from the
spectrum of stream orders sampled in this study. All of the sampled regions (Fig. 1) were
in uplands or piedmont zones where stream order is loosely, and inversely correlated with
frequency of disturbance and degree of scour. Patrick (1964), for example, surveyed the
macroinvertebrate fauna of upland stream habitats in Peru and noted the almost complete
absence there of gastropods and trichopteran and plecopteran larvae. One reason she
hypothesized for this was that high suspended and bed loads of these Andean systems
scoured the bottom too intensely to allow persistence of these elements of the
macroinvertebrate fauna. In comparison, large river habitats are relatively stable and
feature of diversity of both autochthonous primary production and allochthonous detrital
resources (e.g. Putz and Junk, 1997, Roelke et al., 2006). An increase in resource
diversity with increase in stream order is a more difficult alternative hypothesis to
attempt to falsify with the given data set. My study concentrated on regions within the
range of Loricariidae known to exhibit relatively high levels of species richness, but
future studies should contrast the results reported herein with data from habitats that are
relatively depauperate in loricariid diversity.
Even if increases in resource diversity and abundance across the sampled taxa are
98
playing a role in the observed expansion of niche space, contemporaneous decreases in
NND (Fig. 3E) suggest that resources are not expanding at quite the same rate as species
are being added to the assemblage, and that individual trophic niches are becoming
increasingly packed. This suggests that loricariids in species rich assemblages may be
beginning to fully occupy the fundamental loricariid trophic niche, and that limited
opportunities exist for loricariids to expand beyond this fundamental niche. Likewise,
decreases in variation from predicted relationships of CD and NND to species richness
(Fig. 4) indicate that with increasing species richness, assemblages approach a limit to the
amount of variability that can be tolerated; that is, the realized niche of loricariids is
expanding to fully occupy their fundamental niche.
Decreases in variability and convergence upon predicted trophic structure observed
at the species level are mirrored at the level of tribe for both CD and NND (Fig. 5),
despite relatively poor correspondence between species richness and tribe richness across
the sampled assemblages (Table 1). Although not significant (p>0.6), possibly due to
considerable variation among tribe poor assemblages, the flat relationship between tribe
richness and inter-tribe spacing as measured by trCD and trNND (Fig. 5), in conjunction
with generally greater inter-tribe spacing than inter-species spacing (Fig. 6), indicates that
taxa at higher taxonomic ranks are more divergent from each other than species are from
each other regardless of rank. This is consistent with niche conservation theory, which
generally supports the conclusion that extent of niche divergence is correlative with
taxonomic rank (Peterson et al., 1999, Wiens and Graham, 2005). That is, families are
more divergent from each other than tribes, which are more divergent from each other
than species.
99
My study has presented the first data suggesting the existence of rules governing
the trophic structure of loricariid assemblages. When viewed in light of the highly
successful models of river function (e.g. river continuum concept, Vannote et al., 1980)
that grew out of a basic understanding of the diversity of functional roles of temperate
detritivores, the importance of a more detailed understanding of trophic diversity in
Loricariidae is magnified. Traditional stream ecological paradigms must be adjusted to
the scale of tropical river systems, whose scale and trophic and taxonomic diveristy is
orders of magnitude larger than temperate zone ecosystems. As an example of scale and
trophic diversity from among detritivores, two species of macroinvertebrate shredders
(the calamoceratid caddisfly Heteroplectron californicum and the elmid Lara avara;
Anderson et al., 1978, Anderson and Sedell, 1979) are known to specialize on large-scale
detrital resources, like tree boles, in north temperate rivers. Among Loricariidae, two
different lineages (Panaque and the Hypostomus cochliodon group) have converged upon
a wood gouging trophic morphology featuring teeth shaped like carpentry instruments
(Fig. 2B). And in the most species rich of the assemblages sampled in this study
(Mara?on, Fig. 14), five species of these wood-eating loricariids coexist. ?
13
C and ?
15
N
signatures and unpublished jaw morphological data from these wood-eaters suggest that
even across such challenging detrital resources, loricariids have diversified.
100
3.5 REFERENCES
Adriaens, D., T. Geerinckx, J. Vlassenbroeck, L. Van Hoorebeke, and A. Herrel. 2009.
Extensive jaw mobility in suckermouth armored catfishes (Loricariidae): a
morphological and kinematic analysis of substrate scraping mode of feeding.
Physiological and Biochemical Zoology 82:51?62.
Alvim, M. C. C., and A. C. Peret. 2004. Food resources sustaining the fish fauna in a
section of the upper S?o Francisco River in Tr?s Marias, MG, Brazil. Brazilian
Journal of Biology 64:195?202.
Anderson, N. H., and J. R. Sedell. 1979. Detritus processing by macroinvertebrates in
stream ecosystems. Annual Review of Entomology 24:351?377.
Anderson, N., J. R. Sedell, L. M. Roberts, and F. J. Triska. 1978. The role of aquatic
invertebrates in processing of wood debris in coniferous forest streams. American
Midland Naturalist 100:64?82.
Araujo-Lima, C. A. R. M., B. R. Forsberg, R. Victoria, and L. Martinelli. 1986. Energy
sources for detritivorous fishes in the Amazon. Science 234:1256?1258.
Arens, W. 1989. Comparative functional morphology of the mouthparts of stream
animals feeding on epilithic algae. Archiv f?r Hydrobiologie 83
(Supplement):253?354.
101
Arens, W. 1994. Striking convergence in the mouthpart evolution of stream-living algae
grazers. Journal of Zoological Systematics and Evolutionary Research
32:319?343.
Armbruster, J. W. 1998. Modifications of the digestive tract for holding air in loricariid
and scoloplacid catfishes. Copeia 1998:663?675.
Armbruster, J. W. 2002. Hypancistrus inspector, a new species of suckermouth armored
catfish (Loricariidae: Ancistrinae). Copeia 2002:86?92.
Armbruster, J. W. 2003. The species of the Hypostomus cochliodon group (Siluriformes:
Loricariidae). Zootaxa 249:1?60.
Armbruster, J. W. 2004. Phylogenetic relationships of the suckermouth armoured
catfishes (Loricariidae) with emphasis on the Hypostominae and the Ancistrinae.
Zoological Journal of the Linnean Society 141:1?80.
Armbruster, J. W. 2008. The genus Peckoltia with the description of two new species and
a reanalysis of the phylogeny of the genera of the Hypostominae (Siluriformes:
Loricariidae). Zootaxa 1822:1?76.
Arrington, D. A., and K. O. Winemiller. 2002. Preservation effects on stable isotope
analysis of fish muscle. Transactions of the American Fisheries Society
131:337?342.
Bayley, P. B. 1981. Fish yield from the Amazon in Brazil: Comparison with African river
yields and management possibilities. Transactions of the American Fisheries
102
Society 110:351?359.
Bayley, P. B. 1989. Aquatic environments in the Amazon Basin, with an analysis of
carbon sources, fish production, and yield. Pages 399?408 in D. P. Dodge, editor.
Proceedings of the International Large River Symposium. Canadian Special
Publication of Fisheries and Aquatic Sciences 106.
Bayley, P. B., and M. Petrere Jr. 1986. Amazon fisheries: assessment methods, current
status and management options. Pages 385?398 in D. P. Dodge, editor.
Proceedings of the International Large River Symposium. Canadian Special
Publication of Fisheries Aquatic Sciences.
Bellwood, D. R. 2003. Origins and escalation of herbivory in fishes: a functional
perspective. Paleobiology 29:71?83.
Benedito-Cecilio, E., and C.A.R.M. Araujo-Lima. 2002. Variation in carbon isotope
composition of Semaprochilodus insignis, a detritivorous fish associated with
oligotrophic and eutrophic Amazonian rivers. Journal of Fish Biology
60:1603?1607.
Benedito-Cecilio, E., C.A.R.M. Araujo-Lima, B.R. Forsberg, M. M. Bittencourt, and L.C.
Martinelli. 2000. Carbon sources of Amazonian fisheries. Fisheries Management
and Ecology 7:305?315.
Bonetto, A., W. Dioni, and C. Pignalberi. 1969. Limnological investigations on biotic
communities in the Middle Paran? River Valley. Verhandlungen des
103
Internationalen Verein Limnologie 17:1035?1050.
Bowen, S. H. 1980. Detrital nonprotein amino acids are the key to rapid growth of
Tilapia in Lake Valencia, Venezuela. Science 207:1216?1218.
Bowen, S. H. 1983a. Detritivory in neotropical fish communities. Environmental Biology
of Fishes 9:137?144.
Bowen, S. H. 1983b. Quantitative description of the diet. Pages 325?336 in L. A. N. D.
L. Johnson, editor. Fisheries Techniques. American Fisheries Society, Bethesda,
MD.
Bowen, S. H. 1984. Detritivory in neotropical fish communities. Pages 59?66 in T. M.
Zaret, editor. Evolutionary Ecology of Neotropical Freshwater Fish. Dr W. Junk
Publishers, The Hague.
Bowen, S. H., A. A. Bonetto, and M. O. Ahlgren. 1984. Microorganisms and detritus in
the diet of a typical neotropical riverine detritivore, Prochilodus platensis (Pisces:
Prochilodontidae). Limnology and Oceanography 29:1120?1122.
Bowen, S. H., B. Gu, and Z. Huang. 2006. Diet and digestion in Chinese mud carp
Cirrhinus molitorella compared with other ilyophagus fishes. Transactions of the
American Fisheries Society 135:1383?1388.
Buddington, R. K., J. W. Chen, and J. M. Diamond. 1987. Genetic and phenotypic
adaptation of intestinal nutrient transport to diet in fish. Journal of Physiology
393:261?281.
104
Choat, J. H., and K. D. Clements. 1998. Vertebrate herbivores in marine and terrestrial
environments: A nutritional ecology perspective. Annual Review of Ecology and
Systematics 29:375?403.
Cross, W. F., J. P. Benstead, A. D. Rosemond, and B. Wallace. 2003. Consumer-resource
stoichiometry in detritus-based streams. Ecology Letters 2003:721?732.
Cummins, K. W. 1994. Invertebrates. Pages 234?250 in P. C. G. Petts, editor. The Rivers
Handbook. Blackwell Scientific Publications, Oxford, UK.
Cummins, K. W. 1973. Trophic relations of aquatic insects. Annual Review of
Entomology 18:183?206.
Cummins, K. W., and M. J. Klug. 1979. Feeding ecology of stream invertebrates. Annual
Review of Ecology and Systematics 10:147?172.
Delariva, R.L., and A. A. Agostinho. 2001. Relationships between morphology and diets
of six neotropical loricariids. Journal of Fish Biology 58:832?847.
Eggers, D. M., N. W. Bartoo, N. A. Rickard, R. E. Nelson, R. C. Wissmar, R. L. Burgner,
and A. H. Devol. 1978. The Lake Washington ecosystem: the perspective from
the fish community production and forage base. Journal of the Fisheries Research
Board of Canada 35:1553?1571.
Ferreira, K. M. 2007. Biology and ecomorphology of stream fishes from the rio Mogi-
Gua?u basin, Southeastern Brazil. Neotropical Ichthyology 5:311?326.
105
Fisher, T. R. 1979. Plankton and primary production in aquatic systems of the Central
Amazon Basin. Comparative Biochemistry and Physiology 62A:31?38.
Flecker, A. S. 1992. Fish trophic guilds and the structure of a tropical stream: weak direct
vs. strong indirect effects. Ecology 73:927?940.
Flecker, A. S. 1996. Ecosystem engineering by a dominant detritivore in a diverse
tropical stream. Ecology 77:1845?1854.
Flecker, A. S., B. W. Taylor, E. S. Bernhardt, J. M. Hood, W. K. Cornwell, S. R. Cassatt,
M. J. Vanni, and N. S. Altman. 2002. Interactions between herbivorous fishes and
limiting nutrients in a tropical stream ecosystem. Ecology 83:1831?1844.
Forsberg, B. R., C. A. R. M. Araujo-Lima, L. A. Martinelli, R. L. Victoria, and J. A.
Bonassi. 1993. Autotrophic carbon sources for fish of the Central Amazon.
Ecology 74:643?652.
Frost, P. C., J. J. Elser, and M. A. Turner. 2002. Effects of caddisfly grazers on the
elemental composition of epilithon in a boreal lake. Journal of the North
American Benthological Society 21:54?63.
Fry, B., and E. B. Sherr. 1984. ?
13
C measurements as indicators of carbon flow in marine
and freshwater ecosystems. Contributions in Marine Science 27:13?47.
Galvis, G., and J. I. Mojica. 2007. The Magdalena River fresh water fishes and fisheries.
Aquatic Ecosystem Health and Management 10:127?139.
106
Geerinckx, T., M. Brunain, A. Herrel, P. Aerts, and D. Adriaens. 2007. A head with a
suckermouth: a functional-morphological study of the head of the suckermouth
catfish Ancistrus cf. triradiatus (Loricariidae, Siluriformes). Belgian Journal of
Zoology 137:47?66.
German, D. P. 2008. Beavers of the fish world: can wood-eating catfishes actually digest
wood? A nutritional physiology approach. Unpublished Ph.D. dissertation.
University of Florida, Gainesville.
Hamilton, S. K., S. J. Sippel, and S. E. Bunn. 2005. Separation of algae from detritus for
stable isotope or ecological stoichiometry studies using density fractionation in
colloidal silica. Limnology and Oceanography: Methods 3:149?157.
Hecky, R.E., and R.H. Hesslein. 1995. Contributions of benthic algae to lake food webs
as revealed by stable isotope analysis. Journal of the North American
Benthological Society 14:631?653.
Henderson, P. A., and W. G. R. Crampton. 1997. A comparison of fish diversity and
abundance between nutrient-rich and nutrient-poor lakes in the Upper Amazon.
Journal of Tropical Ecology 13:175?198.
Hoeinghaus, D. J., and S. C. Zeug. 2008. Can stable isotope ratios provide for
community-wide measures of trophic structure? Comment. Ecology
89:2353?2357.
Hood, J. M., M. J. Vanni, and A. S. Flecker. 2005. Nutrient recycling by two phosphorus
107
rich grazing catfish: the potential for phosphorus-limitation of fish growth.
Oecologia 146:247?257.
Horn, M. H., and K. S. Messer. 1992. Fish guts as chemical reactors: a model of the
alimentary canals of marine herbivorous fishes. Marine Biology 113:527?535.
Hyslop, E.J. 1980. Stomach content analysis-a review of methods and their application.
Journal of Fish Biology 17:411?429.
Jepsen, D. B., and K. O. Winemiller. 2002. Structure of tropical river food webs revealed
by stable isotope ratios. Oikos 96:46?55.
Jepsen, D. B., and K. O. Winemiller. 2007. Basin geochemistry and isotopic ratios of
fishes and basal production sources in four neotropical rivers. Ecology of
Freshwater Fish 16:267?281.
Konow, N., D. R. Bellwood, P. C. Wainwright, and A. M. Kerr. 2008. Evolution of novel
jaw joints promote trophic diversity in coral reef fishes. Biological Journal of the
Linnean Society 93:545?555.
Kramer, D. L., and M. J. Bryant. 1995. Intestine length in the fishes of a tropical stream:
2. Relationships to diet ? the long and short of a convoluted issue. Environmental
Biology of Fishes 42:129?141.
Layman, C. A., and D. M. Post. 2008. Can stable isotope ratios provide for community-
wide measures of trophic structure? Reply. Ecology 89:2358?2359.
108
Layman, C. A., D. A. Arrington, C. G. Monta?a, and D. M. Post. 2007. Can stable
isotope ratios provide for community-wide measures of trophic structure. Ecology
88:42?48.
Layman, C. A., K. O. Winemiller, and D. A. Arrington. 2005. Describing a species-rich
river food web using stable isotopes, stomach contents, and functional
experiments. Pages 395?406 in P.C. de Ruiter, V. Wolters, and J. C. Moore,
editors. Dynamic Food Webs: Multispecies Assemblages, Ecosystem
Development and Environmental Change. Elsevier, Amsterdam.
Leite, R.G., C.A.R.M. Araujo-Lima, R.L. Victoria, and L.A Martinelli. 2002. Stable
isotope analysis of energy sources for larvae of eight fish species from the
Amazon floodplain. Ecology of Freshwater Fish 11:56?63.
Lewis, W. M., Jr., and F. Weibezahn. 1981. The chemistry and phytoplankton of the
Orinoco and Caroni Rivers, Venezuela. Archiv f?r Hydrobiologie 91:521?528.
Lewis, W. M., Jr., S. K. Hamilton, M. Rodr?guez, J. F. Saunders, III, and M. A. Lasi.
2001. Foodweb analysis of the Orinoco floodplain based on production estimates
and stable isotope data. Journal of the North American Benthological Society
20:241?254.
Lowe-McConnell, R. H. 1964. The fishes of the Rupununi savanna district of British
Guiana, South America: Part 1. Ecological groupings of fish species and effects of
the seasonal cycle on the fish. Zoological Journal of the Linnean Society
45:103?144.
109
Lujan, N. K., M. Arce, and J. W. Armbruster. 2009. A new black Baryancistrus with blue
sheen from the upper Orinoco (Siluriformes: Loricariidae). Copeia 2009:50?56.
Lundberg, J. G., M. Kottelat, G. R. Smith, M. L. J. Stiassny, and A. C. Gill. 2000. So
many fishes, so little time: an overview of recent ichthyological discovery in
continental waters. Annals of the Missouri Botanical Garden 87:26?62.
Maltby, L. 1994. Detritus processing. Pages 331?353 in P. C. G. Petts, editor. The Rivers
Handbook. Blackwell Scientific Publications, Oxford, UK.
Mann, K. H. 1988. Production and use of detritus in various freshwater, estuarine, and
coastal marine ecosystems. Limnology and Oceanography 33:910?930.
Mann, K. H. 1972. Introductory remarks. Pages 13?16 in IBP-UNESCO Symposium on
Detritus and its role in aquatic ecosystems. Memorie dell'Istituto Italiano
Idrobiologia, Pallanza, Italy.
McCutchan, J. H., Jr., and W. M. Lewis, Jr. 2001. Seasonal variation in stable isotope
ratios of stream algae. Verhandlungen des Internationalen Verein Limnologie
27:3304?3307.
McIntyre, P. B., L. E. Jones, A. S. Flecker, and M. J. Vanni. 2007. Fish extinctions alter
nutrient recycling in tropical freshwaters. Proceedings of the National Academy
of Science 104:4461?4466.
Melack, J. M., and B. R. Forsberg. 2001. Biogeochemistry of Amazon floodplain lakes
and associated wetlands. Pages 235?274 in R. L. V. Michael E. McClain, Jeffrey
110
E. Richey, editor. The Biogeochemistry of the Amazon Basin. Oxford University
Press, New York.
Melo, C. E. de, F. de Arruda Machado, and V. Pinto-Silva. 2004. Feeding habits of fish
from a stream in the savanna of Central Brazil, Araguaia Basin. Neotropical
Ichthyology 2:37?44.
M?rona, B. de, B. Hugueny, and F. L. Tejerina-Garro. 2008. Diet-morphology
relationship in a fish assemblage from a medium-sized river of French Guiana: the
effect of species taxonomic proximity. Aquatic Living Resources 21:171?184.
Moe, S. J., R. S. Stelzer, M. R. Forman, W. S. Harpole, T. Daufresne, and T. Yoshida.
2005. Recent advances in ecological stoichiometry: insights for population and
community ecology. Oikos 109:29?39.
Mol, J. H., and P. E. Ouboter. 2004. Downstream effects of erosion from small-scale gold
mining on the instream habitat and fish community of a small Neotropical
rainforest stream. Conservation Biology 18:201?214.
Moore, J. C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C.
Johnson, K. S. McCann, K. Melville, P. J. Morin, K. Nadelhoffer, A. D.
Rosemond, D. M. Post, J. L. Sabo, K. M. Scow, M. J. Vanni, and D. H. Wall.
2004. Detritus, trophic dynamics and biodiversity. Ecology Letters 7:584?600.
Nachi, A. Massao, F. J. Hernandez-Blazquez, R. L. Barbieri, R. G. Leite, S. Ferri, and M.
T. Phan. 1998. Intestinal histology of a detritivorous (iliophagus) fish Prochilodus
111
scrofa (Characiformes, Prochilodontidae). Annales des Sciences Naturelles
2:81?88.
Newsome, S. D., C. M. del Rio, S. Bearhop, and D. L. Phillips. 2007. A niche for isotopic
ecology. Frontiers in Ecology and the Environment 5:429?436.
Nonogaki, H., J. A. Nelson, and W. P. Patterson. 2007. Dietary histories of herbivorous
loricariid catfishes: evidence from ?
13
C values of otoliths. Environmental Biology
of Fishes 78:13?21.
Odum, W. E. 1970. Utilization of the direct grazing and plant detritus food chains by the
striped mullet Mugil cephalus. Pages 222?240 in J. H. Steele, editor. Marine Food
Chains. Cambridge University Press, Cambridge, UK.
Ono, R. D. 1980. Fine structure and distribution of epidermal projections associated with
taste buds on the oral papillae in some loricariid catfishes (Siluroidei:
Loricariidae). Journal of Morphology 164:139?159.
Patrick, R. 1964. A discussion of the results of the Catherwood Expedition to the
Peruvian headwaters of the Amazon. Verhandlungen des Internationalen Verein
Limnologie 15:1084?1090.
Perga, M.E., and D. Gerdeaux. 2005. 'Are fish what they eat' all year round? Oecologia
144:598?606.
Peterson, A. T., J. Sober?n, and V. S?nchez-Cordero. 1999. Conservatism of ecological
niches in evolutionary time. Science 285:1265?1267.
112
Peterson, B. J., and B. Fry. 1987. Stable isotopes in ecosystem studies. Annual Review of
Ecology and Systematics 18:293?320.
Post, D. 2002. Using stable isotopes to estimate trophic position: models, methods, and
assumptions. Ecology 83:703?718.
Power, M. E. 1984a. Habitat quality and the distribution of algae-grazing catfish in a
Panamanian stream. Journal of Animal Ecology 53:357?374.
Power, M. E. 1984b. The importance of sediment in the grazing ecology and size class
interactions of an armored catfish, Ancistrus spinosus. Environmental Biology of
Fishes 10:173?181.
Prejs, A., and K. Prejs. 1987. Feeding of tropical freshwater fishes: seasonality in
resource availability and resource use. Oecologia 71:397?404.
Putz, R., and W. J. Junk. 1997. Phytoplankton and periphyton. Pages 207?222 in W. J.
Junk, editor. The Central Amazon Floodplain: Ecology of a Pulsing System.
Springer, New York.
Quir?s, R., and C. Baig?n. 1985. Fish abundance related to organic matter in the Plata
River Basin, South America. Transactions of the American Fisheries Society
114:377?387.
Reis, R. E., S. O. Kullander, and C. J. Ferraris, Jr. 2003. Check List of the Freshwater
Fishes of South and Central America. EDIPUCRS, Porto Alegre.
113
Ricklefs, R. E., and J. Travis. 1980. A morphological approach to the study of avian
community ecology. Auk 97:321?338.
Roberts, T. R. 1973. Osteology and relationships of the Prochilodontidae, a South
American family of characoid fishes. Bulletin of the Museum of Comparative
Zoology 145:213?235.
Rodr?guez, M. A., and W. M. Lewis, Jr. 1990. Diversity and species composition of fish
communities of Orinoco floodplain lakes. National Geographic Research
6:319?328.
Rodr?guez, M. A., K. O. Winemiller, W. M. Lewis, Jr., and D. C. Taphorn B. 2007. The
freshwater habitats, fishes, and fisheries of the Orinoco River basin. Aquatic
Ecosystem Health and Management 10:140?152.
Roelke, D. L., J. B. Cotner, J. V. Montoya, C. E. Del Castillo, S. E. Davis, J. A. Snider,
G. M. Gable, and K. O. Winemiller. 2006. Optically determined sources of
allochthonous organic matter and metabolic characterizations in a tropical
oligotrophic river and associated lagoon. Journal of the North America
Benthological Society 25:185?197.
Roman, M. R., and K. R. Tenore. 1984. Detritus dynamics in aquatic ecosystems: an
overview. Bulletin of Marine Science 35:257?260.
Rossi, L. M. 1992. Evolucion morfologica del aparato digestivo de postlarvas y
prejuveniles de Prochilodus lineatus (Val., 1847) (Pisces, Curimatidae) y su
114
relacion con la diet. Revue d'Hydrobiologie Tropicale 25:159?167.
Saint-Paul, U., J. Zuanon, M. A. V. Correa, M. Garcia, and N. N. Fabr?. 2000. Fish
communities in central Amazonian white- and blackwater floodplains.
Environmental Biology of Fishes 57:235?250.
Saul, W. G. 1975. An ecological study of fishes at a site in upper Amazonian Ecuador.
Proceedings of the Academy of Natural Sciences of Philadelphia 127:93?134.
Schaefer, S. A., and D. J. Stewart. 1993. Systematics of the Panaque dentex species
group (Siluriformes: Loricariidae), wood-eating armored catfishes from tropical
South America. Ichthyological Exploration of Freshwaters 4:309?342.
Schaefer, S. A., and G. V. Lauder. 1986. Historical transformation of functional design:
evolutionary morphology of feeding mechanisms in loricarioid catfishes.
Systematic Zoology 35:489?508.
Schaefer, S. A., and G. V. Lauder. 1996. Testing historical hypotheses of morphological
change: biomechanical decoupling in loricarioid catfishes. Evolution
50:1661?1675.
Sioli, H. 1964. General features of the limnology of Amazonia. Verhandlungen des
Internationalen Verein Limnologie 15:1053?1058.
Sioli, H. 1967. Studies in Amazonian waters. Atas do Simp?sio s?bre a Biota Amaz?nica
3:9?50.
115
Sioli, H. 1975. Tropical rivers as expressions of their terrestrial environments. Pages 275-
288 in F. B. Golly and E. Medina, editors. Tropical Ecological Systems. Springer-
Verlag, New York, NY.
Sioli, H. 1984. The Amazon and its main affluents: Hydrography, morphology of the
river courses, and river types. Pages 127?165 in H. Sioli, editor. The Amazon.
Limnology and landscape ecology of a mighty tropical river and its basin. Dr W.
Junk Publishers, Dordrecht.
Tan, F. C., and J. M. Edmond. 1993. Carbon isotope geochemistry of the Orinoco Basin.
Estuarine, Coastal and Shelf Science 36:541?547.
Thompson, D. R., S. J. Bury, K. A. Hobson, L. I. Wassenaar, and J. P. Shannon. 2005.
Stable isotopes in ecological studies. Oecologia 144:517?519.
Vannote, R. L., G. W. Minshall, K. W. Cummings, J. R. Sedell, and C. E. Cushing. 1980.
The river continuum concept. Canadian Journal of Fisheries and Aquatic Sciences
37:130?137.
Vari, R. P. 1983. Phylogenetic relationships of the families Curimatidae,
Prochilodontidae, Anostomidae, and Chilodontidae (Pisces: Characiformes).
Smithsonian Contributions to Zoology 378:1?60.
Vaz, M. M., M. Petrere Jr., L. A. Martinelli, and A. A. Mozeto. 1999. The dietary regime
of detritivorous fish from the river Jacar? Pepira, Brazil. Fisheries Management
and Ecology 6:121?132.
116
Wetzel, R. G., P. H. Rich, M. C. Miller, and H. L. Allen. 1972. Metabolism of dissolved
and particulate detrital carbon in a temperate hard-water lake. Memorie
dell'Instituto italiano di Idrobiologia 29(supplement):185?244.
Wiens, J. J., and C. H. Graham. 2005. Niche conservatism: integrating evolution,
ecology, and conservation biology. Annual Review of Ecology and Systematics
36:519?539.
Winemiller, K., C. Marrero, and D. Taphorn. 1996. Perturbaciones causadas por el
hombre a las poblaciones de peces de los llanos y del piedemonte Andino de
Venezuela. BioLlania 12:13?48.
Winemiller, K. O. 1990. Spatial and temporal variation in tropical fish trophic networks.
Ecological Monographs 60:331?367.
Yamaoka, K. 1982. Morphology and feeding behaviour of five species of genus
Petrochromis (Teleostei, Cichlidae). Physiology and Ecology Japan 19:57?75.
Yossa, M. I., and C. A. R. M. Araujo-Lima. 1998. Detritivory in two Amazonian fish
species. Journal of Fish Biology 52:1141?1153.
Zaret, T. M., and A. S. Rand. 1971. Competition in tropical stream fishes: support for the
competitive exclusion principle. Ecology 52:336?342.
117
Table 1. 19 loricariid assemblages sampled in this study, their taxonomic diversity, and
results of community-wide stable isotope analyses proposed by Layman et al. (2008).
NN
 speciestribesNRCRTACDNNDSDNNDtrCDtrNND
Shaapan2 2 0.50.90.00.51.00.00.51.0
Huancabamba3 2 1.04.70.82.41.82.21.34.4
Peixoto3 2 2.41.30.11.21.01.51.40.1
Siasme-lwr3 2 1.83.80.81.71.90.61.71.6
Siasme-upr3 2 0.80.60.00.50.50.20.30.7
Almendro5 4 2.05.33.31.51.41.41.71.1
Bununi5 5 0.82.61.00.90.60.50.80.6
Utcubamba5 3 1.43.92.81.81.31.00.92.1
Takutu6 3 3.33.27.51.81.40.81.01.9
Casiquiare7 3 3.62.15.51.41.00.91.60.5
Essequibo7 5 3.04.87.11.61.00.81.51.3
Nieva7 4 1.64.01.31.60.70.51.11.7
Jamanxim113 2.74.34.21.20.60.40.61.2
Orinoco-1112 2.14.63.81.00.60.51.50.5
Orinoco-2113 3.16.09.41.71.01.02.41.0
Rupununi124 3.66.814.91.80.90.61.32.1
Curu?144 4.05.112.21.90.70.50.62.2
Ventuari143 3.65.812.11.70.90.61.20.9
Mara?on166 3.05.910.91.60.80.71.01.6
118
Figure 1. Drainage map of northern South America with regions sampled in this study marked by dotted ovals. Region 1
(Venezuela) included the Casiquiare Assemblage (7 spp.), the Orinoco-1 Assemblage (11 spp.), the Orinoco-2 Assemblage (11
spp.), and the Ventuari Assemblage (14 spp.). Region 2 (Guyana) included the Bununi Assemblage (5 spp.), Rupununi
Assemblage (12 spp.), and the Takutu Assemblage (6 spp.). Region 3 (Peru) included the Almendro Assemblage (5 spp.), the
Huancabamba Assemblage (3 spp.), the Mara?on Assemblage (16 spp.), the Nieva Assemblage (7 spp.), the Siasme-lwr
Assemblage (3 spp.), the Siasme-upr Assemblage (3 spp.), and the Utcubamba Assemblage (5 spp.). Region 4 (Brazil) included
the Peixoto Assemblage (3 spp.), the Jamanxim Assemblage (11 spp.), and the Curu? Assemblage (14 spp.).
118
119
Figure 2. Representative diversity of loricariid upper and lower jaw morphologies: A.
Leporacanthicus, B. Panaque, C. Chaetostoma.
120
Figure 3. Relationships between loricariid assemblage species richness and six metrics of
trophic structure proposed by Layman et al. (2008; see Introduction): (A) NR, (B) CR,
and (C) TA are indicators of whole assemblage trophic niche breadth, whereas (D) CD,
(E) NND, and (F) SDNND are indicators of species niche breadth and spacing within
each assemblage. Squares represent clearwater shield habitats and circles represent
whitewater Andean habitats.
121
Figure 4. Relationships between loricariid assemblage species richness and the absolute
values of (A) CD and (B) NND residuals. Squares represent clearwater shield habitats
and circles represent whitewater Andean habitats.
122
Figure 5. Relationships between loricariid assemblage tribe richness and (A) CD
calculated as the mean distance from the species centroid to each tribe mean (trCD); and
(B) NND calculated as the mean distance from each species to its nearest neighbor within
its own tribe (trNND). Squares represent clearwater shield habitats and circles represent
whitewater Andean habitats.
123
Figure 6. Comparison of species versus tribe NND for each assemblage sampled in this
study. Assemblages in ascending order of species richness from left to right.
124
Figure 7. ?
15
N and ?
13
C relationships of loricariid plus astroblepid consumers in (A)
Shaapan creek, Mara?on (Amazon) drainage, northern Peru; (B) Peixoto river, Tapaj?s
(Amazon) drainage, Brazil; (C) upper Siasme creek, Mara?on (Amazon) drainage,
northern Peru; and (D) Bununi creek, Rupununi (Essequibo) drainage, Guyana. Sample
sizes in parentheses.
125
Figure 8. ?
15
N and ?
13
C relationships of loricariid consumers in (A) the Utcubamba
river, Mara?on (Amazon) drainage, northern Peru; (B) Casiquiare canal, Negro
(Amazon) drainage, southern Venezuela; and (C) Takutu river, Branco (Amazon)
drainage, Guyana. Sample sizes in parentheses.
126
Figure 9. ?
15
N and ?
13
C relationships of loricariid plus astroblepid consumers in (A) the
Huancabamba river, and (B) lower Siasme creek, both Mara?on (Amazon) drainages in
northern Peru. Sample sizes in parentheses.
127
Figure 10. ?
15
N and ?
13
C relationships of loricariid plus astroblepid consumers in (A)
Almendro creek, and (B) the Nieva river, both Mara?on (Amazon) drainages in northern
Peru. Sample sizes in parentheses.
128
Figure 11. ?
15
N and ?
13
C relationships of loricariid consumers in the (A) Essequibo
river, Guyana, and (B) the Jamanxim river, Tapajos (Amazon) drainage, Brazil. Sample
sizes in parentheses.
129
Figure 12. ?
15
N and ?
13
C relationships of loricariid consumers in the Orinoco river main
channel, (A) below and (B) above the mouth of the Ventuari river. Sample sizes in
parentheses.
130
Figure 13. ?
15
N and ?
13
C relationships of loricariid consumers in (A) the Rupununi river,
Essequibo drainage, Guyana; and (B) the Curu? river, Xingu drainage, Brazil. Sample
sizes in parentheses.
131
Figure 14. ?
15
N and ?
13
C relationships of loricariid consumers in (A) the Ventuari river,
Orinoco drainage, southern Venezuela; and (B) the Mara?on river, Amazon drainage,
northern Peru. Sample sizes in parentheses.
132
CHAPTER 4. GEOLOGICAL AND HYDROLOGICAL HISTORY OF THE
GUIANA SHIELD, AND HISTORICAL BIOGEOGRAPHY OF ITS FISHES
4.1 SUMMARY
The Guiana Shield is a geologically ancient region of high to middle elevation
terrain covering approximately 2,288,000 km
2
 of northern South America, from French
Guiana west across northern Brazil, Suriname, Guyana, southern Venezuela, and eastern
Colombia. The Guiana Shield is drained by most of the major rivers of South America
including upper and right bank tributaries of the Orinoco River (including the Ventuari,
Caura, and Caroni), upper and left bank tributaries of the Essequibo River (including the
Rupununi, Potaro, and Mazaruni/Cuyuni), upper and left bank tributaries of the Negro
River (including the Branco, Uraricoera, Takutu, and Ireng), left bank tributaries of the
lower Amazon River (including the Uatuma, Trombetas, Paru do Oeste, Paru, and Jari),
and an east-to-west series of Atlantic coastal rivers north of the Amazon including the
Oyapock, Marone, Coppename, and Corantijne. Topography of the Guiana Shield is
largely the result of approximately six periods of non-deformational uplift and tilting
since at least the Early Cretaceous, each uplift period followed by a period of drainage
reorganization, erosion, and downcutting. High elevation parts of the shield are now
concentrated like an east-west ridge down its center. Elevations are particularly high in
the west, in southern Venezuela, where a concentrated region of isolated massifs or
133
cerros over 2000 m-asl is known as Pantepui or the Guayana Highlands. Sediments from
erosion of the Guiana Shield have been distributed in all directions around the highlands,
creating peneplanar savannas to the northwest, west, and south, and coastal lowlands to
the northeast and east.
Headwaters of most Guiana Shield rivers interdigitate throughout the highlands,
and the biogeographic patterns of Guiana Shield fishes indicate a strong historical
influence of river capture both in the highlands and the lowlands. The Guiana Shield has
an ancient history of cyclical uplift followed by drainage reorganization, which has
allowed for an abundance of both recent vicariant speciation and relictual persistence of
narrowly endemic basal lineages. Vicariant speciation and relictual persistence is
particularly evident among the suckermouth armored catfishes (family Loricariidae), a
radiation of over 700 species broadly distributed across tropical Central and South
America. Loricariid assemblages in the Guiana Shield are particularly species rich and
they are frequent occupants of upland habitats, making them an ideal group in which to
examine the biogeographic importance of modern topographic patterns and historical
geological events. Several genera in the loricariid tribes Hartiini (Harttia) and Ancistrini
(Pseudancistrus, Pseudacanthicus, Leporacanthicus) are shared between the Guiana and
Brazilian Shields, but are absent from highlands of the Andes. Several other ancistrin
genera (Exastilithoxus, Lithoxus, Neblinichthys, New Genus 2) have ranges restricted
only to uplands of the Guiana Shield, where their distributions across isolated and
restricted ranges serves as important clues to the historical continuity of many currently
disconnected headwaters.
134
At the center of the Guiana Shield, until at least the Plio-Pleistocene, a river called
the proto-Berbice is hypothesized to have united modern headwaters of the Orinoco,
Branco and Essequibo into a single paleodrainage that exited into the Atlantic near the
modern border between Guyana and Suriname. Progressive stream capture by the
Orinoco, Amazon, and Essequibo led to modern drainage patterns. Highlands of the
Guiana Shield also appear to have functioned as an incubator and source for upland
specialized taxa that have radiated into more recently uplifted terrains along the Andean
flanks. Phylogenetic analysis of morphological data indicate that a predominantly
Andean upland radiation of four ancistrin genera (Chaetostoma, Cordylancistrus,
Dolichancistrus, Leptoancistrus) is nested within a clade of Guiana Shield endemic
genera (Exastilithoxus, Lithoxus, New Genus 2). Morphological data likewise indicate
that Lithogenes, which is currently or recently known only from the Guiana Shield and
the nearby Coastal Mountain range of northern Venezuela, is either sister to the Andean
upland genus Astroblepus (family Astroblepidae, sister to Loricariidae), or sister to all
remaining Loricariidae.
In addition to historical geologic events and basal radiations, distributional
patterns of several widespread fish groups support the broad taxonomically significance
of several modern, permanent or seasonal connections between major drainage basins.
The network of major modern and historical avenues for dispersal between drainages
conforms, roughly, to a prone number 8 laid atop the Guiana Shield. Major modern
lowland connections among Guiana Shield rivers include: the Casiquiare Canal, linking
the upper Orinoco with the upper Negro; the seasonally flooded Rupununi Savannas,
linking the Essequibo with the Branco; Atlantic coastal connections consisting of flooded
135
lowlands and freshwater plumes from the Amazon mouth northwest to the Orinoco
mouth; and both southern and northern tributaries of the lower Amazon, which drain
respective northern and southern slopes of the Brazilian and Guiana Shields. High
resolution phylogenetic hypotheses and fish distributional data that might provide greater
resolution to the rank importance of these drainage connections are generally lacking.
4.2 INTRODUCTION
Highland areas that serve as sources and boundaries for the great rivers of South
America can be broadly divided into two categories based on their geologic age and
origin. As reviewed elsewhere in this volume (Chapters 15 and 16), the allochthonous
terrains and massive crustal deformations of the Andes Mountains that comprise the
extremely high-elevation western margin of South America have their origins in
diastrophic (distortional) tectonic activity largely limited to the Late Paleogene and
Neogene (<25 Ma; Gregory-Wodzicki, 2000). In contrast, vast upland regions across
much of the interior of the continent have been relatively tectonically quiescent since the
Proterozoic (>550 Ma; Gibbs and Baron, 1993) and exhibit a topography that is instead
largely the result of epeirogenic (non-distortional) uplift of the Guiana and Brazilian
Shields and subsequent erosion of overlying sedimentary formations.
Topographic and hydrologic evolution of both the Andes and the Amazon
Platform advanced within the late Mesozoic to Cenozoic timeframe recognized as largely
encompassing the evolutionary radiations of Neotropical fishes (Lundberg et al., 1998);
however, early uplifts of the Amazon Platform predate significant Andean orogeny by
several hundred million years. Lundberg (1998), in his review of the temporal context for
136
diversification of Neotropical freshwater fishes, made it clear that, despite the prevailing
attention given to Andean orogeny and the various vicariant speciation events that it
spawned, most major Neotropical fish lineages were already extant long before the
Miocene surge in Andean uplift, and the search for geologic events relevant to basal
nodes in the evolutionary history of Neotropical fish lineages should extend deeper in
time.
In this chapter we describe the geologic, topographic, and hydrologic evolution of
the Guiana Shield since at least the Cretaceous. We then compare these historical
processes with evolution of the region?s fishes. The primary taxonomic focus of this
chapter is on suckermouth armored catfishes (Loricariidae), due to their great diversity,
comprising over 700 described species, their ancient ancestry as part of a superfamily
sister to all other Siluriformes, and their biogeographic tractability due to distributions
across headwater habitats and associated allopatric distribution patterns among sister
taxa. We conclude that the diverse loricariid fauna of the Guiana Shield accumulated
gradually over tens of millions of years over the whole continent, and not as the result of
a rapid, geographically restricted adaptive radiation. We demonstrate the role of the
Guiana and Brazilian shields as ancient highland regions in the origin of frequently
rheophilic loricariid taxa. We also show how diversification was influenced by a
restricted number of landscape scale features; especially dispersal and vicariance across
several geologically persistent corridors, expansion and contraction of ranges due to
tectonic alterations in prevailing slope, and patterns of local and regional climate change.
Continued progress in this area will require increased sampling, especially in the southern
and western portions of the Guiana Shield, both to more fully understand the alpha
137
taxonomy and distribution of species, and for the reconstruction of detailed species-level
phylogenies.
4.3 GEOLOGY AND HYDROLOGY
4.3.1 Overview
Surficial outcrops of the Amazon Platform can be observed as bedrock shoals in
many northern and southern tributaries of the Amazon River, but rarely at elevations
higher than 150 meters above sea level (m-asl). Topography higher than this is largely
comprised by the Roraima Group, an aggregation of fluviolacustrine sediments deposited
over much of the northern Amazon Platform during the Proterozoic and subsequently
uplifted along with the basement. Portions of this formation resistant to erosion now
comprise most of the striking topographic elements for which the shield regions are
famous, including the fabled Mount Roraima (2810 m-asl) and South America?s highest
non-Andean peak, Pico Neblina (3014 m-asl) at the frontier with Brazil in the
southwestern corner of Amazonas State, Venezuela. The relatively recent discovery of
Pico Neblina in the mid-twentieth century illustrates both the long standing
inaccessibility of much of the Guiana Shield, and the tremendous gaps in knowledge that
still challenge summaries of shield geology and biogeography.
Separating the Guiana Shield from the Brazilian Shield is the Amazon Graben, a
structural downwarp underlying the Amazon Basin. This major divide is 300 to 1000 km
wide (from North-South) and is filled with sediments up to 7000 m deep. South of the
138
Amazon Graben to about 20?S latitude, stretches the larger of the Amazon Platform?s two
subunits: the Brazilian (or Guapor?) Shield, whose highlands delineate watershed
boundaries of the major southern Amazon River tributaries Tocantins, Tapajos and
Xingu, as well as northwestern headwaters of the south-flowing Parana River. Middle
reaches of the Madeira River are also interrupted by several major rapids due to their
transect of a western arm of the Brazilian Shield.
The Guiana Shield, the smaller, more northern subunit of the Amazon Platform, is
elongated nearly east to west and roughly oval in shape (Fig. 1). From its eastern margin
along the Atlantic coast, it stretches across French Guiana, Suriname, Guyana, and
Venezuela, to southeastern Colombia in the west (approximately 2000 km distance).
Bounded by the Amazon Basin to its south, and the Orinoco River to its north
(approximately 1000 km distance) and west, the Guiana Shield occupies some 2,288,000
km
2
 (Hammond, 2005). The average elevation of the Guiana Shield is approximately 270
m-asl but disjunct and frequently shear-sided formations exceeding 2000 m-asl, known
variously as tepuis, cerros, massifs, sierras, and inselbergs are common, particularly near
Venezuela?s frontier with Brazil in a region of concentrated high elevation terrain known
as Pantepui. The Pantepui region slopes more or less gently to the north but has a striking
southern scarp boundary along the Venezuela-Brazil border. Ridges along this border
comprise the Sierras Pakaraima and Parima, which stretch some 800 km ENE?WSW and
rarely drop below 1000 m-asl. The Pakaraima and Parima ranges have their eastern origin
in Mount Roraima at the tricorners between Guyana, Brazil, and Venezuela, and their
western terminus in Sierra Neblina.
139
The name ?Guiana? is believed to be derived from an Amerindian word meaning
?water? or ?many waters? (Hammond, 2005). Indeed, as many as 47 medium to large
rivers drain the greater Guiana Shield region (Fig. 1), including the Negro, Orinoco,
Essequibo, Trombetas, Caqueta (Japur?), Jatapu, Marone (Marowijne), and Corentyne
(Correntijne). Discharge from rivers draining or traversing the Guiana Shield totals an
estimated average of 2,792 km
3
 per year, which amounts to approximately a quarter of
South America?s total volume of freshwater exported to the oceans (Hammond, 2005).
This volume of water carries with it considerable erosive power, which, with sporadic
periods of epeirogenic uplift, are primary forces responsible for the region?s remarkable
topography.
4.3.2 Topographic Evolution
Granitic basement rocks that comprise most of the Amazon Platform formed
during orogenic events of the Paleoproterozoic (1,700-2,200 Ma), although the Imataca
Complex of northeastern Venezuela is exceptional for its Archean age (>2,500 Ma). For
much of this time, it is hypothesized based on once-contiguous fault lines that the
Amazon Platform was united with the West African craton, and that together they were
part of a single tectonic plate forming parts of the supercontinents Gondwana, Pangea,
and Columbia. Approximately 1,800 Ma, a major orogenic episode somewhere to the east
and north of the Guiana Shield in what would have been the Supercontinent Columbia,
turned what is now the shield into a foreland basin and depositional zone (Santos et al.,
2003). Over the course of a few hundred million years the northern Amazon Platform
accumulated up to over 3,000 m (avg. 500 m; Gansser, 1974) of sediment from rivers
140
flowing off of this ancient mountain range into fluvio-deltaic and lacustrine environments
(Edmond et al. 1995). The resulting sandstone formations, known as the Roraima Group,
feature ripple marks and rounded pebbles indicating their fluvial origin and the original
east to west direction of deposition (Gansser, 1954; Ghosh, 1985). Now uplifted at least
3000 m and constituting highlands throughout the Guiana Shield, these sediments still
cover a vast area but are much reduced from their original range, which surpassed
2,000,000 km
2
 and stretched about 1,500 km from an eastern origin in or near Suriname
(largely exclusive of French Guiana) to Colombia and across northern Brazil. Eastern
Roraima formations such as the Tafelberg in east-central Suriname and Cerro Roraima
itself are older and deeper than western Roraima sediments now evident as shallow
sandstone caps of the central Colombian Macarena and Garzon massifs, and the
southeastern Colombian mesas of Inirida, Mapiripan, and Yambi (Gansser, 1974).
Transition from the once contiguous, fluvially deposited Roraima formation to the
now disjunct Guiana Shield highlands required loss of an enormous volume of
intervening sediment via erosion. The modern highlands constitute approximately
200,000 km
3
 of comparatively resistant sediment, but this is a small remnant of what was
originally an approximately 1,000,000 km
3
 formation averaging approximately 500 m in
depth (Gansser, 1974). Erosional redistribution of Roraima Group sediments, along with
younger Andean sediment, into structural basins encircling the shield has created
peneplainer savannas north, west, and south of the highlands. To the north, the structure
of the flat Eastern Venezuelan Llanos is that of a basin filled with sediments over 12 km
deep (Hedberg, 1950). This Eastern Venezuela Basin is narrowly contiguous with the
Apure-Barinas back arc basin underlying the Apure Llanos northwest of the Guiana
141
Shield (see Eastern Venezuela Basin below). Around the western side of the Guiana
Shield, the Apure-Barinas basin and a back arc basin underlying the Colombian Llanos
just southeast of the Andes are contiguous with a low-lying cratonic subduction or suture
zone approximately coincident with the Colombia-Venezuela border (Gaudette and
Olszewski, 1985; Hammond, 2005). Lowlands of the Amazon Graben form the shield?s
southern boundary, and the Rupununi Savannas in the middle of the Guiana Shield are
comprised of Cenozoic sediments filling a rift valley up to 5400 m deep (see Proto-
Berbice below). Even basins of the Western Amazon have, since the Cretaceous, been
filling with sediments from the Guiana and Brazilian shields (R?s?nan et al., 1998).
Despite the dramatic topographic results of a long history of erosion and sediment
redistribution, the modern shield highlands are subject to chemical weathering almost
exclusively and shield rivers carry very little sediment (Lewis and Saunders, 1990; see
Limnology and Geochemistry of Shield Rivers below).
Epeirogenic uplift of the Guiana Shield has occurred sporadically almost since its
formation in the Paleoproterozoic. Since at least the middle Paleozoic, when the region
was first exposed at the surface, cycles of uplift and stasis during which erosion occurred
have resulted in elevated erosional surfaces (pediplains or planation surfaces) that are
now observed throughout the northern interior of South America (Table 1; Schaefer and
do Vale, 1997, Gibbs and Barron, 1993). At lower elevations, these appear as steps or
stages of Roraima Formation sediments, vertically separated from each other by 60 to
200 m elevation. At higher elevations, collections of peaks can be identified that share
similar elevations (Fig. 2). Berrang? (1975:813), for example, described the ?remarkable
concordance of summits? of the Kanuku Mountains, which are mostly between 900 and
142
946 m-asl. The heights of Kanuku peaks can be correlated with heights of the Pakaraima-
Parima ranges to the northwest, Wassari Mountains to the south, and several other peaks
to the north and east, each separated from the other by hundreds of kilometers. Five
surfaces, one higher and four lower than that of the Kanuku Mountains, have been
identified and assigned tentative ages (Table 1; Schubert et al., 1986).
The ages and history of Guiana Shield uplift provide important clues to the origin
and evolution of topographic formations such as drainage divides and waterfalls
particularly relevant to the distribution patterns of aquatic faunas. Kaieteur Falls (226 m
high) in Guyana, for example, isolates the only known habitats of Lithogenes villosus, a
basal astroblepid or loricariid, and Corymbophanes andersoni and C. kaiei, the only two
species of this basal genus of hypostomin loricariid (Armbruster et al., 2000). Fossil
calibrated relaxed molecular clock data (Lundberg et al., 2007) suggest that these
relictual taxa predate the Oligocene uplift of the Kaieteur planation surface and may
therefore owe their continued existence to isolation via uplift of this barrier (see Relictual
Fauna below).
4.3.3 Proto-Berbice (Central Shield)
The hydrologic history of South America is a dynamic one, and a large body of
evidence indicates that many of the paleofluvial predecessors of modern drainages were
substantially different from rivers seen today. Regardless, the Guiana Shield has been
embedded among headwaters of the Amazon, Orinoco, Essequibo and their paleofluvial
predecessors, since their inception. Late Mesozoic and Paleogene terrigenous sediments
recorded from the Caribbean Sea (Kasper and Larue, 1986) and Atlantic Ocean (Dobson
143
et al., 2001) are derived from Proterozoic and Archean age sources, indicating that, prior
to Neogene acceleration of Andean uplift, the Guiana and Brazilian Shields were the
continent?s major uplands, and likely the continent?s most concentrated regions of high-
gradient lotic habitat (Galvis, 2006). One of the largest drainages of the central Guiana
Shield during much of the Cenozoic was the proto-Berbice, a northeast-flowing river
draining portions of Roraima State, Brazil, most of Guyana, and parts of southern and
eastern Venezuela and western Suriname (Sinha, 1968, Schaefer and do Vale, 1997).
Central to the historical geography and hydrology of the proto-Berbice is the
Takutu Graben, a deep structural divide between eastern and western lobes of the Guiana
Shield approximately 280 km long by 40 km wide and up to 7 km deep, centered on the
town of Lethem, Guyana. The modern graben is a valley between the Pakaraima and
Kanuku Mountains trending ENE to WSW and approximately equally divided between
Brazil and Guyana. Early rifting of the graben resulted in volcanism in the Late Triassic
to Early Jurassic but the depression has received freshwater sediments since the Middle
to Late Jurassic. Lake Maracanata, an endorheic lake approximately 75 to 100 m deep
(though progressively shallower through time and fluctuating greatly in depth through
periods of aridity) occupied the graben until the Early Cretaceous (Crawford et al., 1985).
This ancient lake received predecessors of the modern Ireng, Cotinga, Takutu,
Uraricoera, Rupununi, Rewa, and Essequibo Rivers (McConnell, 1959, Sinha, 1968,
Berrang?, 1975, Crawford et al., 1985).
From Late Cretaceous to Paleogene, Lake Maracanata transitioned to a fluvial
environment with a trunk stream, the proto-Berbice, that flowed northeast through the
144
North Savannas Gap and exited to the Atlantic between the modern towns of New
Amsterdam, Guyana and Nickerie, Suriname (McConnell, 1959). Head-cutting by the
Branco River, a south-flowing tributary of the Amazon River, into the western end of
what had been Lake Maracanata robbed the proto-Berbice of the Cotinga and Uraricoera
first, at the end of the Pliocene, then the Ireng and Takutu in the Pleistocene. The broader,
flatter bed now apparent in the Takutu relative to the Ireng indicates that the former was
captured and rejuvenated first, whereas the latter, with its entrenched, meandering bed, is
still accommodating to its new slope (Sinha, 1968). The modern Berbice River itself has
withered, and is now dwarfed by its former tributaries the Essequibo and Corentyne to its
northwest and southeast, respectively. Evidence of a shift away from the lower Berbice as
the more important trunk stream can be observed in an elbow of capture near Massara, at
the eastern edge of the Maracanata Basin. This is the point at which the modern upper
Essequibo shifts abruptly westward, away from a nearby north-flowing Berbice tributary,
that has aggraded in response ? raising the level of the stream bed (Gibbs and Baron,
1993).
It seems likely, given their considerable endemism (see Caroni (Orinoco) to
Cuyuni/Mazaruni Corridors below), that the Mazaruni and Cuyuni rivers were also only
recently linked with the Essequibo, and that they historically exited to the Atlantic via
their own mouth, separate from that of the proto-Berbice. In the southern Guiana Shield
highlands of Venezuela, strongly recurved elbows of capture are regular features of the
upper Caroni, Caura, and Erebato (Caura), which, along with biogeographic evidence
(Lujan, 2008), indicate historical confluence of these headwaters with the southeasterly
flowing upper proto-Berbice, now the Uraricoera River.
145
The North Rupununi Savannas occupy the modern Maracanata depression and
form a shallow continental divide between the northeastern versant of South America,
drained in this area predominantly by the Essequibo, and a more southern versant that
drains to the Amazon via the Branco and Negro. Seasonal (May to August) rains
regularly flood this divide, forming a lentic connection extending to over 6,000 km
2
 and
centered between the north-flowing Rupununi River and headwaters of the Pirara River, a
west-flowing tributary of the Ireng. Lake Amuku is the name sometimes applied to the
broad areal extent of these floodwaters (Lowe-McConnell, 1964), as well as to one or
more restricted ponds into which floodwaters retreat (NKL pers. obs.). Lacustrine
sedimentation from the annual inundation continues to contribute to a shallowing of the
Maracanata basin (Sinha, 1968) and a possible long-term reduction of its role as
biogeographic portal between the Essequibo and Negro watersheds.
Tilting of the underlying basement both in the North Rupununi Savannas and
across the Guiana Shield has occurred as recently as the Holocene (Gibbs and Baron,
1993) and is likely a frequent driver of head cutting and stream capture. Evidence of this
can be seen in the disproportionate incision of tributaries on one side of rivers flowing
perpendicular to the direction of tilt, and aggradation of tributaries on the opposite side.
Tilting to the west in the South Rupununi Savannas, for example, has led to rejuvenation
and steepening of eastbank tributaries, and aggradation and sluggishness in westbank
tributaries of the north flowing Takutu, Rupununi, and, in part, Kwitaro rivers (Gibbs and
Baron, 1993). In Southeastern Venezuela, a gradual shift in the prevailing tilt of the Gran
Sabana from north to south is thought by L?pez et al. (1942) to be responsible for
remarkably complex drainage patterns in the upper Caroni River. Abrupt and localized
146
orthogonal shifts in channel direction, with streams of the same drainage flowing in
parallel but opposite directions are common features, as are biogeographic patterns
indicative of frequent stream capture (Lasso et al., 1990; see Caroni to Cuyuni/Mazaruni
Corridors below).
4.3.4 Proto-Orinoco (Western Shield)
The western Guiana Shield features one of the largest and most notable river
capture events in the Neotropics: that of the ongoing piracy of the northwest-flowing
Upper Orinoco River by the southeast-flowing Negro River, via the southwest-flowing
Casiquiare Canal. The Casiquiare Canal diverts up to 20% of the Upper Orinoco?s
discharge away from the Orinoco trunk and into the Amazon via the Negro. This is a
relatively recent phenomenon, however, in the dynamic history of the Orinoco River,
which has given rise to the upper-Amazon and Magdalena Rivers while its own main
channel migrated progressively eastward from an ancestral north-south orientation. Prior
to consolidation by any trunk stream, from at least the Campanian to the Maastrichtian,
westward flowing drainages from highlands of the Guiana and Brazilian Shields likely
followed short, anastomosing channels across a broad coastal plain, into shallow marine
environments that occupied much of what is today Colombia, Ecuador, and Peru. The
Panamanian Isthmus was not yet present and the northern Andes were just beginning to
form.
By the Middle Eocene, uplift of the Central Cordillera had progressed to the
extent that it formed the western margin of a large south to north trending valley, drained
147
by a single fluvial system, then expanded by coalescence of both high gradient left bank
tributaries draining the eastern slope of the young Central Cordillera and right bank
tributaries flowing west from Guiana Shield uplands. The mainstem of this proto-Orinoco
was a large, low-energy, meandering river that deposited ?vast amounts of sediment?
(Villamil, 1999: 245) in a geological formation called the Misoa Delta in what is now the
Maracaibo Basin (D?az de Gamero, 1996). From Late Eocene to Oligocene, marine
incursions pushed the mouth of the proto-Orinoco back as far south as the modern town
of Villavicencio, Colombia up to five times (D?az de Gamero, 1996; Villamil, 1999). In
the Late Oligocene, the proto-Orinoco expanded longitudinally to the north-northeast
(Shagam et al., 1984; Villamil, 1999), and by the Early Miocene, it was flowing into the
eastern end of the La Pascua-Roblecito marine basin, a deep seaway occupying much of
modern-day Falcon State in northwest Venezuela. The proto-Orinoco and its mouth were
isolated at this time from the Maracaibo Basin by uplift of the Merida Andes (Shagam et
al., 1984; Villamil, 1999) and from the Eastern Venezuela Basin by the El Baul structural
arch (Kiser and Bass, 1985; D?az de Gamero, 1996).
4.3.5 Eastern Venezuela Basin (Northern Shield)
The Eastern Venezuela Basin is a structural depression located between the
northern edge the Guiana Shield and the northern coast of South America that receives
lower portions of the northern shield drainages Caura, Aro and Caroni. The basin is
asymmetric in bottom profile, growing shallower to the south and west and opening to
the northeast, where the basement is over 12 km deep. The entire basin is filled and
148
leveled with Mesozoic to Cenozoic sediments now zero to 50 m-asl and comprising the
eastern half of the Venezuelan Llanos. Approximately 800 km east to west and 250 km
north to south, the basin is bounded in the east by the Sierra Imataca, a northern arm of
the Guiana Shield, and in the south by the Guiana Shield proper. In the north, it is
bounded by the Sierra del Interior and Coastal mountain ranges; and in the east, it has an
opening to the Apure-Barinas basin that is constricted as between a thumb and forefinger
by the coastal mountain ranges in the north and the El Baul structural arch in the south.
From at least the Lower Cretaceous to the Early Eocene, the Eastern Venezuela
Basin was a marine environment that received rivers draining the northern slope of the
Guiana Shield directly. In the Early Eocene, however, tectonic convergence of the
Caribbean Plate caused widespread emergence of the Eastern Venezuela Basin and
northward expansion of the coastal plain. In response, the Caura extended its lower
course north-northeast so that it formed a delta in the Sucre region of Venezuela between
the modern islands of Margarita and Trinidad (Rohr, 1991; Pindell et al., 1998).
Emergent, coastal plain conditions largely prevailed in the Eastern Venezuela Basin
throughout the Eocene and into the Oligocene but convergence of the Caribbean Plate in
the Late Oligocene caused southeastward migration of the La Pascua-Roblecito seaway.
While the proto-Orinoco continued to discharge into the seaway?s closed southwestern
end, the Caura and Caroni coalesced in a more restricted coastal plain and delta near the
seaway?s eastern opening, in the northern portion of Anzoategui State, Venezuela (Rohr,
1991; Pindell et al., 1998).
In the Early Miocene, regression of the seaway, and consequent eastward
149
progradation of the proto-Orinoco placed deltas of the proto-Orinoco and Caura in close
proximity at the western margin of the Eastern Venezuela Basin, but uplift of the El Baul
arch at this time ensured that they remained separate until at least the Middle Miocene
(Pindell et al., 1998). In the Late Middle Miocene, southward propagation of rapid uplift
in the Serrania del Interior, along with a possible decrease in the significance of the El
Baul arch, pushed the proto-Orinoco southward to capture the lower course of the Caura
(Pindell et al., 1998). Further progradation and eastward movement of the Orinoco put its
delta in the region of modern Trinidad in the mid-Pliocene. Final conformation to its
modern course, adhering closely to the northern edge of the Guiana Shield, occurred in
the Late Pliocene to Pleistocene (Rohr, 1991; Hoorn et al., 1995; D?az de Gamero, 1996).
4.3.6 Proto-Amazon and eastern Atlantic drainages (Southern and Eastern Shield)
The Amazon River?s birth as a distinctly South American, versus Gondwanan,
river can be dated to at least the Middle Aptian, approximately 120 Ma. Fossil evidence
indicates that by the Late Aptian, an equatorial seaway linked the North and South
Atlantic, thereby dividing the once-contiguous landmass of Gondwana into South
America and Africa (Maisey, 2000). Given the much older Proterozoic structural
evolution of the Amazon Graben as a regional lowland, and its sediment fill dating at
least to the Cambrian (Putzer, 1984), it can be assumed that from the moment the South
Atlantic Seaway opened, a paleofluvial predecessor of the Eastern Amazon drained the
southeastern Guiana and northeastern Brazilian shields east through a mouth
approximately coincident with its modern delta. This proto-Amazon was much smaller
150
than the modern Amazon-Solim?es system. For over 100 My following the breakup of
Gondwana, upper and lower portions of the modern Amazon Basin (approximately
coincident with the modern Solim?es and Amazonas reaches), were separated by the
Purus Arch, a continental divide within the Amazon Graben located near the modern
mouth of the Purus River. Lowland portions of those proto-Amazon tributaries draining
the southern slope of the Guiana Shield would have also been separated by the Purus
continental divide into western and southeastern paleo-drainages. The upper Negro,
Caqueta (Japur?), and upper Orinoco would have flowed west or northwest during this
period, either directly into the Pacific or into the Caribbean via the proto-Orinoco (see
Proto-Orinoco above).
Southeastern drainages of the Guiana Shield would have been further limited in
areal extent, and distanced from the western lobe of the Guiana Shield relative to their
modern pattern by the expanded proto-Berbice draining the central Guiana Shield region
(see Proto-Berbice above). Paleodrainages of the southeastern Guiana Shield that would
have been south of the proto-Berbice?s approximate watershed boundaries, and east of
the Purus Arch, and therefore still been northern tributaries of the proto-Amazon, would
have included the lower Branco/Negro below the Mucujai River, and the Uatuma,
Trombetas, and Paru Rivers. A series of ridges with peaks in the range of 400 to 1000 m
extends east from the Kanuku Mountains and forms another continental divide within the
eastern lobe of the Guiana Shield, in this case separating south-flowing Amazon
tributaries from northeast flowing Atlantic Coastal drainages. Headwaters of respective
northern and southern rivers interdigitate across these highlands, though, rendering them
largely porous to fish dispersal (Nijssen, 1970, Cardoso and Montoya-Burgos, 2009; see
151
Atlantic Coastal Corridors below). The westernmost of these east-to-west ranges,
forming the border between Brazil and Guyana, are the Wassari and Acarai Mountains
which give rise to the Trombetas, the fourth largest watershed on the Guiana Shield
(drainage area 136,400 km
2
). The easternmost of these ranges, forming the southern
borders of Suriname and French Guiana, are the Tumucumaque Mountains, which give
rise to the Paru River (44,250 km
2
). North of this divide, in order from west to east, flow
the Correntyne (68,600 km
2
), Coppename (21,900 km
2
), Suriname (17,200 km
2
), and
Marone (70,000 km
2
) rivers. Finally, draining the eastern slope of the eastern Guiana
Shield is the Oyapok River (32,900 km
2
), which forms the border between French Guiana
and Brazil, and the Approugue River (10,250 km
2
) just to its northwest inside French
Guiana (Fig. 1; drainage area data from Hammond, 2005).
In the Late Miocene, paroxysms of Andean uplift shifted the prevailing slope of
the Andean back arc basin eastward and caused Andean-derived watercourses to breach
the Purus Arch, vastly expanding the proto-Amazon?s watershed westward. New regions
of the Amazon Basin included vast swaths of the modern western and southwestern
Amazon Basin that had been tributary to the proto-Orinoco and are now tributary to the
upper Amazon mainstem (Solim?es). New northern tributaries of the expanded Amazon
included drainages of the western lobe of the Guiana Shield such as the Caqueta and
upper R?o Negro. Uplift of the Vaupes Arch and Macarena Massif contemporaneous with
the Late Miocene Western Cordillera uplift also created a new drainage divide
segregating the upper Negro from the upper Orinoco (Galvis, 2006).
Orographic rainfall effects of the rapidly rising Andes Mountains also contributed
152
to the expansion of the Amazon in the Late Neogene by increasing its discharge beyond
that predicted by its areal expansion alone. With the increase in discharge came an
increase in erosional potential and further watershed expansion via head cutting. The
southeast flowing Branco River, for example, sequentially captured headwater tributaries
of the northeast flowing proto-Berbice throughout the Pliocene and Pleistocene. Indeed,
the Amazon is still expanding, as seen in the ongoing capture of the upper Orinoco by the
Rio Negro. The initiation of this capture and opening of this portal has been hypothesized
to be fairly recent, possibly due to Late Pleistocene or even Holocene tilting (Stern, 1970;
Gibbs and Barron, 1993). Under this scenario, the Orinoco is estimated to have been
largely isolated from the Amazon for some 5?10 My, from the Late Miocene uplift of the
Vaupes Arch to the Pleistocene-Holocene formation of the Casiquiare Canal. The future
seems to be one in which a new drainage divide forms within the Orinoco downstream of
the Tama-Tama bifurcation, and the current headwaters of the Orinoco become entirely
adopted by the Amazon (Stern, 1970).
4.3.7 Aridity and Marine Incursions
We have thus far described, in broad strokes, major trends and events in the
drainage evolution of four hydrologic regions around the Guiana Shield, but we have
done so at the expense of dwelling in too great detail on global cycles and climatological
events that had periodic, widespread effects across all hydrologic units. Aridity and
marine incursions are treated here together because of their similar effects on rivers and
riverine biota ? that of reducing and isolating habitats over a broad geographic range. The
two phenomena are also correlated in their response to global cycles of glaciation with a
153
periodicity of 20?100 thousand years (Milankovitch cycles; Bennett, 1990). In general,
warmer, interglacial climates correspond to higher sea levels, more extensive marine
incursions, and higher levels of precipitation. Cooler, glacial periods result in reduced
precipitation, retreat of the sea, expansion of the coastal plain, and incision of river
channels.
Many lines of geologic and biogeographic evidence indicate that the climate of
South America was much drier in the recent past than it is today. The last major glacial
period, the W?rm or Wisconsin glaciation, lasted throughout the Late Pleistocene, from
approximately 110,000 BP to between 10,000 and 15,000 BP. Several authors (e.g.
Krock, 1969, Hammen, 1972, Tricart, 1985, Schubert et al., 1986, Schubert, 1988)
describe the substantial paleobotanical and geomorphological evidence of aridity in South
America during this period. Hammen (1972) states that within the overall trend of late
Pleistocene aridity, the period from approximately 21,000 to 13,000 BP was the driest.
Terrestrial vegetation throughout much of the Guianas in the Late Pleistocene was
of an open savanna or grassland type, with rainforests likely limited to a few highland
refugia, including parts of the Pantepui highlands, and riparian margins. These refugia
feature heavily in explanations of patterns of terrestrial plant and animal diversity (e.g.,
Haffer, 1969, 1997, Prance, 1973, Vanzolini, 1973, Brown and Ab?S?ber, 1979, Kelloff
and Funk, 2004), but do not appear to useful in explaning freshwater fish distributions
(Weitzman and Weitzman, 1982; see reviews in Chapters 1 and 18). Loss of major forest
cover along with a decrease in sea level had major effects on the geomorphic evolution of
South American rivers. Without forest cover to keep soil intact, and because of lowered
river base levels due to lower sea levels, many rivers cut deeply into their channels
154
(Sternberg, 1975, Tricart, 1985, Latrubesse and Franzinelli, 2005). Channel bottom in the
lower reaches of some Amazon tributaries, for example, can be up to 80 m below modern
sea level (Sioli, 1964, Latrubesse and Franzinelli, 2005). Rapids would have been more
widespread during periods of aridity, and deep-channel habitats that may currently
function as barriers to rheophilic taxa would have been reduced. Decreases in total
discharge would have lead to shallowing, sedimentation and aggradation or braiding of
low gradient habitats where they persisted (Schubert et al., 1986, Latrubesse and
Franzinelli, 2005). Garner (1966) also suggests that the complex drainage pattern of the
upper Caroni in the Gran Sabana may be the result of anastomosing channel development
during a more arid climatic regime, and that the latest humid period has not lasted long
enough for the Caroni to consolidate into a more stable drainage pattern.
Marine incursions have inundated much of northwestern South America during
interglacial periods of globally high sea level since at least the Maastrichtian (Gayet et
al., 1993, Hoorn, 1993; see review in Chapter 17). During much of the Miocene, from
approximately 23?11 Ma, the llanos basins of Colombia and Venezuela were dominated
by coastal and lagunal conditions with occasional marine episodes (Hoorn et al., 1995).
Given their similar elevations and exposure to the coast, it is likely that similar conditions
prevailed in the Rupununi Savannas and coastal plain of Guyana. A more recent marine
incursion, approximately 6?5 Ma, was hypothesized by Hubert and Renno (2006) to have
affected the distribution and diversity of characiform fishes in northeastern South
America by isolating a series of upland freshwater refuges in respective eastern and
western portions of the eastern Guiana Shield highlands. Further support for such an
incursion is provided by Noonan and Gaucher (2005, 2006), who recovered a temporally
155
and spatially congruent vicariance pattern in their molecular phylogenetic studies of
Dendrobates and Atelopus frogs.
4.3.8 Limnology and Geochemistry of Shield Rivers
Rivers of the Guiana Shield are a heterogeneous mix of white-, black-, and
clearwater, with most tributaries initially trending toward black- and clearwater then
mixing to form intermediate main stems. In heavily forested and largely uninhabited
regions such as the shields and the Amazon Basin of South America, topography,
climate, geology and a watershed?s terrestrial vegetative cover are the main influences on
a river?s limnological character. Because of their origins in watersheds of ancient, highly
weathered, and forest covered basement rock, rivers of the Guiana Shield tend to be
nutrient poor with very low levels of suspended solids and alkalinity, but relatively high
levels of dissolved silica. Limestones and evaporates are completely absent in the Guiana
Shield, so the chemical signature of its rivers are largely influenced by primary
weathering of silica-rich felsic granitoids that are the dominant rock type (Edmond et al.,
1995, Hammond, 2005). Concentrations of major ions and nutrients in the Orinoco
mainstem and its Guiana Shield tributaries fall near or even below those expected in
precipitation, though, indicating minimal contributions from geology and significant
sequestration by forests (Lewis and Weibezahn, 1981).
Extreme blackwater conditions of low acidity (pH <5.5), negative to low
alkalinity, and low conductivity (<25 ?ohms) prevail in the Atabapo, Guainia, Negro,
Pasimoni, and other tributaries of the Casiquiare that drain low-lying peneplains between
the upper Orinoco and Negro. Adjacent higher-gradient headwaters of the Orinoco such
156
as the Ventuari, Ocamo, Mavaca, and Guaviare are white- to clearwater, with near neutral
pH, alkalinity up to 85 ?eq/kg, and up to 29 ?ohms conductance (Thornes, 1969,
Edmond et al., 1995). Rivers draining the northern slope of the Guiana Shield, such as
the Caura, Caroni, and Cuyuni, as well as rivers further east in Suriname and French
Guiana, trend toward blackwater conditions despite also having higher gradients.
Drainages in the central, south and southeast of the Guiana Shield, such as the Essequibo,
Branco, Trombetas, and Paru Rivers are clear to whitewater. Guiana Shield whitewater
rivers, it should be noted, are defined based largely on alkalinity and pH, being
considerably lower in suspended solid load relative to those Andean drainages on which
the traditional definition of whitewater rivers is based (Sioli, 1964).
4.4 BIOGEOGRAPHY OF GUIANA SHIELD FISHES
4.4.1 Modern Corridors: the Prone-8
Phylogeography of South American fishes is hampered by a lack of collections
and a lack of studies, and the problems of amassing specimens for phylogeographic
studies have been particularly acute in the Guiana Shield. Most of the region is difficult
to access with few or no roads to important habitats. Among the better-sampled areas are
the lowlands of Amazonas, Venezuela, the lower and upper Caroni of Venezuela
(although not the middle reaches or of the Paragua, a large tributary), the Cuyuni of
Venezuela, the Rupununi and Takutu of Guyana, and much of French Guiana. The
western highlands, the Mazaruni, the Corantijne, and most rivers of the southern edge of
the Guiana Shield have been poorly sampled. Phylogeographic studies are especially
hampered by the scarcity of collections from headwaters throughout the Guiana Shield.
157
Most molecular phylogenetic studies inclusive of the Guiana Shield, including those by
Lovejoy and Ara?jo (2000) of Potamorrhaphis, Turner et al. (2004) and Moyer et al.
(2005) on prochilodontids, and Willis et al. (2007) on Cichla, are based on lowland taxa
potentially capable of great vagility that could obscure fine-scale biogeographic patterns
within the Guiana Shield.
To observe fine-scale biogeographic patterns within and between drainages, low-
vagility taxa that are less likely to have biogeographic patterns erased via migration and
panmixis should be examined. Rheophilic fishes whose movement between drainages
would be expected to be hampered in the absence of stream capture or lowered main
channel base levels are good examples of such taxa. The phylogenetic patterns of
rheophilic taxa distributed allopatrically across isolated headwaters may be particularly
informative when trying to understand the biogeographic significance of such historical
events. Among rheophilic Neotropical fishes, loricariid catfishes of the tribe Ancistrini
(Hypostominae) are a group with several genera and species that appear to be both most
common and most diverse in shield regions. Ancistrin catfishes are, as a whole, also
highly territorial with relatively low vagility (see Power, 1984), and are a group that we
have studied most. We will therefore focus on them in the discussion below.
For the purpose of our discussion, we refer to taxa known only from the Guiana
and/or Brazilian Shields as Shield Endemic Taxa. Taxa that are only most common
within the Guiana and/or Brazilian Shields, but have ranges extending beyond these
regions, are considered Shield Specialist Taxa. Although our discussion of biogeographic
patterns will focus on species in the tribe Ancistrini and Hypostomini that we have
studied, published examples from other taxa will also be discussed.
158
Problematically, the phylogeny and taxonomy of Loricariidae are in their infancy
and are complicated by gross morphological similarity. In many of our studies
(Armbruster, 2005; 2008; Armbruster et al. 2007), we have found little morphological
variability within genera upon which to base robust phylogenies; however, by using what
is known of ancistrin phylogenetics, distributions, and historical and current corridors
between river systems, we support below a conceptual model of biogeographically
significant hydrologic corridors around the Guiana Shield. This model approximates the
appearance of a prone number 8 (Fig. 3). Corridors between hydrologically contiguous
segments of this Prone-8 consist of both recently formed portals such as the Casiquiare
Canal (see Proto-Orinoco above), recently closed or altered corridors such as the
Rupununi Savannas, and numerous intermittent corridors that have likely been present in
the recent past. These corridors allow for dispersal around the shield, and their
intermittent nature serves as the basis for allopatric speciation. Given that our
understanding of the geologic and hydrologic evolution of the Guiana Shield extends
beyond the node-age estimates for most Neotropical taxa, especially ancistrin loricariids,
we assume that such geophysical evolution has been relevant to the dispersal of extant
taxa. The alternative argument that modern Neotropical fish distributions are the result of
widespread extinction versus dispersal will be discussed under Relictual Taxa below, but
will not be considered in the majority of our discussion.
4.4.2 Caroni (Orinoco) to Cuyuni/Mazaruni Corridors
Streams of the lower Caroni interdigitate with streams of the upper Cuyuni, and
streams of the upper Caroni interdigitate with the upper Mazaruni, allowing the
159
possibility of stream capture between Orinoco and Essequibo drainages. Lasso et al.
(1990) found a close similarity between whole fish communities of the Caroni in the
Gran Sabana and those of the Cuyuni-Essequibo system and hypothesized frequent
stream capture as a cause. Despite the relative richness of the loricariid fauna in the
Orinoco and Essequibo basins, there seems to be little evidence that the Caroni to Cuyuni
and Caroni to Mazaruni corridors are particularly important for loricariids. Armbruster
and Taphorn (2008) suggest that the ancestor of Pseudancistrus reus (Ancistrini) may
have entered the Caroni from the Cuyuni as it is the only member of Pseudancistrus
sensu stricto currently known from the Orinoco (all other Orinoco Pseudancistrus are
basal species); however, P. reus has some unique characteristics that make its
relationship to other Pseudancistrus unclear. The only species of Pseudancistrus we
know of in the Cuyuni is a species with large white blotches that may be undescribed,
and that is relatively common in the Essequibo.
Exchange of loricariids between the upper Caroni and upper Mazaruni also seems
to be rare, and consistent with a general tend in which many fish taxa seem to be endemic
to the Mazaruni alone. An undescribed species of Exastilithoxus (Ancistrini) from the
upper Mazaruni has been reported in aquarium literature although we have not examined
specimens, and E. fimbriatus is restricted to the upper Caroni. Two undescribed species
of Neblinichthys (Loricariidae: Ancistrini) were collected during recent fieldwork in the
upper Mazaruni by H. Lopez and D. Taphorn (pers. comm.), while the congeneric N.
yaravi is only known from the upper Caroni. Non-loricariid taxa endemic to the Mazaruni
include a recently described new species, possibly new genus, of parodontid (Apareiodon
agmatos, Taphorn et al., 2008), a basal crenicarine cichlid (Mazarunia mazarunii;
160
Kullander, 1990), a lebiasinid, possibly sister to the Pyrrhulininae (Derhamia
hoffmannorum; G?ry and Zarske, 2002), and a basal Nannostomus (N. espei; Weitzman
and Cobb, 1975). The basal crenuchid Skiotocharax meizon was also described largely
from the Mazaruni (Presswell et al., 2000). Taken together, these taxa provide strong
evidence of long-term isolation of the Mazaruni River. Indeed, if the proto-Berbice
paleodrainage hypothesis is correct, it seems like that the Mazaruni would have
maintained its own mouth to the Atlantic through much of the Miocene-Pliocene when
the proto-Berbice is thought to have exited further to the east.
4.4.3 Casiquiare Portal
The Casiquiare Canal is a large and permanent (navigable year round) corridor
between the upper Orinoco and the upper Rio Negro (Amazon). Distributions of species
across the Casiquiare have been studied by Chernoff et al., (1991), Buckup (1993),
Schaefer and Provenzano (1993), Lovejoy and Ara?jo (2000), Turner et al. (2004),
Moyer et al. (2005), and Willis et al. (2007). Winemiller et al. (2008) and Winemiller
and Willis (Chapter 11 this volume) review this literature and supplement it with fish
community ecology data transecting the entire Casiquiare. They suggest several distinct
distribution patterns: broad distribution in the Orinoco and Negro, distribution in the
upper Orinoco and upper Casiquiare (but not lower Casiquiare or Negro), and distribution
in the lower Casiquiare and the Negro (but not upper Casiquiare or Orinoco). They
attribute the second two distributional patterns to an environmental gradient from
clearwater (Upper Orinoco) to blackwater (Negro). In addition to this limnological
gradient, upper portions of the Orinoco and Negro are isolated from lower portions of
161
their respective drainages by the high energy rapids Atures and Maipures (Orinoco) and
S?o Gabriel (Negro). Several Amazonian species conspicuously absent from the Orinoco
basin (e.g., Osteoglossum spp., Arapaima gigas, Parapteronotus hasemani,
Orthosternarchus tamandua, Symphysodon spp.) are likely more subject to exclusion by
the rapids at S?o Gabriel than by shifts in limnology.
Turner et al. (2004) and Moyer et al. (2005) reported complete segregation
between mitochondrial genotypes of Prochilodus mariae and P. rubrotaineatus in the
Orinoco and P. rubrotaineatus in the Negro and Essequibo. Likewise Lovejoy and
Ara?jo (2000) identified basal haplotypes of Potamorrhaphis that were isolated in the
upper Orinoco and not shared with Negro populations, indicating a barrier at the
Casiquiare. Willis et al. (2007), however, observed that three of the four Cichla taxa
present in the upper Orinoco (C. monoculus, C. orinocensis, C. temensis) were
genetically similar to conspecifics in the upper Negro. Chernoff et al. (1991) list 16
species (11 Characiformes, four Siluriformes, one Gymnotiform) distributed from the
upper Orinoco, across the Casiquiare, into the upper Negro, and the revision of
characidiin fishes by Buckup (1993) gives eight more species whose distribution at least
encompasses this divide.
Several loricariids have broad distribution patterns that include the Orinoco,
Casiquiare, Negro, and possibly even northern tributaries of the Brazilian Shield. We
have studied five species (two shield endemics and three shield specialists) of
hypostomines that occur in the Orinoco, Casiquiare/Negro, and drainages of the Brazilian
Shield - Shield Endemics: Hemiancistrus sabaji (Armbruster, 2008) and Leporacanthicus
galaxias; Shield Specialists: Hypostomus hemicochliodon (Armbruster, 2003),
162
Lasiancistrus schomburgkii (Armbruster, 2005), and Peckoltia vittata (Armbruster,
2008). Two of these species (H. sabaji and L. schomburgkii) are also found in the
Essequibo. These species may offer the best insights into potentially recent movements of
taxa among drainages of the Guiana Shield and between the Brazilian and Guiana Shield;
however, intra-taxon relationships must be explored with genetic techniques to determine
the relative timing of current distributions and degree of population structure. Three
Shield Endemic genera have ranges similar to those outlined for the species above
(Baryancistrus, Hypancistrus, and Leporacanthicus; Werneke et al., 2005a; Armbruster
et al., 2007, Lujan et al., 2009) as do two Shield Specialist genera (Hemiancistrus and
Peckoltia; Armbruster, 2008).
Several loricariids have distributions limited to the Upper Orinoco and upper
Casiquiare. Hemiancistrus guahiborum, H. subviridis, Hypostomus sculpodon,
Pseudancistrus orinoco, P. pectegenitor, and P. sidereus all occur in the upper Orinoco
and Casiquiare, but are not currently known from elsewhere in the Amazon. A few
recently described species from the Orinoco have putative sister species in the
Casiquiare: Hypancistrus inspector (Casiquiare) vs. H. contradens and H. lunaorum
(Orinoco) and Pseudolithoxus nicoi (Casiquiare) vs. P. anthrax (Orinoco). Given the
relatively recent formation of the Casiquiare Portal (Late Pleistocene to Holocene; see
Proto-Amazon above), these species may represent recent invasions from the Orinoco to
the Casiquiare and/or relatively recent speciation events. Aside from a few widespread,
blackwater adapted species (e.g.,, Dekeyseria niveata, D. pulchra), most ancistrin
loricariids appear to be excluded from the lower Casiquiare and upper Negro by
extremely blackwater limnology.
163
4.4.4 Southern Guiana Shield and Northern Brazilian Shield Corridors
The mainstem Amazon River likely acts as a partial barrier for both Shield
Endemic and Shield Specialist taxa on the respective Guiana and Brazilian Shields.
Genera known to withstand more lowland conditions (e.g.,, Ancistrus, Lasiancistrus, and
Hypostomus) may be able to cross the Amazon Basin, but such dispersal is unlikely
among most ancistrins. East-West dispersal around the southern part of the Guiana Shield
may be via either southern Guiana Shield drainages or drainages of the northern part of
the Brazilian Shield. Currently, the fauna of the northern Brazilian Shield is much better
known than that of the southern part of the Guiana Shield. Species and genera mentioned
above from both the Guiana and Brazilian shields offer potential examples of movement
across the northern Brazilian Shield at least to the Tocantins. Dispersal along the southern
flank of the Guiana Shield may be exemplified by Pseudancistrus sensu stricto as several
undescribed species are known from these drainages; however, undescribed species are
also known from the northern Brazilian Shield. Demonstration of ancistrin biogeographic
patterns across the southern Guiana Shield and northern Brazilian Shield must, therefore,
await further collections and analysis of genetic data. Amongst other fishes, the range of
Psectrogaster essequibensis (Characiformes: Curimatidae; Vari, 1987) and Parotocinclus
ariapuanensis and P. britskii (Loricariidae: Hypoptopomatinae) support dispersal via the
northern Brazilian Shield, although we reiterate that collection data in this region are
poor.
164
4.4.5 Rupununi Portal
The Rupununi Savanna floods seasonally, creating a lentic corridor between the
Essequibo and Takutu Rivers, the latter of which was lost to the Negro via stream capture
as recently as the Pleistocene (see Proto-Berbice above). Loricariids of the Essequibo are
nearly identical to those of the Takutu, indicating either regular, recent dispersal across
the flooded savanna or insufficient time for differentiation since stream capture. Among
hypostomines we have examined, Hemiancistrus sabaji, Hypostomus squalinus, H.
macushi, Lasiancistrus schomburgkii, Lithoxus lithoides, and Pseudacanthicus leopardus
are well-represented in collections on either side of the Rupununi Portal and show no
morphological differentiation between drainages. Many other fishes also have ranges that
extend across the Rupununi Portal including Osteoglossum bicirrhosum, Arapaima gigas,
Psectrogaster essequibensis (Vari, 1987), and Rhinodoras armbrusteri (Sabaj et al.,
2008). Molecular phylogenetic studies by Lovejoy and Ara?jo (2000), Turner et al.
(2004), and Willis et al. (2007) support transparency of the Rupununi Portal for
Potamorrhaphis, Prochilodus rubrotaeniatus, and Cichla ocellaris, respectively.
The relative importance of the Rupununi Savannas as either portal or barrier is
difficult to demonstrate. Collections of Hypostomus taphorni have been made from
throughout the Essequibo, but from only one location in the Pirara River, a tributary of
the Ireng (Negro) near the drainage divide, seemingly indicative of recent immigration.
The existence of sister species Peckoltia braueri (Takutu) and P. cavatica (Essequibo) on
either side of the divide seems to support the Rupununi?s role as barrier. An undescribed
species of both Hypancistrus and Panaque have only been collected on the Takutu River
side, as has Cichla temensis (Willis et al., 2007), further supporting its role as barrier.
165
JWA?s lab is currently investigating gene flow across the Rupununi Portal in several fish
groups to determine both relative transparency of this portal for various taxa, and when it
may have become closed to rheophilic species intolerant of conditions in the flooded
savanna.
4.4.6 Atlantic Coastal Corridors
The exchange of fishes between Atlantic coastal drainages of the eastern Guianas
(Guyana, Suriname, French Guiana) and the eastern Amazon Basin may be accomplished
via either a coastal marine corridor with reduced salinity due to the westerly deflected
Amazon River discharge, coastal confluences during times of lower sea level and
expanded coastal plains, and/or headwater interdigitation and stream capture. The region
can be broadly divided into the Western Atlantic Coastal Corridor (from the mouth of the
Orinoco to the mouth of the Essequibo) and the Eastern Atlantic Coastal Corridor (from
the mouth of the Essequibo to (and possibly beyond) the mouth of the Amazon.
The Atlantic Coastal region is poorly represented in molecular biogeographic
studies of northern South America. Willis et al. (2007) report a single species of Cichla
(C. ocellaris) distributed from the Essequibo in the west to the Oyapock in the east, but
the aforementioned studies of Potamorrhaphis and Prochilodus do not cover this region.
In a morphology based taxonomic revision demonstrating a similar pattern to that of C.
ocellaris, Mattox et al. (2006) identified the single species Hoplias aimara in Atlantic
coastal drainages from the eastern Amazon Basin as far west as the northern Guiana
Shield drainages entering the Eastern Venezuela Basin, but not entering the upper
Orinoco. Renno et al. (1990, 1991) investigated the population structure of Leporinus
166
friderici using genetic markers and interpreted their data as providing support for the
existence of an eastern and western Pleistocene refuge from which this species has more
recently expanded its range. Their data identifies the Kourou River in French Guiana as
the point of convergence between historically isolated eastern and western populations.
Low gradient streams of the Western Atlantic coastal plain?s lower drainages are
unsuitable for most ancistrins. Hypostomus plecostomus and H. watwata, coastal plain
species that can be found in some estuaries, may use the low gradient streams and near-
shore marine habitats to move between drainages along the whole Atlantic Coastal
Corridor (Eigenmann, 1912; Boeseman, 1968). Several more rheophilic loricariid species
are restricted to upland habitats across the Eastern Atlantic versant. Lithoxus spp. are
found in upland habitats throughout the eastern Guiana Shield, and morphological
characters suggest they are divided into a western, proto-Berbice subgenus (Lithoxus,
2spp.), and an Eastern Atlantic Coastal subgenus (Paralithoxus, 5 spp.; Boeseman, 1982,
Lujan, 2008). Pseudancistrus sensu stricto is distributed throughout the eastern Guiana
and northern Brazilian shields, with only a single Orinoco species, P. reus, restricted to
the Caroni River (Armbruster and Taphorn, 2008). Pseudancistrus barbatus and P.
nigrescens are distributed from the Essequibo to French Guiana (Eigenmann, 1912; Le
Bail et al., 2000), P. megacephalus is in at least the Essequibo and Suriname Rivers
(Eigenmann, 1912), and P. brevispinnis is found from the Corantijn to the Oyapock and
in several northern tributaries of the Amazon (Cardoso and Montoya-Burgos, 2009).
Several species of Pseudacanthicus are also found across the eastern Atlantic Coastal
drainages of the Guianas, but specimens of these are rare in collections and they appear to
be largely restricted to main river channels. Pseudacanthicus and Pseudancistrus are both
167
Shield Specialists, with ranges throughout the eastern Guiana and northern Brazilian
shields, and Lithoxus is a Shield Endemic, making these groups excellent subjects for
biogeographic studies of the eastern Guiana Shield and adjacent areas.
Cardoso and Montoya-Burgos (2009) conducted a molecular phylogeographic
study of Pseudancistrus brevispinnis and found support for the hypothesis that this
species invaded the Atlantic Coastal river system from the south-flowing Jari River, a
tributary of the Amazon, via headwater interdigitation and stream capture with the north-
flowing Marone River. From the Marone, P. brevispinnis dispersed eastward as far as the
Oyapock River and westward as far as the Corantijn River (Cardoso and Montoya-
Burgos, 2009). Similarly, Nijssen (1970) suggests a seasonal portal between the
Sipalawini River (Corantijn River basin) and the Paru do Oeste River (Amazon River
basin) across the potentially flooded Sipalawini ? Paru Savanna. He used as support the
range of Corydoras bondi bondi, which is found through much of Suriname, the
Essequibo of Guyana, and the Yuruari (Cuyuni-Essequibo) of Venezuela; however, given
the westward extent of C. bondi bondi?s range, the proto-Berbice or Eastern Atlantic
Coastal Corridor might provide a better explanation. Regardless, Nijssen (1970) describes
a variety of potential corridors between north-northeast flowing Atlantic Coastal rivers
and south-flowing Amazon Rivers, and strong support for the transit of at least the
species Pseudancistrus brevispinnis through these corridors was found by Cardoso and
Montoya-Burgos (2009).
Availability of the Eastern Atlantic Coastal Corridor as a means of distribution
between mouths of the Essequibo and the Amazon is suggested by ranges of Curimata
cyprinoids, which is a lowland species that ranges throughout Atlantic Coast drainages
168
from the Orinoco to the Amazon (Vari, 1987), Parotocinclus britskii, which ranges
across Atlantic Coast drainages from the Essequibo to the Amazon (Schaefer and
Provenzano, 1993), and several serrasalmin species with ranges extending from the
Oyapock to the Amazon (J?gu and Keith, 1999).
4.4.7 Relictual Fauna
Inspired by Thurn?s (1885) first ascent of Mount Roraima, Doyle (1912) wrote his
fictional novel ?The Lost World? about a prehistoric landscape isolated atop a table
mountain and populated with ape men and dinosaurs. Although no such archaic member
of the terrestrial fauna has yet been discovered, the Guiana Shield does harbor at least one
aquatic taxon among the Loricarioidea that may have been swimming with dinosaurs of
the Cretaceous. The genus Lithogenes includes three species that currently comprise the
Lithogeninae of either the Astroblepidae of Loricariidae. Lithogenes is similar in external
appearance to basal astroblepids and loricariids (Schaefer and Provenzano, 2008), but has
a morphology so distinct that it does not fit comfortably into either of these loricarioid
families. Armbruster (2004, 2008) and Hardman (2005) hypothesize that Lithogenes is
sister to astroblepids, while Schaefer (2003) hypothesizes that the genus is sister to
loricariids. In a phylogeny with nodes dated by a fossil-calibrated relaxed molecular
clock, Lundberg et al. (2007) hypothesize that the split between the Astroblepidae and
Loricariidae occurred approximately 85-90 Ma (Lithogenes was not included in the
analysis). If Lithogenes is the sister to loricariids, it must also be at least 65?70 million
years old (age of deepest node in the Loricariidae), but if Lithogenes is sister to
169
astroblepids, it may be as young as 20 million years old (age of basal node in the
Astroblepidae).
Two Lithogenes species, L. villosus (Potaro-Essequibo) and L. wahari (Cuao-
Orinoco), are found in the Guiana Shield, and the third species, L. valencia, is thought to
be from the Lago Valencia drainage in the coastal mountains of northern Venezuela (date
and collector of the L. valencia type series are unknown and the species is currently
thought extinct; Provenzano et al., 2003). The disjunct distribution of L. villosus and L.
wahari on opposite sides of the western Guiana Shield is shared by a number of other
rheophilic taxa, and may be the product of sequential capture of proto-Berbice
headwaters by north- and west-flowing tributaries of the Orinoco. Dispersal via
headwater capture seems a likely avenue for Lithogenes, which live in clear, swift
flowing streams and have a specialized pelvic fin morphology adapted to climbing
vertical surfaces (Schaefer and Provenzano, 2008).
Other rheophilic loricariid taxa that seem to represent disjunct East-West relicts of
a more widespread proto-Berbice distribution include Lithoxus, Exastilithoxus,
Neblinichthys, and Harttia. Lithoxus is represented in the west by L. jantjae in the upper
Ventuari River (Orinoco) and in the east by L. lithoides, its putative sister species (Lujan,
2008) in the Essequibo, upper Branco, and Trombetas. Exastilithoxus, the sister of
Lithoxus, is represented in the west by E. hoedemani in the Maraui? River (upper Negro),
and in the east by E. fimbriatus in the upper Caroni. Neblinichthys is represented by N.
pillosus from the Baria River (lower Casiquiare) and by N. yekuana from tributaries of
the upper Caroni River. Harttia is represented by H. merevari in the upper Caura and
upper Ventuari Rivers and by H. platystoma in the Essequibo River.
170
The disjunct distribution of Lithogenes valencia in the Coastal Range, across the
Eastern Venezuela Basin from the Guiana Shield, is more difficult to explain. At no point
in the hydrologic history of the Eastern Venezuela Basin (see above) was there a period
in which high gradient habitat of the coastal mountain range seems to have been
contiguous with that of the Guiana Shield. Periods of low sea level during the Middle
Miocene eastward expansion of the Orinoco may be one period in which such contiguity
existed. Dispersal from the shield, across the Apure Llanos to the Merida Andes and from
there northeast via headwaters to the Coastal Mountain range represents another
possibility. Regardless, the genus seems to have had a much wider distribution at one
time, of which the three known localities represent relicts (Schaefer and Provenzano,
2008).
If Lithogenes is the sister lineage to astroblepids, a Guiana Shield origin is
indicated for this diverse group of Andean-restricted loricarioids. Likewise, competition
with and replacement by more highly derived astroblepids throughout the Merida Andes
provides a compelling explanation for the possible extirpation of Lithogenes from
Andean habitat between the Guiana Shield and the Coastal Mountains. The dispersal of
rheophilic taxa from the Guiana Shield to the Andes, followed by radiation along the
Andean flanks, is a pattern also apparent in the ancistrin clade comprising Chaetostoma,
Cordylancistrus, Dolichancistrus, Leptoancistrus, Exastilithoxus, Lithoxus, and New
Genus 2. New Genus 2 is known only from the upper Orinoco of Venezuela. It is sister to
two clades: Exastilithoxus + Lithoxus and the Chaetostoma group (Chaetostoma +
Cordylancistrus + Dolichancistrus + Leptoancistrus; Fig. 4; Armbruster, 2008).
171
Exastilithoxus and Lithoxus are endemic to the Guiana Shield, but all except two
species of Chaetostoma are distributed across Andean drainages ranging from Panama to
southeastern Peru. Species of Cordylancistrus, Dolichancistrus, and Leptoancistrus are
distributed largely across the Northern Andes of Panama, Colombia, and Venezuela,
although two species currently placed in Cordylancistrus (C. platycephalus and an
undescribed species) are known from the Napo and Mara?on of Ecuador and Peru.
Cordylancistrus torbesensis is basal within the Chaetostoma group (Fig. 4), and it hails
from southeastern slopes of the Merida Andes, across the Apure Llanos from the
northwestern corner of the Guiana Shield. The distribution of sister clades across the
Guiana Shield, a basal species in an adjacent region of the Andes, an intermediate
radiation in the Northern Andes, and derived taxa across the Andes from north to south,
supports a Guiana Shield origin for the Chaetostoma group. This largely Andean
radiation has even contributed two species back to the ancistrin fauna of the Guiana
Shield. Chaetostoma jegui and C. vasquezi are the only two non-Andean Chaetostoma,
and they are present on the respective southern and northern slopes of the western Guiana
Shield. Chaetostoma jegui is from the Uraricoera River (Branco) and C. vasquezi is from
the Caura and Caroni (Orinoco). The derived position of both these species within
Chaetostoma is supported by the presence of a fleshy excrescence (or keel) behind the
head (Armbruster, 2004; Salcedo, 2006).
Another possible relict in the Guiana Shield is the loricariid Corymbophanes
(Armbruster et al., 2000), which was found to be sister to all other hypostomines by
Armbruster (2004, 2008). There are two species of Corymbophanes, both known only
from the Potaro River above Kaieteur Falls where they are sympatric with Lithogenes
172
villosus, and live in habitats occupied elsewhere by members of the Ancistrini. No
ancistrins are present in the upper Potaro, likely because of their restriction to
downstream habitats by Kaieteur Falls, a 226 m drop in the Potaro river over a scarp of
the Guiana Shield uplifted in the Oligocene (Table 1).
The most basal loricariid subfamily (if Lithogenes is not a loricariid) is the
Delturinae (Montoya-Burgos et al., 1997; Armbruster, 2004; 2008), which is known only
from swift rivers of the southeastern Brazilian tributaries of the Brazilian Shield (Reis et
al., 2006). With Lithogenes and the Delturinae in shield regions, it could be speculated
that at least the loricariids (or loricariids + astroblepids) originated in the shields and
subsequently spread through the rest of northern South and southern Central America.
We doubt that the current ranges of the Lithogeninae and Delturinae represent the full
historical distributions of these taxa, and suggest that the modern ranges represent
relictual distributions. The fact that these two basal taxa are on opposite sides of the
shield regions suggest that they or their ancestors had a range that may have included at
least both shields. The origin of the Loricariidae in the shield regions is consistent with
the hypothesis (e.g.,, Galvis, 2006) that shield areas were the most concentrated areas of
high gradient aquatic habitat prior to significant uplift of the Andes. Loricariids share
many elements of their highly-derived morphology with rheophilic specialist taxa in
other parts of the world (e.g.,, sucker-like mouth with Garra and balitorid species in Asia
and Chiloglanis in Africa, and encapsulated swimbladders with Glyptothorax,
Glyptosternum and Pseudecheneis in Asia; Hora, 1922), indicating the selective pressures
required for origination of these structures and supporting the origin of Loricariidae in
high gradient habitats (Schaefer and Provenzano, 2008).
173
4.5 CONCLUSIONS
The Prone-8 biogeographic patterns of the Guiana Shield, coupled with more
ancient drainage patterns within the Amazon-Orinoco basins, provide a conceptual
framework upon which to build phylogeographic hypotheses for stream organisms in
northern South America. The Guiana Shield is not merely an island of upland habitat, but
shares extensive biogeographic connections with upland habitats of the Brazilian Shield,
the Andes, and the Coastal Mountains. Distributions of loricariid taxa suggest that these
connections to other areas have been important, but that within the Guiana Shield there
has been little mixing of upland faunas via the Western Atlantic Coastal, and
Caroni?Cuyuni/Mazaruni corridors. Most distributions within the Guiana Shield can be
explained via currently contiguous rivers, stream capture events in the uplands of larger
systems, and/or ancient river systems such as the proto-Berbice.
Because of temporal fluctuations in these connections, and their differential use
by various taxa, there is no single hypothesis explaining biogeographic patterns across
the Guiana Shield and neighboring uplands. We present a null hypothesis for
biogeographic patterns based solely on our descriptions of basin evolution and geologic
evidence of historical watershed boundaries (Fig. 5). Differential use of modern corridors
of the Prone-8 can obscure these relationships, however, and give rise to a variety of
divergent phylogeographic patterns. The disjunct distribution of Lithogenes, for example,
could represent relicts from a broader Oligocene distribution, or could be due to more
recent distribution via the proto-Orinoco, proto-Caura, or proto-Berbice. Figure 6
provides examples of three such alternative phylogeographic patterns.
174
Multiple biogeographic hypotheses described herein work for most taxa of the
Guiana Shield, but no single explanation works for all taxa. Diverse, species-level
phylogenies will be required to work out the timing and relative importance of proposed
corridors. Further investigations of Guiana Shield biogeographic patterns will require
genetic datasets that can be subjected to molecular-clock analyses, and studies of upland
taxa are especially important because of their frequently smaller ranges, and
corresponding potential for finer-scale resolution. The timing and dispersal rates of the
Chaetostoma group, for example, from the Guiana Shield to the Andes and back again
offer an intriguing opportunity to understand not only the relative importance of the
Andes as a novel upland habitat, but also more general mechanisms of upland fish
dispersal and evolutionary radiation. In today?s advanced age of scientific understanding,
the discoveries of primitive taxa and ancient biogeographic patterns still waiting to be
made among fishes of the Guiana Shield are just as exciting as those fictionalized in
Doyle?s (1912) ?The Lost World.?
175
4.6 REFERENCES
Armbruster, J. W. (2003). The species of the Hypostomus cochliodon group
(Siluriformes: Loricariidae). Zootaxa 249, 1?60.
Armbruster, J. W. (2004). Phylogenetic relationships of the suckermouth armoured
catfishes (Loricariidae) with emphasis on the Hypostominae and the Ancistrinae.
Zoological Journal of the Linnean Society 141, 1?80.
Armbruster, J. W. (2005). The loricariid catfish genus Lasiancistrus (Siluriformes) with
descriptions of two new species. Neotropical Ichthyology 3, 549?569.
Armbruster, J. W. (2008). The genus Peckoltia with the description of two new species
and a reanalysis of the phylogeny of the genera of the Hypostominae
(Siluriformes: Loricariidae). Zootaxa 1822, 1?76.
Armbruster, J. W., Lujan, N. K. & Taphorn, D. C. (2007). Four new Hypancistrus
(Siluriformes: Loricariidae) from Amazonas, Venezuela. Copeia 2007, 62?79.
Armbruster, J. W., Sabaj, M. H., Hardman, M., Page, L. M. & Knouft, J. H. (2000).
Catfish of the genus Corymbophanes (Loricariidae: Hypostominae) with
description of one new species: Corymbophanes kaiei. Copeia 2000, 997?1006.
Armbruster, J. W. & Taphorn, D. C. (2008). A new species of Pseudancistrus from the
R?o Caron?, Venezuela (Siluriformes, Loricariidae). Zootaxa 1731, 33?41.
Bennett, K. D. (1990). Milankovitch cycles and their effects on species in ecological and
evolutionary time. Paleobiology 16, 11?21.
176
Berrang?, J. P. (1975). The geomorphology of southern Guyana with special reference to
the development of planation surfaces. Anais D?cima Confer?ncia Geol?gica
Interguianas 1, 804?824.
Boeseman, M. (1968). The genus Hypostomus Lac?p?de, 1803, and its Surinam
representatives (Siluriformes, Loricariidae). Zoologische Verhandelingen
(Leiden) 99, 1?89.
Boeseman, M. (1982). The South American mailed catfish genus Lithoxus Eigenmann,
1910, with the description of three new species from Surinam and French Guyana
and records of related species (Siluriformes, Loricariidae). Proceedings of the
Koninklijke Nederlandse Akademie van Wetenschappen - Series C: Biological &
Medical Sciences 85, 41?58.
Brice?o, H. O. & Schubert, C. (1990). Geomorphology of the Gran Sabana, Guayana
Shield, southeastern Venezuela. Geomorphology 3, 125?141.
Brown, Jr., K. S. & Ab'S?ber, A. N. (1979). Ice age forest refuges and evolution in the
Neotropics: Correlation of paleoclimatological, geomorphological and
pedological data with modern biological endemism. Paleoclimas 5, 1?30.
Buckup, P. A. (1993). Review of the characidiin fishes (Teleostei: Characiformes), with
descriptions of four new genera and ten new species. Ichthyological Exploration
of Freshwaters 4, 97?154.
Chernoff, B., Machado-Allison, A. & Saul, W. G. (1991). Morphology, variation and
biogeography of Leporinus brunneus (Pisces: Characiformes: Anostomidae).
Ichthyological Exploration of Freshwaters 1, 295?306.
177
Cardoso, Y. P. & Motoya-Burgos, J. I. (2009). Unexpected diversity in the catfish
Pseudancistrus brevispinis reveals dispersal routes in a Neotropical center of
endemism: the Guyanas Region. Molecular Ecology 18, 947?964.
Crawford, F. D., Szelewski, C. E. & Alvey, G. D. (1985). Geology and exploration in the
Takutu Graben of Guyana and Brazil. Journal of Petroleum Geology 8, 5?36.
D?az de Gamero, M. L. 1996. The changing course of the Orinoco River during the
Neogene: a review. Palaeogeography, Palaeoclimatology, Palaeoecology 123,
385?402.
Dobson, D. M., Dickens, G. R. & Rea, D. K. (2001). Terrigenous sediment on Ceara
Rise: a Cenozoic record of South American orogeny and erosion.
Palaeogeography, Palaeoclimatology, Palaeoecology 165, 215?229.
Doyle, A. Conan-. (1912). The Lost World. London: Megaelaon.
Edmond, J. M., Palmer, M. R., Measures, C. I., Grant, B. & Stallard, R. F. (1995). The
fluvial geochemistry and denudation rate of the Guayana Shield in Venezuela,
Colombia, and Brazil. Geochimica et Cosmochimica Acta 59, 3301?3325.
Eigenmann, C. H. (1912). The freshwater fishes of British Guiana, including a study of
the ecological grouping of species, and the relation of the fauna of the plateau to
that of the lowlands. Memoirs of the Carnegie Museum 5, 1?578.
Galvis, G. (2006). La regi?n amaz?nica. In Peces del Medio Amazonas Regi?n de Leticia
(Galvis, J. I. M. G., Duque, S. R., Castellanos, C., S?nchez-Duarte, P., Arce, M.,
Guti?rrez, A., Jim?nez, L. F., Santos, M., Vejarano, S., Arbel?ez, F., Prieto, E.,
Leiva, M. eds.), pp. 28-47. Bogot?, Colombia: Conservation International -
Colombia.
178
Gansser, A. (1954). The Guiana Shield (S. America) geological observation. Eclogae
Geologicae Helvetiae 47, 77?112.
Gansser, A. (1974). The Roraima problem (South America). Verhandlungen der
Naturforschenden Gesellschaft 84, 80?100.
Garner, H. F. (1966). Derangement of the Rio Caroni, Venezuela. Revue de
G?omorphologie Dynamique 16, 54?83.
Gaudette, H. E. & Olszewski, W. J., Jr. (1985). Geochronology of the basement rocks,
Amazonas Territory, Venezuela and the tectonic evolution of the western Guiana
Shield. Geologie en Mijnbouw 64, 131?143.
Gayet, M., Sempere, T.,Cappetta, H., Jaillard, E. & L?vy, A. (1993). La pr?sence de
fossiles marins dans le Cr?tac? terminal des Andes centrales et ses cons?quences
pal?og?ographiques. Palaeogeography, Palaeoclimatology, Palaeoecology 102,
283?319.
G?ry, J. & Zarske, A. (2002). Derhamia hoffmannorum gen. et sp. n. ? a new pencil fish
(Teleostei, Characiformes, Lebiasinidae), endemic from the Mazaruni River in
Guyana. Zoologische Abhandlungen 52, 35?47.
Ghosh, S. K. (1985). Geology of the Roraima Group and its implications. Memoria
Simposium Amaz?nico, 1st, Venezuela, 1981: Caracas, Venezuela, Direci?n
General Sectorial de Minas y Geologia, Publicaci?n Especial 10, 33?50.
Gibbs, A.K. & Barron, C.N. (1993). The Geology of the Guiana Shield. New York:
Oxford University Press.
Haffer, J. (1969). Speciation in Amazonian forest birds. Science 165, 131?137.
179
Haffer, J. (1997). Alternative models of vertebrate speciation in Amazonia: an overview.
Biodiversity and Conservation 6, 451?476.
Hammen, T. van der. (1972). Changes in vegetation and climate in the Amazon Basin
and surrounding areas during the Pleistocene. Geologie en Mijnbouw 51,
641?643.
Hammond, D. S. (2005). Biophysical features of the Guiana Shield. In Tropical Forests
of the Guiana Shield (Hammond, D. S. ed.), pp. 15?194. Cambridge: CABI.
Hardman, M. (2005). The phylogenetic relationships among non-diplomystid catfishes as
inferred from mitochondrial cytochrome b sequences; the search for the ictalurid
sister taxon (Otophysi: Siluriformes). Molecular Phylogenetics and Evolution 37,
700?720.
Hedberg, H. D. (1950). Geology of the Eastern Venezuelan Basin (Anzoategui-Monagas-
Sucre-eastern Guarico portion). Bulletin of the Geological Society of America 61,
1173?1216.
Hoorn, C. (1993). Marine incursions and the influence of Andean tectonics on the
Miocene depositional history of northwestern Amazonia: results of a
palynostratigraphic study. Palaeogeography, Palaeoclimatology, Palaeoecology
105, 267?309.
Hoorn, C., J. Guerrero, Sarmiento, G. A. & Lorente, M. A. (1995). Andean tectonics as a
cause for changing drainage patterns in Miocene northern South America.
Geology 23, 237?240.
Hora, S. L. (1922). Structural modifications in the fish of mountain torrents. Records of
the Indian Museum 24, 31?61.
180
Hubert, N. & Renno, J.-F. (2006). Historical biogeography of South American freshwater
fishes. Journal of Biogeography 33, 1?23.
J?gu, M. & Keith, P. (1999). Le bas Oyapock limite septentrionale ou simple ?tape dans
la progression de la faune des poissons d'Amazonie occidentale. Life Sciences
322, 1133?1143.
Kasper, D. C. & Larue, D. K. (1986). Paleogeographic and tectonic implications of
quartzose sandstones of Barbados. Tectonics 5, 837?854.
Kelloff, C. L. & Funk, V. A. (2004). Phytogeography of the Kaieteur Falls, Potaro
Plateau, Guyana: floral distributions and affinities. Journal of Biogeography 31,
501?513.
Kiser, G. D. & Bass, I. (1985). La reorientacion del Arco de El Baul y su importancia
economica. Memoria Sociedad Venezolana de Geologos 8, 5122?5135.
Krock, L. (1969). Climate and sedimentation in the Guianas during the last glacial and
the Holocene. Proceedings of the Eighth Guiana Geological Conference,
Georgetown, Guyana 18, 1?16.
Kullander, S. O. (1990). Mazarunia mazarunii (Teleostei: Cichlidae), a new genus and
species from Guyana, South America. Ichthyological Exploration of Freshwaters
1, 4?14.
Lasso, C. A., Machado-Allison, A. & Hernandez, R. P. (1990). Consideraciones
zoogeograficas de los peces de La Gran Sabana (Alto Caroni) Venezuela, y sus
relaciones con las cuencas vecinas. Memoria Sociedad de Ciencias Naturales La
Salle 20, 109?129.
181
Latrubesse, E. M. & Franzinelli, E. (2005). The late Quaternary evolution of the Negro
River, Amazon, Brazil: Implications for island and floodplain formation in large
anabranching tropical systems. Geomorphology 70, 372-397.
Le Bail, P.-Y., Keith, P. & Planquette, P. (2000). Atlas des poissons d'eau douce de
Guyane. Tome 2, fascicule II: Siluriformes. Paris : Patrimoines naturels
(M.N.H.N./S.P.N.).
Lewis, W., Jr. & Weibezahn, F. (1981). The chemistry and phytoplankton of the Orinoco
and Caroni Rivers, Venezuela. Archiv f?r Hydrobiologie 91, 521?528.
Lewis, W., Jr. & Saunders, J. F., III. (1990). Chemistry and element transport by the
Orinoco main stem and lower tributaries. In El R?o Orinoco Como Ecosistema
(Weibezahn, F.H., Alvarez, H. & Lewis, W. M., eds), pp. 55?80. Caracas,
Venezuela: Fondo Editorial Acta Cient?fica Venezolana, Universidad Sim?n
Bol?var.
L?pez, V. M., Mencher, E. & Brineman, J. H., Jr. (1942). Geology of Southeastern
Venezuela. Bulletin of the Geological Society of America 53, 849?872.
Lovejoy, N.R. & De Ara?jo, M. L. G. (2000). Molecular systematics, biogeography and
populations structure of Neotropical freshwater needlefishes of the genus
Potamorrhaphis. Molecular Ecology 9, 259?268.
Lowe-McConnell, R. H. (1964). The fishes of the Rupununi savanna district of British
Guiana, South America. Journal of the Linnean Society (Zoology) 45, 103?144.
Lujan, N. K. (2008). Description of a new Lithoxus (Siluriformes: Loricariidae) from the
Guayana Highlands with a discussion of Guiana Shield biogeography.
Neotropical Ichthyology 6, 413?418.
182
Lujan, N. K., Arce, M. & Armbruster, J. W. 2009. A new black Baryancistrus with blue
sheen from the upper Orinoco (Siluriformes: Loricariidae). Copeia 2009, 50?56.
Lundberg, J. G. (1998). The temporal context for the diversification of neotropical fishes.
In Phylogeny and Classification of Neotropical Fishes (Malabarba, L. R., Reis,
R.E., Vari, R.P., Lucena, Z.M. & Lucena, C.A.S., eds.), pp. 49?68. Porto Alegre:
Epipucrs.
Lundberg, J. G., Marshall, L. G., Guerrero, J., Horton, B., Claudia, M., Malabarba, S.L.
& Wesselingh, F. (1998). The stage for neotropical fish diversification: a history
of tropical South American rivers. In Phylogeny and Classification of Neotropical
Fishes (Malabarba, L. R., Reis, R.E., Vari, R.P., Lucena, Z.M. & Lucena, C.A.S.,
eds.), pp. 13?48. Porto Alegre: Epipucrs.
Lundberg, J. G., Sullivan, J. P., Rodiles-Hern?ndez, R. & Hendrickson, D. A. (2007).
Discovery of African roots for the Mesoamerican Chiapas catfish, Lacantunia
enigmatica, requires an ancient intercontinental passage. Proceedings of the
Academy of Natural Sciences of Philadelphia 156, 39?53.
Maisey, J. G. (2000). Continental break up and the distribution of fishes of Western
Gondwana during the Early Cretaceous. Cretaceous Research 21, 281?314.
Mattox, G. M. T., Toledo-Piza, M. & Oyakawa, O. T. (2006). Taxonomic study of
Hoplias aimara (Valenciennes, 1846) and Hoplias macrophthalmus (Pellegrin,
1907) (Ostariophysi, Characiformes, Erythrinidae). Copeia 2006, 516?528.
McConnell, R. B. (1959). The Takutu Formation in British Guiana and the probable age
of the Roraima Formation. In Transactions of the Second Conferencia Geologica
del Caribe, pp. 163?170. University of Puerto Rico.
183
Montoya-Burgos, J.-I., Muller, S., Weber, C. & Pawlowski, J. (1997). Phylogenetic
relationsips between Hypostominae and Ancistrinae (Siluroidei: Loricariidae):
first results from mitochondrial 12S and 16S rRNA gene sequences. Revue suisse
de Zoologie 104, 165?198.
Moyer, G. R., Winemiller, K. O., McPhee, M. V. & Turner, T. F. (2005). Historical
demography, selection, and coalescence of mitochondrial and nuclear genes in
Prochilodus species of Northern South America. Evolution 59, 599?610.
Nijssen, H. (1970). Revision of Surinam catfishes of the genus Corydoras Lac?p?de,
1803 (Pisces, Siluriformes, Callichthyidae). Beaufortia 18, 1?75.
Noonan, B. P. & Gaucher, P. (2005). Phylogeography and demography of Guianan
harlequin toads (Atelopus): diversification within a refuge. Molecular Ecology 14,
3017?3031.
Noonan, B. P. & Gaucher, P. (2006). Refugial isolation and secondary contact in the
dyeing poison frog Dendrobates tinctorius. Molecular Ecology 15, 4425?4435.
Pindell, J. L., Higgs, R. & Dewey, J. F. (1998). Cenozoic palinspaztic reconstruction,
paleogeographic evolution and hydrocarbon setting of the northern margin of
South America. Society of Economic Paleontologists and Mineralogists Special
Publication 58, 45?85.
Power, M. E. (1984). Habitat quality and the distribution of algae-grazing catfish in a
Panamanian stream. Journal of Animal Ecology 53, 357?374.
Prance, G. T. (1973). Phytogeographic support for the theory of Pleistocene forest
refuges in the Amazon Basin, based on evidence from the distribution patterns in
184
Caryocaraceae, Chrysobalanaceae, Dichapetalaceae and Lecythidaceae. Acta
Amazonica 3, 5?28.
Presswell, B., Weitzman, S. H. & Bergquist, T. (2000). Skiotocharax meizon, a new
genus and species of fish from Guyana with a discussion of its relationships
(Characiformes: Crenuchidae). Ichthyological Exploration of Freshwaters 11,
175?192.
Provenzano R., F., Schaefer, S. A., Baskin, J. N. & Royero-Leon, R. (2003). New,
possibly extinct lithogenine loricariid (Siluriformes, Loricariidae) from Northern
Venezuela. Copeia 2003, 562?575.
Putzer, H. (1984). The geological evolution of the Amazon basin and its mineral
resources. In The Amazon. Limnology and landscape ecology of a mighty tropical
river and its basin (Sioli, H., ed.), pp. 15?46. Dordrecht: Dr W. Junk Publishers.
R?s?nen, M. E., Salo, J. S. & Kalliola, R. J. (1998). Fluvial perturbance in the Western
Amazon Basin: Regulation by long-term sub-Andean tectonics. Science 238,
1398?1401.
Reis, R. E., Pereira, E. H. L. & Armbruster, J. W. (2006). Delturinae, a new loricariid
catfish subfamily (Teleostei, Siluriformes), with revisions of Delturus and
Hemipsilichthys. Zoological Journal of the Linnean Society 147, 277?299.
Renno, J.F., Berrebi, P., Boujard, T. & Guyomard, R. (1990). Intraspecific genetic
differentiation of Leporinus friderici (Anostomidae, Pisces) in French Guiana and
Brazil: a genetic approach to the refuge theory. Journal of Fish Biology 36,
85?95.
185
Renno, J.F., Machardom, A., Blanquer, A. & Boursot, P. (1991). Polymorphism of
mitochondrial genes in populations of Leporinus friderici (Bloch, 1794):
intraspecific structure and zoogeography of the Neotropical fish. Genetica 84,
137?142.
Rohr, G.M. (1991). Paleogeographic maps, Maturin Basin of E. Venezuela and Trinidad.
In Transactions of the 2nd Geological Conference of the Geological Society of
Trinidad & Tobago (Gillezeau, K. A., ed.), pp. 88?105. GSST.
Sabaj, H. H., Taphorn, D. C. & Castillo G., O. E. (2008). Two new species of thicklip
thornycats, genus Rhinodoras (Teleostei: Siluriformes: Doradidae). Copeia 2008,
209?226.
Salcedo, N. J. (2006). New species of Chaetostoma (Siluriformes: Loricariidae) from
Central Peru. Copeia 2006, 60?67.
Santos, J. O. S., Potter, P. E., Reis, N. J., Hartmann, L. A., Fletcher, I. R. & McNaughton,
N. J. (2003). Age, source, and regional stratigraphy of the Roraima Supergroup
and Roraima-like outliers in northern South America based on U-Pb
geochronology. Geological Society of America Bulletin 115, 331?348.
Schaefer, C. E. R. & do Vale, J. F., Jr. (1997). Mudan?as clim?ticas e evolu??o da
paisagem em Roraima: uma resenha do Cret?ceo ao Recente. In Homem,
Ambiente e Ecologia na Estado de Roraima (Barbosa, R. I., Ferreira, E. J. G. &
Castell?n, E. G. eds.), pp. 231?265 Manaus, Brazil: INPA.
Schaefer, S. A. (2003). Relationships of Lithogenes villosus Eigenmann, 1909
(Siluriformes, Loricariidae): evidence from high-resolution computed
microtomography. American Museum Novitates 3401, 1?26.
186
Schaefer, S. A. & Provenzano R., F. (1993). The Guyana Shield Parotocinclus:
systematics, biogeography, and description of a new Venezuelan species
(Siluroidei: Loricariidae). Ichthyological Exploration of Freshwaters 4, 39?56.
Schaefer, S. A. & Provenzano R., F. (2008). The Lithogininae (Siluriformes,
Loricariidae): anatomy, interrelationships, and description of a new species.
American Museum Novitates 3637,1-49.
Schubert, C. (1988). Climatic changes during the last glacial maximum in northern South
America and the Caribbean: a review. Interciencia 13, 128?137.
Schubert, C., Brice?o, H. O. & Fritz, P. (1986). Paleoenvironmental aspects of the
Caroni-Paragua River Basin (Southeastern Venezuela). Interciencia 11, 278?289.
Shagam, R., Kohn, B. P., Banks, P. O., Dasch, L. E., Vargas, R., Rodr?guez, G. I. &
Pimentel, N. (1984). Tectonic implications of cretaceous-Pliocene fission-track
ages from rocks of the circum-Maracaibo Basin region of western Venezuela and
eastern Colombia. Geological Society of America Memoir 162, 385?412.
Sinha, N. K. P. (1968). Geomorphic evolution of the Northern Rupununi Basin, Guyana,
McGill University Savanna Research Project, Savanna Research Series 11.
Montreal: McGill University.
Sioli, H. (1964). General features of the limnology of Amazonia. Verhandlungen des
Internationalen Verein Limnologie 15, 1053?1058.
Stern, K. M. (1970). Der Casiquiare-Kanal, einst und jetzt. Amazoniana 2, 401?416.
Sternberg, H. O. (1975). The Amazon River of Brazil. Geographische Zeitschrift 40,
1?74.
187
Taphorn B., D. C., L?pez-Fern?ndez, H. & Bernard, C. R. (2008). Apareiodon agmatos, a
new species from the upper Mazaruni river, Guyana (Teleostei: Characiformes:
Parodontidae). Zootaxa 1925, 31?38.
Thornes, J. B. (1969). Variability in specific conductance and pH in the
Casiquiare?Upper Orinoco. Nature 221, 461?462.
Thurn, E. im. (1885). The ascent of Mount Roraima. Proceedings of the Royal
Geographical Society and Monthly Record of Geography, New Monthly Series 7,
497?521.
Tricart, J. (1985). Evidence of Upper Pleistocene dry climates in northern South America.
In Environmental Change and Tropical Geomorphology (Douglas, T. S. I., British
Geomorphological Research Group, eds.), pp. 197-217. London: Allen and
Unwin.
Turner, T. F., McPhee, M. V., Campbell, P. & Winemiller, K. O. (2004). Phylogeography
and intraspecific genetic variation of prochilodontid fishes endemic to rivers of
northern South America. Journal of Fish Biology 64, 186?201.
Vanzolini, P. E. (1973). Paleoclimates, relief, and species multiplication in equatorial
forests. In Tropical Forest Ecosystems in Africa and South America (Meggers, B.
J., Ayensu, E. S. & Duckworth, W. D. eds.), pp. 255-258. Washington:
Smithsonian Institution Press.
Vari, R. P. (1987). The Curimatidae, a lowland neotropical fish family (Pisces:
Characiformes); distribution, endemism, and phylogenetic biogeography. In
Proceedings of a Workshop on Neotropical Distribution Patterns (Vanzolini, P.
188
E. & Heyer, W. R., eds.), pp. 343-377. Rio de Janeiro: Academia Brasileira de
Ci?ncias.
Villamil, T. (1999). Campanian-Miocene tectonostratigraphy, depocenter evolution and
basin development of Colombia and western Venezuela. Palaeogeography,
Palaeoclimatology, Palaeoecology 153, 239?275.
Weitzman, S. H. & Cobb, J. S. (1975). A revision of the South American fishes of the
genus Nannostomus G?nther (Family Lebiasinidae). Smithsonian Contributions to
Zoology 186, 1?36.
Weitzman, S. H. & Weitzman, M. (1982). Biogeography and evolutionary diversification
in neotropical freshwater fishes, with comments on the refuge theory. In
Biological Diversification in the Tropics (Prance, G. T., ed.), pp. 403?422. New
York: Columbia University Press.
Werneke, D. C., Armbruster, J. W., Lujan, N. K. & Taphorn, D. C. (2005).
Hemiancistrus guahiborum, a new suckermouth armored catfish from Southern
Venezuela (Siluriformes: Loricariidae). Neotropical Ichthyology 3, 543?548.
Willis, S. C., Nunes, M. S., Monta?a, C. G., Farias, I. P. & Lovejoy, N. R. (2007).
Systematics, biogeography, and evolution of the Neotropical peacock basses
Cichla (Perciformes: Cichlidae). Molecular Phylogenetics and Evolution 44,
291?307.
Winemiller, K. O., Fern?ndez, H. L., Taphorn, D. C., Nico, L. G. & Barbarino Duque, A.
(2008). Fish assemblages of the Casiquiare River, a corridor and zoogeographic
filter for dispersal between the Orinoco and Amazon basins. Journal of
Biogeography 35, 1551?1563.
189
Table 1. Planation surfaces, their age, elevation, and name in each country of the Guiana Shield (after Schubert et al., 1986,
Briceno and Schubert, 1990, Gibbs and Barron, 1993).
Age of UpliftCountrySurfaceElevation (m asl)
Pre-Late CretaceousVenezuelaAuyantepui2000-2900
BrazilRoraima Sedimentary Plateau1000-3000
Pre-Late CretaceousVenezuelaKamarata-Pakaraima1000-1200
GuyanaKanuku 900-1200
BrazilGondwana900-1200
Late Cretaceous-PaleoceneVenezuelaImataca-Nuria600-700
GuyanaKopinang600-700
SurinameE.T.S.-Brownsberg700-750
French GuianaFirst Peneplain525-550
BrazilSul-Americana700-750
* no uplift from Lower Eocene to Lower Oligocene indicated by bauxite formations (McConnell, 1968)
Oligocene-MioceneVenezuelaCaroni-Aro400-450
Guyana-NorthKaieteur 250-350
Guyana-SouthMarudi 400-500
SurinameLate Tertiary I300-400
French GuianaSecond to Third Peneplain200-370
BrazilEarly Velhas200-450
Plio-PleistoceneVenezuelaLlanos 80-150
Guyana-NorthRupununi110-160
Guyana-SouthKuyuwiniup to 200
189
190
SurinameLate Tertiary II80-150
French GuianaFourth Peneplain150-170
BrazilLate Velhas80-150
HoloceneVenezuelaOrinoco floodplain0-50
GuyanaMazaruni>80
SurinameQuaternary fluvial cycle0-50
 BrazilPargua?u0-50
190
191
Figure 1. Major rivers and drainage basins of the Guiana Shield: 1. Orinoco River, 2. Caroni River with Paragua River as its
eastern tributary, 3. Caura River, 4. Ventuari River, 5. Orinoco headwater rivers, from north to south: Padamo, Matacuni, Ocamo,
Orinoco, Mavaca, 6. Casiquiare Canal, 7. Siapa River, 8. Negro River, 9. Demini River, 10. Branco River, 11. Uatuma River, 12.
Trombetas River, 13. Paru do Oeste River, 14. Paru River, 15. Jari River, 16. Oyapok River, 17. Marone River, 18. Coppename
River to the west and Surinam River to the east, 19. Corentyne River, 20. Essequibo River, 21. Potaro River, 22. Cuyuni River, 23.
Uraricoera River, 24. Rupununi Savanna bordered on the west by the Takutu River and on the east by the Rupununi.
191
192
Figure 2. Schematic showing relationships among planation surfaces in Guyana, their historical contiguity (dashed lines) and their
modern remnants (solid lines). Elevation of each surface relative to contemporary sea level in meters on the left and feet on the
right (from McConnell, 1968).
192
193
Figure 3. The Prone-8: hypothesized areas of movement between basins of the Guiana Shield. Area of some connections are
approximate.
193
194
Figure 4. Phylogenetic hypothesis for the Chaetostoma-group (from Armbruster, 2008).
194
195
Figure 5. Null hypothesis of areal relationships among Guiana Shield fishes based only upon hydrologic history. Basal node
represents the historical continental divide between eastern and western drainages at the Purus Arch. Terminal nodes represent
modern river drainages with text color indicating major modern drainage basin: green = Amazon River, blue = Orinoco River.
Three major clades of modern river drainages (proto-Orinoco, proto-Berbice, and NE Atlantic Coast) represent historical
contiguity and regional affinities. Historical geologic and hydrologic events at internal nodes labeled accordingly.
195
196
Figure 6. Three different biogeographic hypotheses based on differential use of connections in the Prone-8. A. Hypothesis based
on current drainage patterns. B. Hypothesis if the Mazaruni-Caroni, Cuyuni- Caroni or Western Atlantic Coastal Connections were
used. C. Hypothesis considering the Casiquiare to be Orinoco in origin.
196
197
APPENDIX I. Citations and abstracts of published or submitted manuscripts.
Armbruster, J. W., L. A. Tansey, and N. K. Lujan. 2007. Hypostomus rhantos
(Siluriformes: Loricariidae), a new species from southern Venezuela. Zootaxa
1553:59?68.
Hypostomus rhantos is described for a uniquely pigmented species of loricariid catfish
from the upper R?o Orinoco of Amazonas, Venezuela. Hypostomus rhantos can be
separated from all other Hypostomus except H. micromaculatus by having its head and
dorsal and lateral surfaces of body densely covered in very small spots (greater than 15
spots on the first plate in the dorsal series of specimens less than 100 mm SL vs. less than
10; greater than 30 spots in specimens greater than 100 mm SL vs. less than 15). The new
species is distinguished from H. micromaculatus by having round spots (vs.
longitudinally oval) that are unordered (vs. in longitudinal lines), by having well-
developed keels on the lateral plates (vs. keels weak), by the presence of a ridge on the
pterotic that is contiguous with the supraorbital ridge (vs. pterotic ridge absent), and by
having the abdomen fully plated (vs. partially plated or naked).
Armbruster, J. W., N. K. Lujan, and D. C. Taphorn. 2007. Four new Hypancistrus
(Siluriformes: Loricariidae) from Amazonas, Venezuela. Copeia 2007:62?79.
198
Hypancistrus contradens, H. debilittera, H. furunculus, and H. lunaorum are described
based on specimens from the upper R?o Orinoco of southern Venezuela. Hypancistrus
furunculus differs from other Hypancistrus based on color pattern: distinct dark oblique
stripes ending at posterior insertion of dorsal fin and vertical bands in caudal fin (vs.
oblique stripes ending at end of caudal fin in H. zebra and thin, indistinct, light-colored
bands and vermiculations on a dark background in H. debilittera) and color pattern dark
with white spots in H. contradens, H. inspector, and H. lunaorum. Hypancistrus
contradens and H. lunaorum differ from H. inspector by having the dorsal fin reaching
the adipose fin when adpressed (vs. not reaching), having spots on the head the same size
as the body or spots absent (vs. spots smaller on head) and by usually having 22?23 mid-
ventral plates (vs. 24); and from H. debilittera, H. furunculus, and H. zebra by lacking
bars, saddles, or stripes on the body and bands in the fins. Hypancistrus lunaorum differs
from H. contradens by having white spots on the body smaller than nasal aperture
diameter (vs. white spots larger than the nasal aperture diameter).
Lujan, N. K. 2008. Description of a new Lithoxus (Siluriformes: Loricariidae) from the
Guayana Highlands with a discussion of Guiana Shield biogeography. Neotropical
Ichthyology 6:413?418.
Lithoxus jantjae, new species, is described from above Tencua Falls in headwaters of the
Ventuari River, a white- to clearwater river flowing west from the Maigualida and Parima
mountains in the Guayana Highlands of southern Venezuela. Lithoxus jantjae represents a nearly
600 km westward range expansion for a genus historically known only from Guyana, Suriname,
French Guiana, and Brazil. Lithoxus jantjae shares with other species of Lithoxus a
199
dorsoventrally depressed body and a large, papilose oral disk with small toothcups and few teeth.
It can be distinguished from congeners by a unique combination of characters including 12
branched caudal-fin rays, medial premaxillary tooth cusps enlarged, and a convex posterior
margin of the adipose-fin membrane. With the discovery of L. jantjae, Lithoxus becomes the
most recent example of a growing list of rheophilic loricariid genera with disjunct distributions
on east and west sides of the Guayana Highlands. A biogeographic hypothesis relying on the
existence of a proto-Berbice River uniting the southern Guayana Highlands with rivers of the
central Guiana Shield is advanced to partially explain the modern distribution of these species.
Lujan, N. K., and C. C. Chamon. 2008. Two new species of Loricariidae (Teleostei:
Siluriformes) from main channels of the upper and middle Amazon Basin, with
discussion of deep water specialization in loricariids. Ichthyological Exploration of
Freshwaters 19:271?282.
Hemiancistrus pankimpuju, new species, and Panaque bathyphilus, new species, are described
from the main channel of the upper (Mara?on) and middle (Solim?es) Amazon River
respectively. Both species are diagnosed by having nearly white bodies and fins, long
filamentous extensions of both simple caudal-fin rays, small eyes, absence of an iris operculum
and unique combinations of morphometrics. The coloration and morphology of these species,
unique within Loricariidae, are hypothesized to be apomorphies associated with life in the dark,
turbid depths of the Amazon mainstem. Extreme elongation of the caudal filaments in these and
a variety of other main channel catfishes is hypothesized to have a mechanosensory function
associated with predator detection.
Lujan, N. K., and J. W. Armbruster. 2009a. Geological and hydrological history of the
200
Guiana Shield and historical biogeography of its fishes. In press in J. Albert and R.
Reis, editors. Historical Biogeography of Neotropical Freshwater Fishes. Pfeil-
Verlag, Munich.
See Chapter 4.
Lujan, N. K., and J. W. Armbruster. 2009b. Two new ancistrin genera and species
(Siluriformes: Hypostominae) from the western Guiana Shield with discussion of
swimbladder and jaw morphological variation across the Loricariidae.
Ichthyological Exploration of Freshwaters in review.
Two new ancistrin genera and species are described from the upper Orinoco River watershed in
Amazonas, Venezuela. Micracanthicus vandragti is black with white spots and distinguished by
its small body-size, large swimbladder capsules, and highly protrusible lower jaws with short
tooth cups and five to eight long teeth. The known range of Micracanthicus vandragti is
restricted to the confluence of the Ventuari and Orinoco Rivers. Soromonichthys stearleyi is
green with small yellow-gold spots on the head and thin vertical bars on the body, and has long
jaw rami with 39 to 69 teeth. It is distinguished by its coloration and by its unique pattern of
snout deplatation (plates missing from mesethmoid surface, anteriormost margin of snout, and
small region posterior from anterolateral snout margin; plates present in column along either side
of mesethmoid). Soromonichthys stearleyi is known only from Soromoni Creek, a tributary of
the upper Orinoco draining southern slopes of Mount Duida. Phylogenetic analysis recovered
Micracanthicus at the base of the Acanthicus clade within the larger Panaque clade, and
Soromonichthys at the base of the Lithoxus clade within the larger Ancistrus clade.
Micracanthicus and Soromonichthys represent near opposite ends of the spectrum of jaw and
201
swimbladder morphologies in their respective clades, and are discussed in relation to variation
and functional evolution in these characters across Loricariidae.
Lujan, N. K., J. W. Armbruster, and M. H. Sabaj. 2007. Two new species of Pseudancistrus
from southern Venezuela (Siluriformes: Loricariidae). Ichthyological Exploration of
Freshwaters 18:163?174.
Two new species of the loricariid genus Pseudancistrus are described from the upper R?o
Orinoco and R?o Negro in Southern Venezuela. Pseudancistrus pectegenitor was collected in the
main channel of the R?o Orinoco near the mouth of the R?o Ventuari and in the middle reaches of
the R?o Casiquiare. It differs from congeners by having 10-11 dorsal-fin rays (vs. seven),
adpressed cheek odontodes reaching to three or more plates beyond the opercle in adults (vs.
maximally to rear edge of the opercle), plates of ventral row of caudal peduncle with dorsal
laminae strongly concave, accentuating the medial keel of the ventral plate row (shared with P.
sidereus), and large oral papillae internal to the dentary tooth cup (shared with P. coquenani, P.
orinoco, and P. yekuana). Pseudancistrus yekuana is known only from the type locality,
immediately upstream of Salto Tencua in the upper R?o Ventuari. It differs from congeners by
having large oral papillae internal to the dentary tooth cup (shared with P. coquenani, P. orinoco,
and P. pectegenitor), lower lip reaching to middle of pectoral girdle (vs. to anterior edge of
pectoral girdle), pectoral-fin spine maximally reaching posterior base of the pelvic-fin spine
when adpressed ventral to the pelvic fin (vs. at least halfway through pelvic-fin insertion) and by
several morphometric differences.
Lujan, Nathan K., M. Arce, and Jonathan W. Armbruster. 2009. A new black
Baryancistrus with blue sheen from the upper Orinoco (Siluriformes: Loricariidae).
202
Copeia 2009:50?56.
Baryancistrus beggini, new species, is described from the upper R?o Orinoco and lower portions
of its tributaries, the R?o Guaviare in Colombia and R?o Ventuari in Venezuela. Baryancistrus
beggini is unique within Hypostominae in having a uniformly dark black to brown base color
with a blue sheen in life, and the first three to five plates of the midventral series strongly bent
forming a distinctive keel above the pectoral fins along each side of the body. It is further
distinguished by having a naked abdomen, two to three symmetrical and ordered predorsal plate
rows including the nuchal plate, and the last dorsal-fin ray adnate with adipose fin via a posterior
membrane that extends beyond the preadipose plate up to half the length of the adipose-fin spine.
Rengifo, B., N. K. Lujan, D. Taphorn, and P. Petry. 2008. A new species of Gelanoglanis
(Siluriformes: Auchenipteridae) from the Mara?on River (Amazon Basin),
northeastern Peru. Proceedings of the Academy of Natural Sciences of Philadelphia
157:181?188.
We describe a new species of driftwood catfish, Gelanoglanis travieso, (Siluriformes:
Auchenipteridae) from the Mara?on River, a whitewater tributary of the Amazon River in
northeastern Per?. It shares with the two described species in this genus, G. stroudi, from left
bank whitewater tributaries of the Orinoco River in Colombia and Venezuela, and G.
nanonocticolus from blackwater tributaries of the upper Orinoco and Negro Rivers in Amazonas,
Venezuela and northern Brazil, the following synapomorphies: reduced size, compressed body,
conical snout, a single pair of mental barbels, premaxillae widely separated at rostral border of
upper jaw, premaxillary and dentary tooth patches narrow, posterior naris long and narrow and
positioned immediately anterior to orbit, and small eyes. Gelanoglanis travieso differs from all
203
congeners in having second dorsal-fin lepidotrichium filamentous, simple, not a spine, and not
serrate (shared with G. nanonocticolus); pectoral-fin spine stout, serrate along posterior margin
(shared with G. stroudi); and a terminal mouth (vs. subterminal in G. nanonocticolus and G.
stroudi).
Werneke, D. C., J. W. Armbruster, N. K. Lujan, and D. C. Taphorn. 2005. Hemiancistrus
guahiborum, a new suckermouth armored catfish from Southern Venezuela
(Siluriformes: Loricariidae). Neotropical Ichthyology 3:543?548.
Hemiancistrus guahiborum, new species, is described from the Orinoco River drainage of
Venezuela. Hemiancistrus guahiborum can be separated from all other Hemiancistrus
and all Peckoltia except P. braueri and P. cavatica by having an orange edge to the
dorsal and caudal fins. Hemiancistrus guahiborum can be separated from Peckoltia
cavatica and P. braueri by having the dorsal fin with separated light spots or uniformly
colored (vs. with dark spots forming bands) and the sides either solidly colored or with
tan blotches (vs. with dark dorsal saddles).
Werneke, D. C., M. H. Sabaj, N. K. Lujan, and J. W. Armbruster. 2005. Baryancistrus
demantoides and Hemiancistrus subviridis, two new uniquely colored species of
catfishes from Venezuela (Siluriformes: Loricariidae). Neotropical Ichthyology
3:533?542.
Baryancistrus demantoides and Hemiancistrus subviridis are two new species of uniquely
pigmented loricariids from southern Venezuela with an olive ground coloration and white to
cream-colored or golden-yellow spots. Baryancistrus demantoides is known only from the upper
204
r?o Orinoco drainage while H. subviridis is also known from the r?o Casiquiare drainage. In
addition to its coloration, B. demantoides can be distinguished from all other ancistrins by having
dorsal and adipose fins connected by an expanded posterior section of the dorsal-fin membrane,
lemon-colored spots confined to the anterior portion of the body, and greater than 30 teeth per
jaw ramus. H. subviridis can be separated from all other ancistrins by lacking a connection
between the dorsal and adipose fins and by having fewer than 30 teeth per jaw ramus.
205
APPENDIX II. Vectors from loricariid assemblage ?
13
C and ?
15
N centroids to species
means. Data from 19 localities and 83 species pooled and sorted according to taxon.
When data permit, direction of mean vector for each taxon shown as dashed radius, and
95% circular confidence intervals are shown as either dashed (significant, p<0.05) or
dotted (not significant, p?0.05) arcs. Vectors for species represent individual samples.
Vectors for all higher taxa represent species means.
Figure 1. Hypostominae: Ancistrini, Hypostomini, Pterygoplichthyini; and
Hypoptopomatinae: Hypoptopomatini.
206
Figure 2. Loricariinae: Loricariini, Harttiini, Farlowellini.
207
Figure 3. Hypostominae, Ancistrini: Ancistrus, Chaetostoma, Lasiancistrus, and
Dekeyseria.
208
Figure 4. Hypostominae, Ancistrini: Baryancistrus, Hemiancistrus, Hopliancistrus,
Lithoxus, Oligancistrus, and Pseudancistrus.
209
Figure 5. Hypostominae, Ancistrini: Peckoltia, Pseudacanthicus, Pseudolithoxus, and
New Genus 3.
210
Figure 6. Hypostominae, Ancistrini: Hypancistrus, Leporacanthicus, and
Scobinancistrus.
211
Figure 7. Loricariinae, Hartiini: Harttia, Lamontichthys, and Sturisoma; Farlowellini,
Farlowella; and Hypoptopomatinae, Hypoptopomatini, Hypoptopoma.
212
Figure 8. Loricariinae, Loricariini: Loricaria, Pseudoloricaria, Limatulichthys,
Rineloricaria, Loricariichthys, and Spatuloricaria.
213
Figure 9. Hypostominae, Ancistrini: Ancistrus macrophthalmus, A. temminckii, Ancistrus
sp. ?longjaw?, Ancistrus sp. ?shortjaw?, Ancistrus sp. ?wormline?, and Ancistrus sp.
214
Figure 10. Hypostominae, Ancistrini: Baryancistrus beggini, B. demantoides,
Baryancistrus sp. ?B&W?, Baryancisrus sp. ?gold nugget?, and Baryancistrus sp. ?green
nugget?.
215
Figure 11. Hypostominae, Ancistrini: Chaetostoma microps, C. cf. milesi, and
Chaetostoma sp.
216
Figure 12. Hypostominae, Ancistrini: Hemiancistrus guahiborum, H. sabaji, H.
snethlageae, H. subviridis, and Hemiancistrus sp. ?gold spot?.
217
Figure 13. Hypostominae, Ancistrini: Hypancistrus contradens, H. furunculus, H.
inspector, and H. lunaorum.
218
Figure 14. Hypostominae, Ancistrini: Hopliancistrus tricornis, New Genus 3,
Lasiancistrus schomburgkii, L. tentaculatus, and Dekeyseria scaphirhyncha.
219
Figure 15. Hypostominae, Ancistrini: Panaque albomaculatus, P. gnomus, P. nocturnus,
P. cf. maccus, Panaque n. sp. ?Mara?on?, and P. cf. nigrolineatus.
220
Figure 16. Hypostominae, Ancistrini: Peckoltia braueri, P. cavatica, P. vermiculata, and
Peckoltia sp. ?big spot?.
221
Figure 17. Hypostominae, Ancistrini: Pseudancistrus nigrescens, Pseudolithoxus
anthrax, Pseudancistrus pectegenitor, Pseudolithoxus dumus, Pseudancistrus sidereus,
and Pseudolithoxus tigris.
222
Figure 18. Hypostominae, Ancistrini: Leporacanthicus cf. galaxias, L. triactis.
Oligancistrus puctatissimus, Lithoxus lithoides, Pseudacanthicus leopardus, and
Scobinancistrus sp.
223
Figure 19. Hypostominae, Hypostomini (Hypostomus cochliodon-group): Hypostomus
macushi, H. pyrineusi, H. taphorni, Hypostomus sp. ?dirty?, and Hypostomus sp. ?spots?.
224
Figure 20. Hypostominae, Hypostomini: Hypostomus hemiurus, H. niceforoi, H. rhantos,
and Hypostomus sp.
225
Figure 21. Hypostominae, Hypostomini (Hypostomus emarginatus-group): Hypostomus
emarginatus, H. cf. emarginatus, H. squalinus, and H. unicolor.
226
Figure 22. Loricariinae, Harttiini: Harttia platystoma, Harttia sp., Sturisoma monopelte,
Sturisoma nigrirostrum, and Lamontichthys filamentosus.
227
Figure 23. Loricariinae, Loricariini: Limatulichthys griseus, Pseudoloricaria sp.,
Rineloricaria fallax, R. lanceolata, R. stewarti, and Rineloricaria sp.
228
Figure 24. Loricariinae, Loricariini: Loricaria clavipinna, Loricaria sp. 1, Loricaria sp.,
and Loricariichthys brunneus.
229
Figure 25. Loricariinae, Loricariini: Spatuloricara puganensis, Spatuloricaria sp. 1, and
Spatuloricaria sp. 2.
230
Figure 26. Loricariinae, Farlowellini: Farlowella acus, F. amazona; Hypoptopomatinae,
Hypoptopomatini, Hypoptopoma guianense; and Hypostominae, Pterygoplichthyini,
Pterygoplichthys gibbiceps.