Herpetologists' League

Guilds of Anuran Larvae: Relationships among Developmental Modes, Morphologies, and

Habitats
Author(s): Ronald Altig and Gail F. Johnston
Reviewed work(s):
Source: Herpetological Monographs, Vol. 3 (1989), pp. 81-109
Published by: Herpetologists' League
Stable URL: http://www.jstor.org/stable/1466987 .
Accessed: 18/06/2012 07:33

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Herpetologists' League is collaborating with JSTOR to digitize, preserve and extend access to Herpetological
Monographs.

http://www.jstor.org
HerpetologicalMonographs,3, 1989, 81-109
? 1989 by The Herpetologists'League, Inc.

GUILDS OF ANURAN LARVAE: RELATIONSHIPS

AMONG DEVELOPMENTAL MODES,
MORPHOLOGIES, AND HABITATS
RONALD ALTIG AND GAIL F. JOHNSTON1
Department of Biological Sciences, Mississippi State University,
Drawer GY, Mississippi State, MS 39762, USA

ABSTRACT: Anurans show a distinct dichotomy among the sources of energy used during larval
growth and development, endotrophy and exotrophy. This distinction allows more exact definitions
of direct development, ovoviviparity, viviparity, and related terms. Endotrophic larvae, whether
as a non-hatched embryo or a free-swimming, non-feeding tadpole, gain immediate developmental
nutrition solely from parental sources, most frequently from the yolk allotted to each egg during
vitellogenesis. Exotrophic larvae are always free-swimming, feeding tadpoles and orally consume
numerous sources of energy not derived from a parent.
The morphology of exotrophic tadpoles is surveyed with an emphasis on oral structures. Spec-
ulative scenarios and hypotheses concerning the functions of the various morphologies and the
relationship of trophic structures to the ecology of the tadpoles are offered.
Larval morphology and behavior permit the recognition of six developmental guilds of endo-
trophic larvae and 18 ecomorphological guilds of exotrophic larvae; these guilds are fully charac-
terized.

Key words: Salientia; Tadpoles; Energetics; Development; Morphology; Ecology; Guilds

"Larval adaptive radiation and convergent ad- vergent evolution of large-mouthedstream

aptation have produced an abundance of mod- tadpoles and tadpoles with upturned
ifications for pond life, stream life, arboreal life,
terrestrial life (direct development), for surface mouths each from several families were
feeding, and for predatory habits, and other illustrated. Informal concepts of ecomor-
specializations of less obvious ecological signif- phological kinds of tadpole morphologies
icance" [G. L. Orton, 1957]. have arisen from these categorizations.
ORTON(1953) recognized four morpho- Heyer et al. (1990), Liu (1940), and van
logical types of tadpoles based largely on Dijk (1972) applied these ideas to the tad-
spiracularand external oral morphologies. poles of specific geographic regions, and
Starrett(1973) provided names for Orton's Inger (1986) recognized five feeding types
types and added a diversity of morpho- among Bornean tadpoles that correspond
logical information to the framework of reasonably well with ecomorphological
the four basic types. Orton's types, Star- types. A scheme that includes all known
rett's names and the included families are tadpole diversity and provides more exact
as follows: Type I (xenoanura),pipids and boundaries and definitions of the various
rhinophrynids;Type II (scoptanura), mi- groups has not been attempted. With de-
crohylids; Type III (lemnanura), asca- scriptive data now available on the tad-
phids and discoglossids; and Type IV poles of almost 1000 species (update of
(acosmanura), all other families. In ad- Altig and Johnston, 1986), the diversity of
dition, Orton (1953) illustrated major tadpoles seems reasonablywell known, al-
adaptive groups of tadpoles; six types though new morphologies continue to ap-
(i.e., arboreal, surface-feeding, direct pear (e.g., Dubois, 1986). A classification
development, mountain stream, nektonic, of all known tadpole types would 1) pro-
and carnivorous) were "derived" from a vide a better understanding of the taxo-
generalized pond form. Examples of con- nomic and ecological diversity of tadpoles
with more precise conceptual images of
types or adaptive peaks, i.e., guilds and 2)
PRESENT ADDRESS: Department of Biology, Den- identify pertinent areas for further study.
ison University, Granville, OH 43023, USA. As recognized by Orton (1957), any clas-
81
82
82 HERPETOLOGICAL
HERPETOLOGICAL MONOGRAPHS
MONOGRAPHS [No.
[No. 3

sification of ecomorphological kinds of in the available developmental energy

tadpoles must reflect the fact that the supplied by the mother must exist (Elin-
scheme is actually a continuum. son, 1987; Raff, 1987), because oral con-
The paper is divided into three sections. sumption of nutrients is usually omitted.
The first part addresses the interactions By recognizing a dichotomy rooted at the
among developmental energy and devel- point of divergence of the two types of
opmental pattern. Secondly, in support of embryonic energy sources, 1) develop-
the second comment above, we survey mental patterns are partitioned into two
morphological diversity of oral and body groups and 2) developmental modes not
configurationsand propose many hypoth- involving a feeding tadpole become more
eses and speculative scenariosto stimulate precisely defined. The following terms ap-
investigations of the interactions between ply equally well to all orders of amphib-
functional morphology and ecology. A dis- ians.
cussion of the interaction of oral structures Endotrophic larvae, regardless of any
and tadpole ecology follows. Lastly, six de- other factor(s)of the reproductive process,
velopmental guilds of endotrophic larvae obtain immediate developmental nu-
and 18 ecomorphological guilds of exo- trients only from parentally derived
trophic larvae are described. Tadpole ter- sources. The first grade includes the vast
minology of Altig (1970) and staging by majority of species in which only yolk al-
Gosner (1960) are followed throughout. lotted to a given egg during vitellogenesis
is available. The second grade includes a
DEVELOPMENTAL ENERGY, few cases in which vitellogenic yolk from
PATTERNS, AND SITES the mother is eventually supplemented by
Classically stated, anuran breeding bi- non-yolk, oviducal (e.g., some Nectophry-
ology includes an aquatic, non-reproduc- noides) or pseudo-oviducal (e.g., paternal
tive, energy-gathering larval stage that vocal sac secretions from male Rhinoder-
eventually metamorphosesinto a frog. En- ma) materialsfor oral consumptionwithin
ergy for early development is derived from the parent's body. Embryological devel-
enzymatic degradationof yolk supplied by opment ranging from entirely typical to
the mother. Developmental regimens highly modified results in a free-swim-
through hatching are surprisingly conser- ming, non-feeding tadpole of various sorts
vative among species. After yolk exhaus- or an embryonic froglet. Regardless, all
tion, the free-swimming, feeding tadpole endotrophic juveniles first enter the tro-
orally consumes numerous other kinds of phic environment outside the parent as
nutrientsthat usually are not derived from miniature replicas of their parents. Con-
a parent. Conversely, developmental pat- cepts of oviparity, ovoviviparity, and vi-
terns and breeding biologies of many taxa viparity do not change, and this develop-
diverge from this standard in numerous mental-energetic categorization does not
ways. For example, feeding may not oc- imply that all developmental patterns di-
cur, even when a tadpole morphotype ex- vergent from normal development are
ists. monophyletic.
We suggest that the source of develop- Alternately, for some variable period
mental energy is the ultimate discrimi- during development, exotrophic larvae
natorbetween the typical patternand those orally consume a number of kinds of nu-
that diverge from it, regardlessof all other trients that are rarely derived from a par-
developmental or biological (e.g., female ent. A feeding-tadpole morphotype always
size, breeding mode, clutch size, etc.) fac- occurs. A few cases that appear as excep-
tors of anuran breeding. Developmental tions are classified as exotrophic, because
patternsthat diverge from the norm imply even if parental yolk is involved, it is ex-
some degree of escape from 1) the reliance traneous to the vitellogenic yolk allotted
on free water and 2) a feeding tadpole as for the initial development of a given em-
dictated by the standard breeding mode. bryo and never while the embryo or tad-
To accomplish such a shift, an alteration pole is intimately associated with a par-
1989]
1989] HERPETOLOGICAL MONOGRAPHS 83

ent's body. For example, some dendrobatid

females purposelysupply unfertilized eggs
to a specific tadpole (Weygoldt, 1980).
Cannibalism of siblings or eating of non-
sibling but homospecific or heterospecific FIG. 1.-Proposed method to quantify the angular
eggs or embryos is common in arboreal orientation of the oral disc of anuran tadpoles relative
tadpoles (Lannoo et al., 1987). Tadpoles of to a defined longitudinal body axis. Two extremes,
Bufo periglenes (Crump, 1989) and prob- ventral (acute angle A) and upturned (obtuse angle
ably Bufo haematiticus (R. W. Mc- B) mouths, are shown; angles for other mouths would
fall between these two values. See text for further
Diarmid, personal communication) devel-
explanation.
op in a typical manner from large eggs for
Bufo and facultatively alternate between
endotrophy and exotrophy depending on scribe consistently the body axis, and this
the availability of food. method demands that the entire specimen
The physical environments of endo- be straight. Secondly, the plane of the oral
trophic larvae are varied, although consid- disc is described by a line connecting the
erably less so than those of exotrophs.Mor- junction of the base of the lower labium
phologicalvariationof the few endotrophic with the body (or similar point, as in mi-
tadpoles is low relative to exotrophic tad- crohylids)and a similarpoint dorsally.This
poles. Details of embryology are poorly line is extended until it contacts the lon-
known for most endotrophs(see Townsend gitudinal axis, and the angle between de-
and Stewart, 1985, for an exception), and scribes mouth position. Species with ex-
the development of hemiphractine hylids tremely ventral mouths may produce an
is unique. Developmental patterns may be angle of 0? (lines parallel), those with ter-
similar among certain of the highly mod- minal mouths would be at 90?, and those
ified endotrophic groups. Six develop- with upturned mouths would be greater
mental guilds of endotrophs (Appendix) than 90?. Pertinent points for this mea-
are based on the site of development, sex surement can be quickly located with a
of the associated parent when pertinent, camera lucida without drawing the tad-
type of association with a parent, degree pole. When the plane of the mouth is not
of lossof the tadpole morphotype,and post- easily determined, as with some micro-
hatching activities of early larvae. Eigh- hylids and pipids, the orientation of the
teen ecomorphological guilds of exotrophs oral structuresmay be difficult to evaluate.
(Appendix) are based on morphological, Oral disc papillae.-Configurations of
anatomical, and behavioral features that marginal papillae are commonly used in
reflect adaptations to specific but varied the identification of tadpoles and strongly
environments. influenced our designation of guilds, par-
ticularly of stream tadpoles. Distributions
SURVEY OF TADPOLE MORPHOLOGY
of marginal oral disc papillae fall into five
groups:absent, dorsal gap only (taxonom-
Oral and Buccal Structures ically and ecologically common), dorsal
Oral disc orientation.-The orientation and ventral gaps (predominate in bufo-
of the oral disc reflects ecological differ- nids, few ranids, rhacophorids, and hy-
ences. Qualitative terms (e.g., ventral, an- lids), ventral gap only (uncommon), and
teroventral,subterminal,and terminal) are no gaps (= complete). At least some species
used commonly to describe the variations. of Cardioglossa have papillae around the
Figure 1 presents a means of measuring entire "lower labium" so that some pa-
oral disc orientation. The longitudinal axis pillae stand in front of the recessed jaw
of the tadpole is described as the extended sheaths. Within the above groups, there
line connecting the tip of the tail and the are large variationsin the number, shape,
point where the axis of the tail myotomes and size of papillae. Most categories are
contacts the body. Because of morpholog- scattered throughout a diverse taxonomic
ical variations, no other two points de- and ecological range of tadpoles, whereas
84
84 HERPETOLOGICALMONOGRAPHS [No.
[No. 3

E an object in flowing water. We suggest that

the following series of morphological
D ,,
changes allows a progressive occupation of
faster water: relatively few marginal pa-
C
B 0 I
pillae with an anterior gap and a labial
x
tooth row formula (LTRF) of 2/3 - pa-
a.
B
pillae similar with increased number of
tooth rows in various formulae -, many
I A
small marginal papillae in a complete se-
o
ries and often in multiple rows with a
LTRF of about 2/3 - papillae similar and
FASTER CURRENT ? with numerous formulae larger than 2/3
FIG. 2.-Diagrammatic relationship between tad-
up to at least 17/21 -- papillae as before,
LTRF 2/3 or greater, and belly modified
pole oral complexity and occupation of water of in-
creasing velocity: (A) lotic-benthic among others, la- into a sucker (= gastromyzophorous). This
bial tooth row formula (LTRF) 2/3, few, large scheme does not imply that the changes
marginal papillae with anterior gap; (B) clasping, occurred in a linear fashion as shown in
marginal papillae largely unchanged, LTRF > 2/3;
(C) adherent, many small, closely spaced marginal Fig. 2.
papillae complete around oral disc, LTRF usually
Without knowing more about the func-
2/3; (D) suctorial, marginal papillae as in (C), LTRF tions of the papillae (e.g., chemosensory,
> 2/3-17/21; (E) gastromyzophorous, mouth parts tactile, flow and water content monitors,
as in (C) or (D) and belly modified as a sucker, sucker modification of mouth shape during feed-
surface papillate and some with rugose keratinized
areas. In (B), (C), and (D), the shown row numbers ing, substrate manipulation), comments
are not indicative of specific formulae; see text for concerning the functional differences of
additional discussion. lesser patterns of variation [number/dis-
tance, size (extremely long posterior pa-
pillae in some Hemisus and Phrynobatra-
complete marginal papillae are more lim- chus), and shape] cannot be made. Van
ited: stream inhabitants of several families, Dijk (1981) believed that the ventral gap
some species with inverted funnel-mouths common in bufonid tadpoles acted as a
(= umbelliform; Fig. 8B), and some species "wier-like flow controlling structure," and
that are totally or partially carnivorous Gradwell (1975a) believed that these gaps
(Fig. 8F) that may inhabit tree holes and gave needed flexibility to the disc for feed-
other small bodies of water. Those with ing and attaching to irregular objects.
umbelliform mouths that feed at the sur- Configurations of submarginal papillae
face of the water may use the complete, seem less correlated with habitats. Large
but sparse, papillary row as a crude filter numbers of small, closely spaced papillae
for water entering the small pool within occur in lentic-nektonic forms (some North
the oral disc. American Hyla) and many rheophilous
In those tadpoles that adhere to sub- forms. Lotic forms other than clasping
strates via their mouths, the particular mi- types often have submarginal papillae in
crohabitat that the tadpoles occupy ob- a row(s) distal to A-1 (cannot occur in
viously reflects the degree to which a species without complete marginal pa-
particular disc configuration affords adhe- pillae) and below the last posterior row
sion. In species with incomplete marginal (particularly hylids). Bufonid and ranid
papillae (Fig. 8A), the upper labium is ab- tadpoles have few submarginal papillae re-
sent for the width of the papillary gap, so gardless of habitat.
that complete marginal papillae also de- Labial teeth.-Other than characteriza-
mands a complete upper labium. Small, tion (e.g., Altig, 1973; Altig and Pace, 1974;
closely packed papillae may allow better Faier, 1972; Gosner, 1959; Heron-Royer
adhesion to the irregularities of substrates and van Bambeke, 1889; Nicholls, 1937)
and perhaps lie within the thin layer of and explanations of formation (Deunff and
still water (boundary layer) surrounding Beaumont, 1959; Luckenbill, 1964; Tachi-
1989] HERPETOLOGICALMONOGRAPHS 85

bana, 1978), little can be said concerning other rows and the teeth are much smaller
the diversityof tooth forms. Mertens(1960) and probably a different form. Different
and Nachtigall (1974) discuss morpholo- lengths of adjacent teeth (e.g., Bragg and
gies of tadpole mouths and other structures Hayes, 1963) are probable misinterpreta-
that serve analogous functions. Our un- tions of where transparentsoft tissue of the
published data and those of Faier (1972), tooth ridge stops. Retention of a tooth seat-
Gosner(1959), and othersreveal only weak ed upon the presently erupting tooth caus-
ecological or taxonomic correlations. es some teeth to appear longer than others.
Functional differences are largely un- Retention of just the basal sheath of an old
known. Suggested functions of labial teeth tooth on a newly erupted tooth sometimes
include rasping material from a substrate presents an unexpected shape. We know
(Savage, 1952), cellular piercing (Savage, of no case where adjacent teeth are grossly
1952), holding food (Luckenbill, 1965), different in size, although size and cusp
trapping food (Tyler, 1963), and substrate complexity generally reduce gradually as
attachment (Altig and Brodie, 1972; Grad- one moves laterally within a row.
well, 1975a). Acting as a broom or current Two cases will illustratemany pertinent
generator to sweep material toward the traits. Amolops larutensis (ranid) has a
mouth, breaking up mucilagenous layers, suctorial,gastromyzophoroustadpole with
sieving particles, and combing strandsinto complete marginal papillae. The tadpoles
alignment for easier cutting could be attach to rocks in fast flowing, often tur-
added. Even so, these functions would not bulent, water. The head of a tooth from
appear to require the large diversity of the medial part of A-1 (stage 25; 10 mm
shapes, sizes, and cusping patternsthat ex- TL) is ca. 8.4 Am wide with ca. 1.1 ,/m
ist. between teeth; about 96% of the row is
Cusps commonly occur on the terminal teeth and 4% is space. With 16 cusps per
and lateral surfaces of tooth heads where tooth and 115 teeth/mm, over 1800 work-
they act only when a tooth is inserted into ing points/mm occur. If all cusps of all
a substrate or the row folds backwards. rows are implanted, the large surface sure-
Assuming the points of the cusps are the ly provides an efficient tool for food re-
primary working surfaces, the points of moval and a strong anchor for position
some medially flexed cusps would seem maintenance. In suctorial species, the oral
never to contact the surface (e.g., Heleo- disc obviously cannot flex during jaw op-
phryne). There are interspecific and row erations, so the teeth cannot rasp in the
differences in the degree of interdigitation same manner as non-suctorial tadpoles;
of replacement teeth (Fig. 3C), although much of the substrateremoval may be via
the degree of interdigitation and the num- the serrationson the jaw sheaths.
ber of replacement teeth (usually 2-3) The lentic-nektonic, "rasping"tadpole
seemingly show no ecological or taxonom- of Rana heckscheri illustratesthe opposite
ic correlations. Lotic tadpoles commonly extreme. Teeth are spikes so that width
have smaller teeth arranged more densely depends on where the tooth is measured,
than lentic forms, although it is not clear intertoothdistance is 1.5,m, tooth density
which part(s) of the tooth are reduced 39/mm, number of working surfaces/mm
(Gosner,1959). These differences likely re- is 39 (no cusps), 29% of the row is teeth
flect physical characteristics of the foods and 71%is space. Drawing an analogyfrom
eaten and the substratesto which the tad- mechanical abrasives, tools for working
poles attach. Intraspecific row differences hard materials have many small, closely
in tooth density are most commonly less spaced surfaces, pressure applied to the
than 10 teeth/mm. A Litoria or Nycti- tool is high, and the removal per stroke is
mystes from New Guinea with tooth den- small (metal file analog). Tips of non-
sities of 88, 33, 56, 85, and 140 teeth/mm cusped teeth (few points/distance) of sim-
in rows A-1 through P-3 is exceptional. ilar dimensions to cusps of cusped teeth
Tooth density in the distal posterior rows (more points/distance) may perform quite
of Ascaphus tadpoles is much higher than different functions. Conversely, tools for
86 HERPETOLOGICALMONOGRAPHS [No. 3
[No.

workingsoft materialshave relatively few, er such adhesive abilities pose a removal

larger, widely spaced surfaces, work with problem for tadpoles, but the thinness of
less pressure, and remove more material the cells or cell layers provides some pro-
per stroke (wood rasp analog). Tool sta- tection from ranid tadpoles with relatively
bility, angle of attack, stroke rate, and the straight, non-cusped teeth (Fig. 3A; Dick-
ability of the abrasive to remain clean re- man, 1968).
late to shape, size, and longevity of the Labial teeth may break up the relatively
working surface. Wear and turnover rates strong sheets of diatom mucilage at the
of labial teeth are not known (see Kuang soil-water interface (Holland et al., 1974).
and Kollros,1976 for data on jaw sheaths). The flocculent nature of bottom debris in
We have observed wear only in darkly temporaryponds is partiallycaused by this
pigmented teeth. Because darkly pig- mucilage layer being disrupted by such
mented structures are typically stronger organisms as tadpoles. Tadpoles of Bufo
than lightly pigmented ones, we suggest and Scaphiopus that appear to be filled
that darkly pigmented teeth are retained only with soil are likely feeding on this
longer and thus exhibit wear, whereas soil-mucilage-diatom mixture plus parti-
lightly pigmented teeth are simply re- cle attached bacteria (Bowen, 1980; Test
placed. and McCann, 1976).
In summary,one could predict that tad- Labial tooth rows and tooth ridges.-
poles of R. heckscheri forage on thicker, The number and arrangement of tooth
less dense substratesand remove a larger rows are common traits used in identifi-
amount per feeding stroke than tadpoles cation; intraspecificvariation in row num-
of Amolops. Large numbers of small, ber is known (e.g., Bragg and Bragg, 1959;
closely spaced, cusped teeth apparently Hampton and Volpe, 1963; Potthoff and
supply a useful tool for harvesting thin, Lynch, 1986) but is relatively uncommon
perhaps tough layers of organics attached in most taxa. Inter-row spacing and tooth
to rocks in fast flowing water. Thicker ridge topography seemingly reflect eco-
growths of less tensile strength common to logical specializations. Inter-row spacing
lentic periphyton suggest different har- and tooth size are reduced distally on at
vesting mechanics. A Rana-type of tooth least the lower labium in species with large
arrangement may either "pick" diatoms numbers of tooth rows. This reduces the
or "comb" fibrillar periphyton into par- potential of the oral disc becoming im-
allel arrays for easier cutting by the jaw mense and provides an increase in contacts
sheaths. per area as distance from the mouth in-
Labialtooth shape surelyaffectsthe style creases. The musculature of oral discs is
and length of substrate contact and the probablynot well developed distally to the
pressureneeded to keep the tooth implant- mouth in most species. Tooth ridges with
ed or in properalignment. It would appear biserial tooth rows (e.g., ascaphids, disco-
that strongly curved teeth (e.g., Fig. 3B; glossids) may represent an alternate solu-
Bufo in Altig, 1973) contact the substrate tion to increasingthe total number of tooth
for a short distance before labial disc flex- rows.Tadpolesof Bombina orientalis have
ure rotatesthe cusps off the substrate.Such all rows presentand uniserialat about stage
teeth may be used primarilyto pry objects, 26, so the anlagenof the second serieseither
such as diatom frustules, off a substrate. are not presently differentiated or tooth
Adhesion of diatoms may equal 2.9-4800 production in the second series is delayed.
(x = 634; n = 8 species) nanoNewtons It is not known if the production of two
(Harper, 1977);Drum and Hopkins (1966) tooth rows/tooth ridge represents fusion
found it difficult to pipette diatoms be- of tooth ridges, fission of developmental
cause they adhered firmly to glass. With- anlage, or the production of two cell lines
out knowing the mechanics of the entire from one mitotic zone. Transitory ap-
feeding mechanism (Gradwell,1968,1970, pearance of tooth ridges indicative of the
1971, 1973, 1975b; Gradwell and Pasztor, second series in each row in Bombina does
1968) of a tadpole, we cannot judge wheth- not occur. We have seen a few specimens
 1989] HERPETOLOGICAL MONOGRAPHS 87

I
V,
A

406
. A- .
.

FIG. 3.-Scanning electron micrographs of labial teeth and an example of a feeding mark: (A) short-cusped,
relatively straight tooth of Megaelosia goeldi, x440; (B) strongly curved, long-cusped teeth in single row/
tooth ridge of Hyla biobeba, x710; (C) emergent tooth (upper left) and replacement teeth in Leptodactylus
chaguensis, x550; (D) lateral view of non-cusped, short-sheathed tooth of Scaphiopus bombifrons, tooth
head broadly spatulate in face view, x1720; (E) negative cast of two feeding marks made by a tadpole of
Rana utricularia in an agar-based food penetrated by the jaw sheaths (upper jaw sheath entered rather
shallowly at left, lower jaw sheath entered steeply at right; lower sheath skipped once before actual entry
on left mark); (F) cusped teeth in two rows/tooth ridge in Bombina orientalis, x510.
88 HERPETOLOGICAL MONOGRAPHS [No. 33
[No.

gressive changes with occupation of faster

water. Tooth ridges become broad-based
pyramids with flattened apices as viewed
in cross-section.Inter-row valleys are shal-
low, slit-like, and immediately posteriorto
the preceding row so that the plane of the
oral disc appearsalmost uniform (Fig. 4B).
Extended front bases of the teeth (e.g.,
Ascaphus, Altig and Pace, 1974) extend
into the inter-row tissue to form strong
attachmentsfor both teeth and rows. Rows
have little flexibility and cannot fall back-
wards. These arrangements aid active
gripping by the upper and lower rows that
face and act in opposite directions.
Shapes of tooth rows suggest potential
actions and functions. Rows are typically
arched away from the mouth, with lower
rowscommonly straighterthan upper rows.
This shape remains during use (Johnston,
1982). Regardlessof the arch of a row, each
FIG. 4.-Schematic longitudinal section of the low- tooth is set parallel with the longitudinal
er labia of (A) lotic-suctorial and (B) lentic-nektonic
axis of the tadpole. Without this configu-
tadpoles showing the topography and shapes of tooth
ridges and associated structures. ration,either the row would have to change
orientation during a strike or only medial
teeth would face the substrate squarely.
of ranid and rhacophorid tadpoles nor- Because the row is arched, at least the lat-
mally with uniserial tooth rows that have eral sections skid even though each tooth
biserial tooth rows. Thibaudeau and Altig does not. This action should provide ad-
(1988) showed a close spatial-temporal de- ditional abrasion and plow material to-
velopmental relationshipbetween rows A- ward the mouth.
1 and P-2, and tooth row use indicates Lastly, tooth row lengths relative to oral
these two rows are likely of prime impor- disc dimensions and the positions of rows
tance in some lentic species (G. F. John- on the disc are considered. Suctorial tad-
ston, unpublished data). Accessory tooth poles use their mouths for position main-
rowsof some ranids(e.g., Webb and Korky, tenance, and long rows, particularly ad-
1977) and pelobatids are not parallel with jacent to the edges of the oral disc, would
the main tooth rows and are situated lat- seem important. This is typical, however
eral to the jaw sheaths. Their functions Hyla uranochroa (Fig. 8H), H. rufioculis,
appear unassociatedwith the suspected di- and part of the genus Ptychohyla (Fig. 261
rection of action of either labium. in Duellman, 1970) have all short rows
Lentic tadpoles have high, distally situated adjacent to the jaw sheaths. Large
rounded tooth ridges that often are widest lateral and posteriorzones lack teeth. The
near the apex and separatedby wide, deep large oraldisc with weakly developed tooth
inter-row valleys situated about equidis- rows and jaw sheaths of at least some tad-
tant between rows (Fig. 4B; Noble, 1931). poles of Taudactylus (Watsonand Martin,
Such arrangementsafford high flexibility, 1973) is similar. Oral discs of lotic species
small forces against a substrate, and little are less resilient than those of lentic forms,
retardationof the entire row folding back- but how are these toothless areas held
wards during use. If rasping is the ex- againstthe substrate?We suspectthat these
pected action, efficiency of removal would forms inhabit a zone different from those
seem to be low. Teeth are inflexible but species that have rows nearly equal to the
rows are flexible. Lotic forms show pro- oral disc width and closer to the edges. As
199
1989] MONOGRAPHS
HERPETOLOGICALMONOGRAPHS
HERPETOLOGICAL 89
89

expected,lowertoothrowsof lotic species Echeverria, 1984), and Hyla chrysoscelis

are typically long, whereas lentic forms and Rana utricularia (Thibaudeauand Al-
may have a distalmostrow as shortas 5% tig, 1988) [all pond forms with 2/3 rows]
of the lengthof the proximalmost
posterior appears to be A-i, P-2, P-l, A-2, P-3
row. (sequentialformulae = 1/0, 1/1, 1/2, 2/2,
Tooth row formula diversity and row 2/3). The timing and developmental to-
homology.-We categorizetoothrowfor- pography of A-1 and P-2 are closely
mulaeas balanced(equalnumberof rows linked, and these rows may appear simul-
on upperand lowerlabia),negativelyim- taneously or slightly reversed in time. Lit-
balanced(morerowson lowerlabium),or erature data on partial developmental se-
positivelyimbalanced(morerows on up- ries suggest that this pattern is common in
per labium).A tabulationis alwayspoten- species with 2/3 rows and for the appear-
tially biasedby the speciesincluded,but ance of the first five rows in species with
of 627 speciesexcludingspecieswith 0/0 more than five rows. These five rows in a
(Table 1), imbalancedmaximawere +7 2/3 configuration appearing in the above
and -11, and 88% of the species were order are designated the prime rows or
between+2 and -2. Untilmoreis known formula. Adding subsequent rows alter-
aboutthefunctionalaspectsof varioustooth nately on each labium seems to be a com-
row formulae,the significanceof this ba- mon occurrence, but discordance of stage
sicallyunimodaldistributioncannotbe ad- and tooth row appearance, especially in
dressed, but interestingquestions arise. species with formulae much larger than
What are the functionaldifferencesbe- 2/3, is common. Species with many tooth
tween negatively (4/15) and positively rows do not attain the eventual comple-
(10/3) imbalancedformulae?Whatis the ment of rows until later in ontogeny and
significanceof the sequencesbetween -2 retain teeth into later stages at metamor-
and +4 wherethe mostcommonformulae phosis compared to species with 2/3 rows,
areusuallyformedby increasingthe upper and this phenomenon is accentuated in
labialrowsby 1 while keepingthe lower stream forms that use their mouths for po-
labiumat 3 (i.e., 1/3 through7/3). Does sition maintenance.
the relative rarity of highly imbalanced Evaluation of whether ontogenetic gain
formulaeimply specificfunctionsor mi- mimics phylogenetic additions of rows is
crohabitats?What is the relationshipbe- difficult, because 1) the ontogeny and de-
tweenthe sametotalrownumberandbal- velopmental controls of median row gaps
ancevalue(e.g.,9/9 vs.4/14)?Theformula are not understood (e.g., Dutta and Mo-
2/3 at BV -1 occursin taxa from many hanty-Hejmadi, 1984; Michalowski, 1966)
habitatsand in 51%of the taxa sampled. and 2) data are available for so few species
Total rows ranged from 1-38 in 49 row with more than 2/3 rows. We hypothesize
configurationswithout considering row that there are four basic patterns (Fig. 5)
gaps, and at least 124 formulaeoccur if of tooth row additions to the prime for-
rowgapsareincluded(AltigandJohnston, mula. Less common patterns of row ad-
1986). ditions (or losses) involve combinations or
The sequenceof tooth row appearance parts (one labium) of these patterns. Two
is known within a narrowrange of taxa, contrasting examples will illustrate.
guilds,and tooth row formulae,and pat- Ranids and pelobatids include many
ternsof metamorphicatrophyarelesswell stream forms with many tooth rows and
understood.Evenso, thesedatalikelypro- incomplete marginal papillae. We suggest
vide the best evaluationsof toothrow ho- that these tadpoles do not use the mouth
mology,i.e., which rows in a formulaof extensivelyfor attachmentto substratesand
17/21 equalthe rowsin a formulaof 2/3. have more jaw mobility, larger gapes, and
The ontogeneticsequenceof row appear- different actions for substrate removal
ance for Bufo valliceps (Limbaughand compared with suctorial forms. Gaps in
Volpe, 1957), Rana pipiens (Nicholls, rows are progressively wider towards the
1937), Bufo arenarum (de Lopez and large,stronglyarchedjaw sheaths,and rows
90 HERPETOLOGICAL
HERPETOLOGICALMONOGRAPHS
MONOGRAPHS [No.
[No. 33

TABLE 1.-Number of species (n = 627) that occurs at each labial tooth row formula (LTRF) and balance
value (BV); see text for further discussion.

Number of species
BV LTRF 0 10 20 30 320

Negative imbalance
-11 4/15 XX
-10 4/14 X
-6 3/9 X
-5 8/13 X
-4 3/7 XX
-4 7/11 X
-3 1/4 X
-3 2/5 XXXXXXXXX
-3 4/7 XX
-3 6/9 XXX
-3 8/10 X
-3 17/21 X
-2 1/3 XXXXXXXXXXXXXXXXXXXXXXXXXXXXX
-2 2/4 XXXXXXXXXXXXX
-2 3/5 XXXXX
-2 4/6 XXX
-2 7/9 X
-2 8/10 X
-1 0/1 XXXXXXXXX
-1 1/2 XXXXXXXXXXXXXXXXXXXXXXXX
-1 2/3 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXX
-1 3/4 XXXXX
-1 4/5 XXXXXX
-1 5/6 X
-1 6/7 XXX
-1 7/8 XX
Balance
0 1/1 XXXXXX
0 2/2 XXXXXXXXXXXXXXXXXXXXXXXXXX
0 3/3 XXXXXXXXXXXXXXXXXXXXXXXXX
0 4/4 XXXXXXXXXXX
0 5/5 XXXXXX
0 6/6 XXXXXXX
0 7/7 XXX
0 8/8 X
0 9/9 X
Positive imbalance
+1 1/0 XXX
+1 2/1 X
+1 3/2 X
+1 4/3 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
+1 5/4 XXXXXXX
+1 7/6 XXXXX
+2 5/3 XXXXXXXXXXXXXXXXXXXXXX
+2 6/4 XXXX
+3 6/3 XXXXXXXXXXXXX
+3 7/4 X
+3 9/6 X
+4 7/3 XXXXX
+5 8/3 X
+7 10/3 X
1989]
1989] HERPETOLOGICAL MONOGRAPHS 91

appear to be added proximally on both B c

labia after the formula of 2/3 is attained
as above. Therefore, the upper prime rows
retain early ontogenetic designations, 0 % D 0
whereas lower prime rows become at least A
numerically more distal (Type E, Fig. 5).
Conversely, in stream-dwelling sucto-
rial forms (e.g., Ascaphus, many Hyla, He-
D 0
leophryne), only the row adjacent to each
jaw sheath is commonly broken. Rows ap-
pear to be added distally on both labia
(Type C, Fig. 5) so that A-1 (= A-i') of 0 % % 0
a young tadpole becomes A-5 of a species
with six eventual upper rows, whereas P-
1', 2', and 3' retain original designations.
In species that add rows distally on the FIG. 5.-Schematic tadpole tooth row formulas in-
lower labium and proximally on the upper dicating proposed means of phylogenetic increase and
decrease of tooth rows from the suggested prime for-
labium, prime rows would retain original mula of 2/3. Heavy lines = prime rows, light lines =
designationsregardlessof the eventual for- additions, circle = jaw sheaths.
mula (Type B, Fig. 5). Such variations in
the patternsof row additions preclude any
numbering system that will preserve the with fewer than 2/3 rows, 88 (71%)species
same designation for homologous rows of fit the expected sequence of 2/3, 2/2 (loss
all formulae throughout ontogeny. of P-3), 1/2 (loss of A-2), 1/1 (loss of P-
Metamorphicoral atrophy appearshap- 1), and 1/0 (loss of P-2) or, based on the
hazardous,but the general pattern follows similar reversal between A-1 and P-2
the reverse of ontogenetic gain (Takahama above, 0/1 (loss of A-i). Formulae such as
et al., 1988; Thibaudeau and Altig, 1988). 0/2 (n = 1), 2/1 (1), and 1/3 (34; mostly
P-3 appears last and is lost first in meta- ranids and hyperoliids) indicate alternate
morphic B. valliceps (Limbaugh and but generally uncommon patterns.
Volpe, 1957). Hillis and de Sa (1988) sug- Few data are available on the genetics
gested that the presence of three lower of tooth row configurations (Michalowski,
tooth rows in Rana sierramadrensis was a 1966), because 1) few studies involve
reversal from the typical four rows in the species with different formulae or config-
R. palmipes group; since P-1 is broken as urations of the same formula, 2) people
expected, it was presumably P-4 that was have not collected these data from hybrids
lost. Tadpoles of Hyla crucifer have few that are available, and 3) few data are
teeth in P-3, and up to 50%of the tadpoles available beyond the ratheruninformative
may normally have a formula of 2/2 (Gos- F1 generation. Hybrids with tooth for-
ner and Black, 1957); when P-3 is absent, mulae of 2/4 from crosses of Rana cas-
the tooth ridge of P-3 may or may not be cadae (3/4) x R. pretiosa (2/3; Haertel
present, and the ventromedial gap in the and Storm, 1970) imply different genetic
marginal papillae that is normally present or developmental controls for the two la-
may or may not be present in 2/2 indi- bia. The observations of Grillitsch and
viduals (personal observation). P-3 is the Grillitsch (1989) concerning more fre-
first row lost in those that have it. We quent abnormalities in the lower, distal
suggest that phylogenetic loss of tooth rows rows may suggest decreased develop-
may occur in reverse order of the onto- mental control distally.
genetic gain of the prime formula, that is, Systematistshave included tadpole data
in the order of metamorphic atrophy. If in their analysesto varying degrees (Gran-
so, certain formulae should not exist. Of dison, 1981; Griffiths and de Carvalho,
124 species (from Altig and Johnston,1986) 1965; Inger, 1967; Kluge and Farris, 1969;
92 HERPETOLOGICALMONOGRAPHS
HERPETOLOGICAL MONOGRAPHS [No.
[No. 33

sion of cellular arrays (Luckenbill, 1964,

1965) that resemble labial teeth, and large
sections of the sheaths are shed similarly
to individual teeth. At metamorphosis, en-
tire sheaths often disengage from the jaw
cartilages as intact units. Jaw serrations are
not cusped. Young tadpoles of Heleo-
phryne (Visser, 1985) and all sizes of Oto-
phryne (Wassersug and Pyburn, 1987) ap-
FIG. 6.-Schematic drawings of jaw sheaths. Buc- pear to have hypertrophied jaw serrations
cal faces of flattened upper sheaths (stipple = lightly without a common sheath at their bases.
keratinized; cutting surface downward) of (A) Hyla Because jaw serrations usually slant more
femoralis and (B) Rana utricularia showing keratini- posteriorly as one moves laterally along
zation patterns presumed to indicate different inter- either sheath, jaw occlusion may provide
actions between upper and lower jaw sheaths; (C) a different action laterally than medially.
front surfaces of both broken sheaths composed of
large units in Amolops sp.; (D) front surface of the
In most cases, the lower jaw passes inside
upper sheath of Plectrohyla matudai with hypertro- the upper, but the jaws are quite mobile
phied serrations (drawn from Duellman, 1970); (E) so that the sheaths may sometimes meet
front surfaces of both cuspate sheaths of the carniv- in a chiseling action. Keratinization of the
orous tadpole of Rana tigerina (drawn from Kirti-
oral face of the upper sheath (Fig. 6A, B;
singhe, 1957); (F) front surfaces of both sheaths of
Hyla pictipes showing typical cutting surface and Johnston and Altig, 1986) shows various
M-shaped accessory working surface on the outer face patterns that may reflect the kind of con-
of the upper sheath (drawn from Duellman, 1970). tact between the two sheaths. Enlarged
serrations, perhaps analogs of mammalian
canines (Fig. 6D), occur on parts of the
Sokol, 1975). Lynch (1971) regarded a for- upper jaw sheaths of lentic-benthic tad-
mula of 3/3 or greater (associated with poles of Leptopelis and suctorial Plectro-
other traits that form a mosiac of traits hyla. Some carnivores have protrusions on
common to stream versus pond tadpoles) either the upper and/or lower sheaths, and
as primitive within leptodactylids. Larger the upper protrusion is often complemen-
tooth formulae within the conservative ar- tary to a notch in the lower jaw sheath
ray of tooth row formulae in hyperoliids without enlargement of serrations (e.g.,
belonged to the more primitive genera Fig. 6E; Ceratophrys, Scaphiopus).
based primarily on analyses of adult traits Serration density, shape, and length plus
(Drewes, 1984). Non-cusped teeth have a number of sheath dimensions, curva-
been considered primitive (Noble, 1931) tures, and shapes that have not been mea-
or derived (Gosner, 1959). Detailed inter- sured likely reflect substrate characteristics
specific developmental data should allow that can be manipulated most effectively.
judgements of character polarity so that Assuming that jaw mechanics are similar,
data on oral structures could be used more tadpoles of Hyla gratiosa, H. regilla, and
effectively in systematics. Without the on- Rana utricularia with wide, thick, robust
togenetic data, one cannot differentiate sheaths with curved front surfaces must be
between identical formulae derived by dif- able to feed on resources unavailable to
ferent developmental patterns. Accessory species with narrow, thin sheaths with flat-
rows in some ranids (Webb and Korky, ter surfaces (e.g., Bufo spp., R. catesbei-
1977) and pelobatids, keratinized areas ad- ana). Johnston (1982) showed that tad-
jacent to the lower jaw sheaths in some poles of Rana utricularia consistently could
ranids and hyperoliids, and the keratinized penetrate and remove pieces from an
patch in the roof of the buccal cavity in agar-based food that tadpoles of Bufo
some Scaphiopus (Spea) also need atten- woodhousei of the same size could not pen-
tion. etrate. Several physical and biological fac-
Jaw sheaths.-Jaw sheaths form by fu- tors (e.g., substrate thickness, consistency,
1989]
19891 H T C
HERPETOLOGICALMONOGRAPHS 93

conformity, and position; water depth; being implanted too deeply into a sub-
suitabilityof food; hunger) likely influence strate. Tadpoles of Hyla tica and H. pic-
jaw actions, and lotic species seem to ex- tipes (Fig. 6F; Fig. 136 in Duellman, 1970)
hibit the most profound morphological appear to have two working surfaces on
modifications. The analogies between la- the upper sheath. The serrated edge may
bial teeth and abrasives apply here, and be used in slower water when the disc-
lotic tadoles can be placed in a continuum substrate space is not severely restricted.
from those with finely spaced, narrow- A roughened area on the front surface of
based sharp serrationsto those with broad- the sheathis probablyused when the sheath
ly spaced,wide-basedblunt serrations.Fine is nearly parallel with the substrateduring
serrationsseem to occur in species where oralhitching in fasterwater when the disc-
the oral disc-substrate space may not be substratespace is restricted. M-shaped up-
severely restricted. Tadpoles may adjust per sheaths of some lotic forms probably
the orientation and action of the sheaths add strength,retardwarping, and increase
to accommodate one of several actions working surface length without increasing
when in water with different flow rates. jaw length in species having short gapes.
In extreme suctorial species, the sheaths Within a crowded space, the serrationsof
may not occlude, as they appear to be ro- such a sheath may remain free of debris
tated to a more "mouth open" position so more effectively than a straight surface
that the bluntly serrated edges probably and plow material toward the mouth.
act in a scouring or scratching motion Correlationsbetween feeding activities,
nearly perpendicular to the substrate.As- morphology,and substratesformulated for
caphus tadpoles representanother solution larval aquatic insects (e.g., chewer, shred-
to action in a restricted space; the huge, der, scraper; Cummins and Klug, 1979)
flat upper sheath lies almost parallel with and snails (Breureand Gittenberger, 1982;
the substrate and acts at an acute angle. Steneck and Watling, 1982) have never
Some suctorial forms have jaw sheaths di- been recognized for tadpoles. We propose
vided into lateral sections separated by a tentative functional designations for the
medial, non-keratinized area (e.g., some actions of jaw sheaths, while suggestions
Amolops, Ansonia, and pelodryadinine for the actions of labial tooth rows would
hylids) or sheaths are absent (e.g., most be premature. The unique large upper
Heleophryne); the functions of these kinds sheath of Ascaphus tadpoles surely attacks
of sheaths relative to typical forms are not at a small angle and thus shaves thin sub-
known. Broken sheaths are often com- stratesattached to stones in flowing water.
posed of fewer large cellular columns The sheath moves by protraction-retrac-
(Luckenbill, 1965) where each column ap- tion cycles of the anterior end of the body
pears to produce a serration [8/mm in (oral hitching of Altig and Brodie, 1972).
Amolops sp. (Fig. 6C) versus 30-80/mm The small lower sheath probably has no
in more typical sheaths] with square tips. feeding function. Many tadpoles remove
Substrate removal during oral hitching filamentousor projectingmaterial (bite) or
(Altig and Brodie, 1972) or ratcheting extract a portion (gouge, scoop) of periph-
would seem essential in these cases. yton from submerged surfaces. Tadpoles
Many lotic species have a perpendicular with only jaw sheaths (e.g., Fig. 8E, Hyla
shelf of variable extent projectingfrom the microcephala and leucophyllata groups)
medial buccal curvature of the upper are limited to these types of feeding,
sheath slightly orad of the occlusive sur- whereas many other tadpoles with labial
face (Fig. 209 in Duellman, 1970; Hyla teeth may commonly feed in this manner.
sumichrasti, H. smaragdina). The shelf In any case, sheaths contact surfaces at
presumably acts as an anti-warping strut variable but large attack angles and then
to the arch and prevents the lower jaw rotate towards each other (Fig. 3E). The
from passing far beyond the face of the sheathsmay not necessarilyocclude (lower
upper sheath or prevents the sheath from orad of upper); material may be ripped
94 HERPETOLOGICALMONOGRAPHS [No. 33
[No.

from position after only a partial cut by well with macro- and microhabitats, al-
the jaw sheaths. Understanding feeding though hydrodynamic data to verify their
modes between these two extremes must functions are lacking. Although many of
await additional examination of the ac- the generalities formulated for fish loco-
tions of both jaw sheaths and labial teeth. motion apply, body form and lack of lat-
Multiple mastication (chewing) of mo- eral fins on tadpoles must impart specific
lariform surfaces does not apply. Cuspate modifications on their locomotion. Tad-
sheaths of some carnivorous species may poles that live in flowing water may be
allow more efficient cutting of animal tis- strongly depressed (pelodryadininehylids,
sues, and hypertrophied serrations may Heleophryne) or bluntly rounded ante-
pierce or penetrate viscous or dense sub- riorly (Fig. 7M; Ascaphus, many Hyla);
stratesmore effectively than a uniform se- these shapes likely reflect the influences of
ries of serrations. The actions of an in- drag, but analyses are lacking. Other de-
creased number of tooth rows or increased pressed tadpoles include various lentic-
tooth densities in some rows adjacent to benthic (perhaps for shadow reduction)
the jaw sheathsin stream species may sub- forms, burrowers(Fig. 70), lentic-nekton-
stitute for jaw actions (Tyler, 1963). Rheo-ic forms that either remain motionless
philous tadpoles sometimes appear to feed much of the time (microhylids, Xenopus)
repeatedly in the same area rather than or swim most of the time (Rhinophrynus),
moving to a slightly different area at each and variousarborealtypes that live in con-
feeding strokeas is often the case with non-fined spaces. Cylindrical (inactive swim-
suctorialtadpoles.For those with few, blunt mers) and compressedbodies (more active
sheath serrations, such activity would swimmers) are limited to lentic-nektonic
seemingly scour the substrate.Those with forms. Hoff and Wassersug (1985, 1986),
numerous, fine serrations may produce a Nishikawa and Wassersug (1989), Was-
scratching action, especially if they move sersug (in press), and Wassersugand Hoff
between each feeding cycle. (1985) address numerous factors concern-
Buccopharyngeal structures. -The ing tail kinematics and morphology of
feeding mechanism of a tadpole has sev- swimming tadpoles.
eral interacting phases. Materialdislodged Tail fins range from almost absent in
by the mouth parts or suspended in the semiterrestrialforms to immense sizes in
water must be moved into the buccal cav- some lentic-nektonic forms. Stream tad-
ity by the buccal pump. What arrives in poles can be characterized (comparisons
the buccal cavity is selectively captured for lentic tadpoles in parentheses) as fol-
by the gill filters. An array of buccal pa- lows: low, relatively stiff fins roughly par-
pillae (Viertel, 1982; Wassersug, 1980) allel with the tail muscle (higher, more
probably monitors and maneuvers (Inger, flexible fins, often with considerable arch
1986) materialwithin the incoming stream. in silhouette), neither fin extends onto the
Although the filter system is well under- body (dorsal fin commonly does; ventral
stood (Seale, 1982; Seale and Wassersug, fin rarely, e.g., Xenopus, some Rhacopho-
1979; Viertel, 1984a,b, 1985, 1987), the rus, Ololygon, and Phyllomedusa), tail tip
data include too few species to allow rounded (typically pointed, often atten-
uate as a flagellum), and tail muscle mas-
grouping tadpoles by filtering abilities. In-
ger (1986) recognized five feeding types sive in all dimensions (much less massive
among 16 species of Borneantadpoles, and at same or larger total length). Variations
most types could be associated with mor- of fin area/muscle volume, body shape,
phological specializations that were func- and center of gravity probably reflect the
tionally related to the particles ingested. different attitudes in the water (attached
to substratevs. horizontalin water column
Body Forms and Habitats vs. head up or down), swimming styles
Body shapes and eye positions.-Body (positionmaintenance vs. cruise vs. sprint;
and tail shapes and eye positions correlate braking vs. coasting stop; highly maneu-
1989] HERPETOLOGICALMONOGRAPHS 95

verable vs. not so; fast vs. slow), and swim- made only for the extremely long tube in
ming mechanics (strokes/time, stroke am- Otophryne (Wassersugand Pyburn, 1987).
plitude). Johnstonand Altig (1986) recognized a
The tail terminus of taxa within several number of character states for vent con-
lentic guilds that inhabit the water column figurations other than dextral versus me-
is attenuated, often non-pigmented, and dial, but functional, developmental, or
may move independently of the remain- ecological correlationsare not known. Ob-
der of the tail. This attenuate portion is servationof neuromast often requires spe-
called a flagellum because it is a locomotor cific procedures (Lannoo, 1985), and data
device, a function not necessarily implied on these structuresare uncommonly used
in filament, etc. Fins of tadpoles that have (e.g., Boulenger, 1891; Scott and Jennings,
a flagellum vary from very tall (e.g., some 1985; Shelton, 1970). Many lotic forms in
Hyla, Ololygon, and Rhacophorus) to low several families have an enlarged neuro-
and parallelwith the tail muscle (e.g., Hyla mast ventrolaterally on the body just an-
leucophyllata group) to reduced (e.g., terior to the tail. The morphology of the
Xenopus, some phyllomedusine hylids). external nares varies in a number of ways
The high, sail-like anteriorpart of the dor- (aperturerimmed, variously papillate, tu-
sal fin followed by a greatly reduced por- bular; aperture shape; position relative to
tion in the lentic-benthic tadpoles of Oc- the surrounding surface and orientation;
cidozyga is unique. Channing, 1984;Johnstonand Altig, 1986),
Size, position, orientation, lenticular but no ecological correlationsare known.
protrusion,and bulge of the eye above the
surface changes as one moves from lentic- Future Studies
nektonic (large, bulging, lateral, oriented Creative use of myography (e.g., Liem,
perpendicular) through lentic-benthic 1980), high-speed cinematography, high-
(smaller, slightly bulging, dorsolateral to speed x-ray cinematography,electrophysi-
dorsal, oriented mostly horizontally) to ology (e.g., sensory capabilities of oral and
stream inhabitants (smaller, dorsal, non- buccal papillae), and physiographicequip-
bulging eyes) to lotic burrowers(very small, ment (e.g., Gradwell, 1971) on intact and
non-bulging eyes that are distinctly para- manipulated individuals (i.e., selective re-
sagittal). moval of oral papillae, tooth rows, etc.) will
Spiracles, vent tubes, neuromast topog- answer many questions concerning the
raphy, and nares.-Even though varia- functions of various oral structures and
tions of these structures aid identification configurations. Many descriptive projects
(Altig, 1970; Johnston and Altig, 1986; concerning development (e.g., Thibau-
Lannoo, 1985), ecological correlationsare deau and Altig, 1988), comparative anat-
rare. Starrett (1973) correlated spiracle omy and mechanics of the chondrocra-
configurations and modes of feeding and nium (e.g., de Jongh, 1968) and oral
swimming. She pointed out that dual lat- apparatus (Johnston, 1982), and creative
eral, single midventral, or parasagittalspi- analysesof intra- and interspecific oral ap-
raclesoften occur in midwaterforms, many paratusvariation (e.g., labial teeth and jaw
of which are suspension feeders. Flows sheaths) would be helpful. Mendelian and
from these types of spiracles may not dis- developmental/biochemical studies of the
rupt the stability of the tadpole as the flow genetics of various structures would be
from a sinistral spiracle may, or at least valuable, especially in the light that none
not disturbfood items in the water column. of the oral disc structureshas adult deriv-
We have observed tadpoles of Gastro- atives. The temporal-spatial distributions
phryne moving via the midventral spirac- of tadpoles within and among communi-
ular jet as Starrett (1973) speculated. A ties remain poorly known, particularly of
suggestion for the function of long spirac- stream forms, and the enormous realm of
ular tubes that are free from the body (e.g., feeding ecology, including what the tad-
Amolops, Acris, Cardioglossa) has been pole eats and the qualities (e.g., abun-
96 HERPETOLOGICAL
HERPETOLOGICAL MONOGRAPHS
MONOGRAPHS (No.
[No. 33

*t.:':',-...'... -
..'.-;
r "'

B K
i???
;? --
-
?
*, ... ...
J t-i? 1
-e r,*
,.g Te ii?
....
; . ."
I.. .~" . ... -r -e
- :*i?- ?

C L
4,.~~~~~~~~r ~~;~~.a, 4~
t~4 .4.. -

N
.-C:~~~~~~~~t
-?

O
F ?rlcr
r ( k_

Y -; , 1??;.
I ,, alLjj-lu I";*
"i " l?zr.CcT J +

G P

H - A, --~~~~~~~~~~~~~~~~~~~~~
I
f I-.

R
'.9.,-. - . ~ ~ ~~~ ~ ~ ~ ~ abnm?iA.~~.
- .
, - -z~
026 ?- .t-. ? t.
-
-.

FIG. 7.-Representative body forms of tadpoles in various guilds: (A) lentic-benthic to lotic-clasping, Rana
palmipes; (B) lotic-neustonic, Megophrys montana; (C) lentic-nektonic, Type 1, Ololygon nebulosa; (D)
lotic-suctorial, Type 1, Hyla rivularis; (E) lotic-gastromyzophorous, Atelopus ignescens, (F) lotic-neustonic,
19891
1989] HERPETOLOGICAL MONOGRAPHS 97

dance, distribution, nutrient value, phys- ent material that a tadpole swallows may
ical properties) of the prey items, is a not be the primary material that is di-
notable void. gested. The following are available com-
monly as food: free and attached algae plus
RELATIONSHIPS OF LARVAL their extracellular products; higher plant
MORPHOLOGY AND ECOLOGY tissue and plant products alive or degrad-
The naturalfeeding ecology (e.g., when, ed; zooplankton, periphyton, and auf-
where, how, and what, including changes wuchs; particulate organic debris plus at-
induced by environmental factors, ontog- tached bacteria and meiofauna; and
eny, and coinhabitants) of no tadpole is polypeptide chains and amino acids
well known. Studiesof larval insects, snails, throughout the system including the sur-
and microcrustaceans are well ahead of face film (Goldacre, 1949). Live material
studies in the feeding mechanics and ecol- from higher plants and some algae prob-
ogy of tadpoles. Our goal should be to un- ably are ingested to obtain the attached
derstand how the morphology, behavior, periphyton, as much of this material is pro-
and physiology of tadpoles influence their tected by thick cell walls and remains vi-
ecology and community structure. able during gut passage. Other than bro-
Tadpoleseat large quantitiesof food and ken cells, proportionately little of the
have short gut clearance times (Altig and material is digested, although some nu-
McDearman, 1975), non-acidic gut pH trients probably can be removed from in-
(Thrall, 1972), and a long gut with weak tact cells. Organic debris is likely ingested
peristalsisand no featuresother than length to obtain the resident meiofauna, bacteria,
to increase surface area (Savage, 1952, and molecular fragments. We know al-
1955; Wilcznska, 1981). At least on arti- most nothing about energetic, physiolog-
ficial diets, conversion efficiencies are high ical, or behavioral cost-benefit character-
(Culley and Sotiaridis, 1983). High den- istics associated with the "prey."
sities of numerous species of various mor- Studies of gut contents sometimes in-
phological types are commonly syntopic: dicate interspecific and site differences in
15 species in one brook in Madagascar diet of tadpoles (Belova, 1964; Bugayski,
(Blommers-Schlosser and Blommers,1984); 1930; Diaz-Paniagua,1985; Farlowe, 1928;
16 lentic tadpoles in Ecuador (Berger, Gosner, 1959; Hendricks, 1973; Heyer,
1985); 11 species of calling males in south- 1974; Inger, 1986; Jenssen, 1967; Osborne
ern Mississippi(personalobservations).We and Maclahan, 1985; Sin and Gavrila,
assume that the means by which tadpoles 1977) and often suggest non-selective
partition food resources include the me- (composition,spatial, or temporal; see Ca-
chanics of oral structures.Three pertinent rothersand Jaksic, 1984) feeding and per-
questions arise: 1) What do tadpoles eat? haps uniform or random distribution of
2) How does oral morphology relate to food items. We have little idea how diet
feeding methodology? 3) How does feed- affects morphology (Meyer, 1987), life his-
ing methodology influence resource par- tory features (Anderson and Cummins,
titioning and therefore community struc- 1979; Hota and Dash, 1986), and rather
ture? little on how diet affects physiology (e.g.,
The first question might be cast better Culley and Sotiaridis,1983; Marshallet al.,
as "what do tadpoles digest?" The appar- 1980; Steinwascher and Travis, 1983).

Colostethus nubicola; (G) lotic-semiterrestrial, Thoropa petropolitana; (H) lotic-benthic, Leptopelis hyloides;
(I) lentic-arboreal, Type 3, Hyla bromeliacia; (J) lentic-suspension feeder, Type 2, Gastrophryne carolinensis;
(K) lentic-macrophagous, Type 2, Hyla microcephala; (L) lentic-macrophagous, Type 1, Occidozyga lima;
(M) lotic-adherent, Type 1, Hyla lindae; (N) lentic-adherent, Type 2, Hyla uranochroa; (0) lotic-fossorial,
Type 1, Centrolenella griffithsi; (P) lentic-arboreal, Type 2, Anotheca spinosa; (Q) lentic-carnivorous, Type
1, Ceratophrys cornuta; (R) lentic-arboreal, Type 5, Stephapaedes anotis. Original drawings by Linda Trueb;
reproduced courtesy of McGraw-Hill Publishing Co. from Duellman and Trueb, Biology of Amphibians.
98 HERPETOLOGICALMONOGRAPHS [No. 3
[No.

A E

I
1' 11'1l 1
"~l~lui'11i 1111
111l1
1k~111111111
1IIIp-IIIM,i
"'LLI'r
LIIIIICI.--IIIIII lliL Atirlil,lf11i%iitu"III 1 11,
11h '
I,S'IIIII

B F
'.
Aq
I
r i
o,
i?
i'

. I
:
il IAIA I

?~- .
v ;b:^',~
w

%..---

C G
9

D H

j 4.

_ - -
4 r. , ., ,LY ,k,
1989]
1989] HERPETOLOGICAL MONOGRAPHS 99

Three important data sets are small or of tadpoles,2) thesedifferencescrosstaxo-

lacking: temporal-spatial distribution of nomic lines, but within certain bounds
tadpoles under the influence of a number generalideas of feeding functionscan be
of variables (e.g., Alford, 1986; Alford and suggestedfrom oral morphology,and 3)
Crump, 1982; Diaz-Paniagua, 1988; within a major ecomorphologicaltype
Mitchell, 1983), microdistribution, abun- thereare othermoresubtledifferencesre-
dance, availability, and nutrient value of lating to oral function.It shouldbe kept
food items; and detailed comparative food in mind that the functionsof the gill fil-
habits of tadpoles at one site and time. As tering systemdeterminewhat eventually
normally done, gut content analyses may arrivesin the gutfordigestion(Seale,1982;
be largely non-productivebecause the ma- Sealeand Beckvar,1980;Sealeand Was-
terial digested cannot be identified and sersug,1979;Sealeet al., 1975;Wassersug,
measured easily; Bogdan and Gilbert 1984).
(1987), Caron (1983), Caswell et al. (1973), Primarilybecausewe cannotanswerthe
Coats and Heinbokel (1982), Coleman secondquestionsufficientlywell, the third
(1980), Feller (1984), Fleming and Couch- question-how feedingmethodsinfluence
lan (1978), and Rau and Anderson (1981) resourcepartitioningand thereforecom-
offer alternative but often labor-intensive munity structure-cannot be answered.
methods. Diatoms are the only easily We suggest reasonswhy present infor-
quantified objects in the gut that are likely mationhas not revealedconcordantideas
purposelyingested and digested. Available on tadpole trophic relationships.Either
information implies efficient feeding and tadpolesshowmorphological oraldiversity
digestion of abundant, nutritious, and withouttrophicspecializationsor we have
probably widespread food. Stringent par- not recognizedthe detailsof trophicdif-
titioning of food sourceshas not been dem- ferences.Thisdoes not imply that all tad-
onstrated. Perhaps the circumstances (i.e., poles are euryphagous,but oral features
non-reproductive, temporary inhabitants thatsuggestfeedingdifferencesmay have
that feed low on a short food chain with been selectedfor by that raredemandfor
abundant types and amounts of food) do actual resource partitioning. The real
not demand stringent or continuous re- questionmay concernwhat a tadpolecan
source partitioning. do when resourcesare limited, not what
There are few data (e.g., Gosner, 1959; it doesroutinely(i.e.,"competitivecrunch"
Johnston, 1982; Kenny, 1969) concerning of Wiens,1977,as comparedin Grossman
the second question-how do oral mor- et al., 1982).
phologies relate to feeding methodology? At the trophiclevel at which tadpoles
The relative utility of labial teeth and jaw presumablyfeed, food scarcityor unavail-
sheathsin different ecological types of tad- ability is probablyuncommon.Structure
poles is not yet clear. Only Johnston(1982) allows function but does not necessarily
has made interspecific and ontogenetic predict it. High morphologicaldiversity
comparisonsinvolving a substratethat can suggestshigh niche diversity.The diverse,
be penetrated by the oral structures.Only specialized morphologyof tadpoles ac-
general statements can be made: 1) there commodatesthese ideas,but the circular-
are consistent differences in oral mor- ity is obvious:the observeddiversityof
phology relative to major ecological types structures,the presumedcommonalityof

FIG. 8.-Representative oral discs of tadpoles in various guilds: (A) lotic-clasping, Rana palmipes; (B)
lotic-neustonic, Megophrys montana; (C) lotic-adherent, Hyla rivularis; (D) lotic-gastromyzophorous, Ateo
lopus ignescens; (E) lentic-macrophagous, Hyla microcephala; (F) lentic-carnivorous, Ceratophrys cornuta;
(G) lotic-suctorial, Hyla lindae; (H) lotic-adherent, Type 2, Hyla uranochroa. Original drawings by Linda
Trueb; reproduced courtesy of McGraw-Hill Publishing Co. from Duellman and Trueb, Biology of Am-
phibians.
100
100 HERPETOLOGICALMONOGRAPHS [No.
[No. 3

functions, and the availability of food ap- 1979; Collins, 1979; Crump, 1981, 1983,
pear contradictory. Dietary specializations 1984; DeBenedictus, 1974; Dupre and Pe-
can be coupled with generalized mor- tranka, 1985; Freda and Dunson, 1985;
phologies (Greene, 1982), but perhaps tad- Gatten et al., 1984; Herreid and Kinney,
poles couple dietary generalization with 1966; Heyer, 1973, 1974; Hota and Dash,
morphological specialization. Liem (1980) 1986; Inger et al., 1986; Licht, 1974; Ling
supplied pertinent comparative data on et al., 1986; Loschenkohl, 1986; Morin,
cichlid fishes. Species classified as algal eat- 1981, 1983; Noland and Ultsch, 1981;
ers were more specialized morphologically Odendaal and Bull, 1983; Odendaal and
but had a relatively wide range of func- Nias, 1982; Seale, 1980; Semlitsch and
tional activities; pursuit hunters had a more Caldwell, 1982; Smith, 1983; Smith-Gill
generalized morphology coupled with a and Gill, 1978; Steinwascher, 1978; Travis,
restricted functional repertoire. Such data 1981, 1983a,b; Wassersug and Murphy,
emphasize the danger of predicting func- 1987; Wassersug and Seibert, 1975; Weig-
tion and ecological latitude solely from mann, 1982; Wilbur, 1972; Wilbur et al.,
morphology. 1983; Wiltshire and Bull, 1977; Wood-
We predict that most tadpoles can feed ward, 1982a,b, 1983). Laboratory data on
in a variety of probably minutely different competition for food (e.g., Steinwascher,
ways on a number of resources, but all 1979a,b; Travis, 1980a,b; Wilbur, 1982)
tadpoles do not have equal capabilities. reveal competitive interactions that cause
Variations in keratinized oral features may profound effects, but design constraints
function selectively only on those occasions disallow many factors (e.g., space, time,
when food limitations demand partition- alternate food sources) that may reduce or
ing. Also, many differences (e.g., number eliminate these effects in the field. It would
and arrangement of cusps on labial teeth) seem that factors associated with the feed-
may have arisen through fixation of var- ing niches of at least representative tadpole
ious contingencies of polygenic inheri- morphotypes must be unveiled before de-
tance (e.g., linkage, pleiotropy, position ef- tailed descriptions of communities of tad-
fect), while selection was for an poles can be made.
unidentified factor(s) or heterochronic Spatial segregation suggested by body,
changes in development. The cellular basis fin, and eye configurations, especially when
for pattern formation must be resolved be- coupled with diel and seasonal differences,
fore an understanding of some traits will may be a prime mechanism to allow par-
be understood. Crosses of species with dif- titioning of food resources that are difficult
ferent tooth row formulae carried through to partition directly (Toft, 1985). The de-
several generations would suggest mech- gree of interspecific or ontogenetic differ-
anisms of genetic control. Proper analyses ences in feeding ecology may be less than
of variation (e.g., Palmer and Strobeck, or at least not concordant with the differ-
1986) may allow ranking of the relative ences in magnitude of the more easily seen
importance of features without knowing morphological traits. We must retreat to
their actual functions. The functional sig- autecology (Greene, 1986), functional
nificance of the numerous configurations anatomy, digestive physiology, and closely
of tadpole vent tubes (Johnston and Altig, controlled studies of many facets of field
1986) and microdermatoglyphics of snake ecology before feeding ecology can be used
scales judged to be of taxonomic but not in community definitions of tadpoles to
ecological significance (Price, 1982) may any accurate degree. Functional mor-
be analogous cases (see Gans, 1976). phology of oral features to complement the
Data on various facets of tadpole ecol- data on gill filter systems and detection of
ogy, environmental effects, and metabolic the subtle differences in resource parti-
functions are available (e.g., Alford, 1986; tioning are the most pertinent and prob-
Alford and Crump, 1982; Berven, 1982a,b; ably the most difficult data to obtain.
Black, 1976; Breden and Kelly, 1982;
Brockelman, 1969; Brodie and Formano- Acknowledgments. -We graciously acknowledge
wicz, 1983; Calef, 1973; Cecil and Just, Linda Trueb, William E. Duellman, and McGraw-
1989]
1989] HERPETOLOGICAL MONOGRAPHS 101

Hill Publishing Company for allowing us to reprint Environments Madagascar. Pergamon Press, Ox-
figures appearing originally in Biology of Amphib- ford.
ians (1986). D. G. Thibaudeau prepared the other BOGDAN, K. G., AND J. J. GILBERT. 1987. Quanti-
figures. A. Channing, M. L. Crump, R. W. Mc- tative comparison of food niches of some fresh-
Diarmid, and R. J. Wassersug were the latest in a water zooplankton. A multi-tracer-cell approach.
long series of persons to substantially improve this Oecologia 72:331-340.
constantly evolving manuscript; we thank them and BOULENGER, C. A. 1891. A synopsis of the tadpoles
all the others for numerous comments and sugges- of the European batrachians. Proc. Zool. Soc. Lon-
tions. Specimens were examined at the California don 61:593-627.
Academy of Sciences, Field Museum of Natural His- BOWEN, S. H. 1980. Detrital nonprotein amino acids
tory (R. M. Louer Fund Award to GFJ), United States are the key to rapid growth in Tilapia in Lake
National Museum of Natural History (Smithsonian Valencia, Venezuela. Science 207:1216-1218.
Visiting Student Research Appointment to GFJ), and BRAGG, A. N., AND W. N. BRAGG. 1959. Variations
the University of Michigan Museum of Zoology; we in the mouth parts in tadpoles of Scaphiopus (Spea)
thank the respective curators for their assistance. bombifrons Cope (Amphibia: Salientia). Southw.
Nat. 3:55-69.
LITERATURE CITED BRAGG, A. N., AND S. HAYES. 1963. A study of labial
teeth rows in tadpoles of Couch's spadefoot. Was-
ALFORD, R. A. 1986. Habitat use and positional mann J. Biol. 21:149-154.
behavior of anuran larvae in a northern Florida BREDEN, F., AND C. H. KELLY. 1982. The effect of
temporary pond. Copeia 1986:408-423. conspecific interactions on metamorphosis in Bufo
ALFORD,R. A., AND M. L. CRUMP. 1982. Habitat americanus. Ecology 63:1682-1689.
partitioning among size classes of larval southern BREURE,A. S. H., AND E. GITTENBERGER.1982.
leopard frogs, Rana utricularia. Copeia 1982:367- The rock-scraping radula, a striking case of con-
373. vergence (Mollusca). Netherlands J. Zool. 32:307-
ALTIG, R. 1970. A key to the tadpoles of the con- 312.
tinental United States and Canada. Herpetologica BROCKELMAN, W. Y. 1969. An analysis of density
26:180-207. effects and predation in Bufo americanus tadpoles.
.1973. Preliminary scanning electron ob- Ecology 50:632-644.
servations of keratinized structures in amphibian BRODIE,E. D., JR., ANDD. FORMANOWICZ, JR. 1983.
larvae. HISS News-J. 1:129-131. Prey size preference of predators: Differential vul-
ALTIG, R., ANDE. D. BRODIE,JR. 1972. Laboratory nerability of larval anurans. Herpetologica 39:67-
behavior of Ascaphus truei tadpoles. J. Herpetol. 74.
6:21-24. BUGAYSKI, W. 1930. O pozywieniu kijanek zaby
ALTIG, R., AND G. F. JOHNSTON. 1986. Major char- ostronosej (Rana arvalis Nils.) w stawach. Arch.
acteristics of free-living anuran tadpoles. Smith- Hydrobiol. Rybactwa 5:285-304.
sonian Herpetol. Info. Serv. (67):1-75. CALEF, G. W. 1973. Natural mortality of tadpoles
ALTIG,R., AND W. MCDEARMAN. 1975. Clearance in a population of Rana aurora. Ecology 54:741-
times and percent assimilation of five species of 758.
anuran tadpoles. Herpetologica 31:67-69. CARON, D. A. 1983. Technique for enumeration of
ALTIG, R., AND W. L. PACE. 1974. Scanning elec- heterotrophic and phototrophic nanoplankton, us-
tron photomicrographs of tadpole labial teeth. J. ing epifluorescence microscopy, and comparison
Herpetol. 8:247-251. with other procedures. Appl. Environ. Microbiol.
ANDERSON,N. H., AND K. W. CUMMINS. 1979. In- 46:491-498.
fluences of diet on the life histories of aquatic in- CAROTHERS, J. H., AND F. M. JAKSIC. 1984. Time
sects. J. Fish. Res. Bd. Canada 36:335-342. as a niche difference: The role of interference com-
BELOVA, Z. A. 1964. On the feeding habits of the petition. Oikos 42:403-406.
tadpoles of Rana ridibunda Pall. in the Volga Delta. CASWELL,H., F. REED, S. N. STEPHENSON,AND P.
Zool. Zhurnal 43:1188-1192. A. WERNER. 1973. Photosynthetic pathways and
BERGER, T. J. 1985. Community Ecology of Pond- selected herbivory: An hypothesis. Am. Nat. 107:
dwelling Anuran Larvae. Ph.D. Dissertation, Uni- 465-480.
versity of Kansas, Lawrence. CECIL, S. G., AND J. J. JUST. 1979. Survival rate,
BERVEN, K. A. 1982a. The genetic basis of altitu- population density and development of a naturally
dinal variation in the woodfrog Rana sylvatica. I. occurring anuran larva (Rana catesbeiana). Copeia
An experimental analysis of life history traits. Evo- 1979:447-453.
lution 36:962-983. CHANNING, A. 1984. Identification of ghost frog
1982b. The genetic basis of altitudinal vari- tadpoles. Proc. Electron Micros. Soc. S. Africa 14:
ation in the woodfrog Rana sylvatica. II. An ex- 89-90.
perimental analysis of larval development. Oeco- COATS,D. W., ANDJ. F. HEINBOKEL.1982. A study
logia 52:360-369. of reproduction and other life cycle phenomena in
BLACK, J. H. 1976. Environmental fluctuations in planktonic protists using an acridine orange fluo-
central Oklahoma temporary ponds. Proc. Okla- rescence technique. Mar. Biol. 67:71-79.
homa Acad. Sci. 56:1-8. COLEMAN,A. W. 1980. The use of DAPI for iden-
BLOMMERS-SCHLOSSER, R. M. A., AND L. H. M. tifying and counting aquatic microflora. Limnol.
BLOMMERS. 1984. The amphibians. Pp. 89-104. Oceanog. 25:943-948.
In A. Jolly, P. Oberle, and R. Albignac (Eds.), Key COLLINS, J. P. 1979. Intrapopulation variation in
102 HERPETOLOGICAL MONOGRAPHS
HERPETOLOGICAL MONOGRAPHS [No.
[No. 33

the body size at metamorphosis and timing of meta- DUPRE, K., AND J. W. PETRANKA. 1985. Ontogeny
morphosis in the bullfrog Rana catesbeiana. Ecol- of temperature selection in larval anurans. Copeia
ogy 60:739-749. 1985:462-467.
COSTA,H. H., ANDS. BALASUBRAMANIAN. 1965. The DUTTA, S. K., AND P. MOHANTY-HEJMADI.1984.
food of the tadpoles of Rhacophorus cruciger cru- Ontogeny of teeth row structure in Rana tigerina
ciger (Bluth). Ceylon J. Sci. 5:107-109. tadpoles. J. Bombay Nat. Hist. Soc. 80:517-528.
CRUMP,M. L. 1981. Energy accumulation and am- ELINSON,R. P. 1987. Change in developmental pat-
phibian metamorphosis. Oecologia 48:167-169. terns: Embryos of amphibians with large eggs. Pp.
1984. Ontogenetic changes in vulnerability 1-21. In R. A. Raff and E. C. Raff (Eds.), Devel-
to predation in tadpoles of Hyla pseudopuma. Her- opment as an Evolutionary Process. Alan R. Liss,
petologica 40:265-271. Inc., New York.
1986. Cannibalism by younger tadpoles: FAIER, J. M. 1972. Larval Anuran Teeth and Some
Another hazard of metamorphosis. Copeia 1986: Ecologic Implications. M.S. Thesis, University of
1007-1009. Nebraska, Lincoln.
.1989. Life history consequences of feeding FARLOWE,V. 1928. Algae of ponds as determined
versus non-feeding in a facultatively non-feeding by an examination of the intestinal contents of tad-
toad larva. Oecologia:In press. poles. Biol. Bull. 35:443-448.
CULLEY, D. D., JR., AND P. K. SOTIARIDIS. 1983. FELLER,R. J. 1984. Dietary immunoassay of Ilya-
Progress and problems associated with bullfrog tad- nassa obsoleta, the eastern mud snail. Biol. Bull.
pole diets and nutrition. Third Annual Dr. Scholl 166:96-102.
Conf. Nutr. Captive Wild Animals.
FLEMING,J. M., AND J. COUCHLAN. 1978. Preser-
CUMMINS,K. W., AND M. J. KLUG. 1979. Feeding vation of vitally stained zooplankton for live/dead
ecology of stream invertebrates. Ann. Rev. Ecol. sorting. Estuaries 1:135-137.
Syst. 10:147-172.
FREDA, J., AND W. A. DUNSON. 1985. Field and
DEBENEDICTUS,P. A. 1974. Interspecific compe-
tition between tadpoles of Rana pipiens and Rana laboratory studies of ion balance and growth rates
of ranid tadpoles chronically exposed to low pH.
sylvatica: An experimental field study. Ecol.
Copeia 1985:415-423.
Monog. 44:129-151.
DE JONGH, H. J. 1968. Functional morphology of GANS, C. 1976. The biological roles of taxonomic
the jaw apparatus of larval and metamorphosing characteristics utilized in amphisbaenian classifi-
Rana temporaria L. Netherlands J. Zool. 18:1-103. cation. British J. Herpetol. 5:611-615.
DE LOPEZ,L. E. F., AND D. D. ECHEVERRIA.1984. GATTEN, R. E., JR., J. P. CALDWELL, AND M. E.
STOCKARD.1984. Anaerobic metabolism during
Morfogenesis de los dientes larvales y pico corneo
de Bufo arenarum (Anura, Bufonidae). Zoologia intense swimming by anuran larvae. Herpetologica
13:573-578. 40:164-169.
DEUNFF,J., AND A. BEAUMONT. 1959. Histogenese GOICOECHEA,O., O. GARRIDO,AND B. JORQUERA.
des dents et dueber conres chez les larves de Dis- 1986. Evidence for a trophic paternal-larval re-
coglossus pictus Otth. C. R. Soc. Biol. 1959:1162- lationship in the frog Rhinoderma darwinii. J. Her-
1164. petol. 20:168-178.
DE VILLIERS,C. G. S. 1929. Some features of the GOLDACRE, R. J. 1949. Surface films on natural
early development of Breviceps. Ann. Transvaal bodies of water. J. Anim. Ecol. 18:36-39.
Mus. 13:142-151. GOSNER,K. L. 1959. Systematic variations in tad-
DIAZ-PANIAGUA, C. 1985. Larval diets related to pole teeth with notes on food. Herpetologica 15:
morphological characters of five anuran species in 203-210.
the biological reserve of Donana (Huelva, Spain). 1960. A simplified table for staging anuran
Amphibia-Reptilia 6:307-322. embryos and larvae with notes on identification.
1988. Temporal segregation in larval am- Herpetologica 16:183-190.
phibian communities in temporary ponds at a lo- GOSNER,K. L., AND I. H. BLACK. 1957. Larval
cality in SW Spain. Amphibia-Reptilia 9:15-26. development in New Jersey Hylidae. Copeia
DICKMAN,M. 1968. The effect of grazing by tad- 1957(1):31-36.
poles on the structure of a periphyton community. GRADWELL,N. 1968. The jaw and hyoidean mech-
Ecology 49:1188-1190. anism of the bullfrog tadpole during aqueous ven-
DREWES,R. C. 1984. A phylogenetic analysis of the tilation. Canadian J. Zool. 46:1041-1052.
Hyperolidae (Anura): Treefrogs of Africa, Mada- 1970. The function of the ventral velum
gascar, and the Seychelles Islands. Occas. Pap. Cal- during gill irrigation in Rana catesbeiana. Cana-
ifornia Acad. Sci. 139:1-70. dian J. Zool. 48:1179-1186.
DRUM, R. W., AND J. T. HOPKINS. 1966. Diatom 1971. Ascaphus tadpole: Experiments on
locomotion: An explanation. Protoplasma 62:1-33. the suction and gill irrigation mechanisms. Cana-
DUBOIS,A. 1986. Diagnose preliminaire d'un nou- dian J. Zool. 49:307-332.
veau genre de Ranoidea (Amphibiens, Anoures) du . 1973. On the functional morphology of
sud de 1'Inde. Alytes 4:113-118. suction and gill irrigation in the tadpole of Asca-
DUELLMAN,W. E. 1970. Hylid frogs of Middle phus, and notes on hibernation. Herpetologica 29:
America. Univ. Kansas Monog. (1):1-753. 84-93.
DUELLMAN,W. E., AND L. TRUEB. 1986. Biology 1975a. The clinging mechanism of Pseu-
of Amphibians. McGraw-Hill Book Co., New York. dophryne bibroni (Anura: Leptodactylidae) to an
1989] HERPETOLOGICAL MONOGRAPHS 103

alga on glass. Trans. Roy. Soc. S. Australia 99:31- from a seasonal tropical location in Thailand. Ecol-
34. ogy 55:651-656.
.1975b. Experiments on oral suction and HEYER,W. R., A. S. RAND, C. A. G. DA CRUZ,O. L.
gill breathing in five species of Australian tadpole PEIXOTO,AND C. E. NELSON. 1990. Frogs of Bo-
(Anura: Hylidae and Leptodactylidae). J. Zool. raceia. Arq. Zool.:In press.
London 177:81-98. HILLIS,D. M., AND R. DE SA. 1988. Phylogeny and
GRADWELL,N., AND V. M. PASZTOR.1968. Hydro- taxonomy of the Rana palmipes group (Salientia:
static pressures during normal ventilation in the Ranidae). Herpetol. Monog. 2:1-26.
bullfrog tadpole. Canadian J. Zool. 46:1169-1174. HOFF, K., AND R. WASSERSUG.1985. Do tadpoles
GRANDISON,A. G. C. 1981. Morphology and phy- swim like fishes? Fortsch. Zool. 30:31-34.
logenetic position of the West African Didynami- 1986. The kinematics of swimming in lar-
pus sjoestedti Andersson, 1903 (Anura Bufonidae). vae of the clawed frog, Xenopus laevis. J. Exp.
Monit. Zool. Italia, Suppl. 15:187-215. Biol. 122:1-12.
GREENE, H. W. 1982. Dietary and phenotypic di- HOLLAND, A. F., R. G. ZINGMARK, AND J. M. DEAN.
versity in lizards: Why are some organisms spe- 1974. Quantitative evidence concerning stabili-
cialized? Pp. 107-128. In D. Mossakowski and G. zation of sediments by marine benthic diatoms.
Roth (Eds.), Environmental Adaptation and Evo- Mar. Biol. 27:191-196.
lution. Gustav Fischer, Stuttgart. HOTA, A. K., AND M. C. DASH. 1986. Growth and
1986. Natural history and evolutionary bi- metamorphosis of anuran larvae: Effect of diet and
ology. Pp. 99-108. In M. E. Feder and G. V. Lauder temperature. Alytes 5:165-172.
(Eds.), Predator-Prey Relationships. Perspectives INGER,R. F. 1967. The development of a phylogeny
and Approaches from the Study of Lower Verte- of frogs. Evolution 21:369-384.
brates. Univ. Chicago Press, Chicago. 1986. Diets of tadpoles living in a Bornean
GRIFFITHS,I. G. 1951. The form and function of rain forest. Alytes 5:153-164.
the fore-gut in anuran larvae (Amphibia, Salientia) INGER, R. F., H. K. VORIS, AND K. J. FROGNER. 1986.
with particular reference to the manicotto glan- Organization of a community of tadpoles in rain
dulare. Proc. Zool. Soc. London 137:249-283. forest streams in Borneo. J. Tropical Ecol. 2:193-
GRIFFITHS,I. G., AND A. L. DE CARVALHO. 1965. 205.
On the validity of employing larval characters as JENSSEN,T. A. 1967. Food habits of the green frog,
major phyletic indices in Amphibia, Salientia. Rev. Rana clamitans, before and during metamorphosis.
Brasil. Biol. 25:113-121. Copeia 1967:214-218.
GRILLITSCH,B., AND H. GRILLITSCH.1989. Tera- JOHNSTON, G. F. 1982. Functions of the Keratinized
tological and ontogenetic alterations to external oral Oral Features in Anuran Tadpoles with an Analysis
structures in some anuran larvae (Amphibia: An- of Ecomorphological Tadpole Types. M.S. Thesis,
ura: Bufonidae, Ranidae). Fortschrit. Zool. Stutt- Mississippi State University, Mississippi State.
gart:In press. JOHNSTON,G. F., AND R. ALTIG. 1986. Identifica-
GROSSMAN, G. D., P. B. MOYLE, AND J. O. WHITAKER, tion characteristics of anuran tadpoles. Herpetol.
JR. 1982. Stochasticity in structural and func- Rev. 17:36-37.
tional characteristics of an Indiana stream fish as- KENNY, J. S. 1969. Feeding mechanisms in anuran
semblage: A test of community theory. Am. Nat. larvae. J. Zool. 157:225-246.
120:423-454. P. 1957. Amphibia of Ceylon. Pri-
KIRTISINGHE,
HAERTEL, J. D., AND R. M. STORM. 1970. Experi-
mental hybridization between Rana pretiosa and vately published.
Rana cascadae. Herpetologica 26:436-446. KLUGE, A. G., AND J. S. FARRIS. 1969. Quantitative
HAMPTON,S. H., AND E. P. VOLPE. 1963. Devel- phyletics and the evolution of anurans. Syst. Zool.
18:1-32.
opment and interpopulation variability of the
KUANG, H.-L. C., AND J. J. KOLLROS. 1976. Cell
mouthparts of Scaphiopus holbrooki. Am. Midi. Nat.
70:319-328. turnover in the beak of Rana pipiens. Anat. Rec.
188:361-370.
HARPER,M. 1977. Movements. Pp. 224-249. In The
Biology of Diatoms. Biol. Monog., Univ. Calif. Press LANNOO, M. J. 1985. Neuromast topography in
(13). Ambystoma larvae. Copeia 1985:535-539.
HENDRICKS,F. S. 1973. Intestinal contents of Rana LANNOO, M. J., D. S. TOWNSEND, AND R. J. WAS-
pipiens, Schreber (Ranidae) larvae. Southw. Nat. SERSUG.1987. Life in the leaves: Arboreal tadpole
18:93-114. types, with special attention to the morphology,
HERON-ROYER,AND C. VAN BAMBEKE. 1889. Le ecology, and behavior of the oophagous Osteopilus
vestibule de la bouche chez les tetards des batra- brunneus (Hylidae) larva. Fieldiana, Zool. (38):1-
ciens anoures d'Europe; sa structure, ses caracteres 31.
chez les diverses especies. Arch. Biol. 9:185-309. LIGHT, L. E. 1974. Survival of embryos, tadpoles,
HERREID,C. F., AND S. KINNEY. 1966. Survival of and adults of the frogs Rana aurora aurora and
Alaskan woodfrog (Rana sylvatica) larvae. Ecology Rana pretiosa pretiosa sympatric in southwestern
47:1039-1041. British Columbia. Canadian J. Zool. 52:613-627.
HEYER,W. R. 1973. Ecological interactions of frog LIEM, K. F. 1980. Adaptive significance of intra-
larvae at a seasonal tropical location in Thailand. and interspecific differences in the feeding reper-
J. Herpetol. 7:337-361. toires of cichlid fishes. Am. Zool. 20:295-314.
1974. Niche measurements of frog larvae LIMBAUGH,B. A., AND E. P. VOLPE. 1957. Early
104 HERPETOLOGICAL MONOGRAPHS [No.
[No. 33

development of the Gulf Coast toad, Bufo valliceps movements, tadpole competition and limits of the
Wiegmann. Am. Mus. Novitates (1842):1-32. distribution of the frogs Ranidella riperia and R.
LING, R. W., J. P. VAN AMBERG,ANDJ. K. WERNER. signifera. Oecologia 57:361-367.
1986. Pond acidity and its relationship to larval ODENDAAL,F. J., AND R. C. NIAS. 1982. Habitat
development of Ambystoma maculatum and Rana selection in tadpoles of Ranidella signifera and R.
sylvatica in upper Michigan. J. Herpetol. 20:230- riperia (Anura: Leptodactylidae). Oecologia 52:411-
236. 414.
LIu, C. C. 1940. Tadpoles of West China Salientia. ORTON,G. L. 1953. The systematics of vertebrate
J. W. China Border Res. Soc. 12B:7-62. larvae. Syst. Zool. 2:63-75.
LOSCHENKOHL,A. 1986. Niche partitioning and . 1957. The bearing of larval evolution on
competition in tadpoles. Stud. Herpetol.:399-402. some problems in frog classification. Syst. Zool. 6:
LUCKENBILL,L. M. 1964. Fine Structure and De- 79-86.
velopment of Horny Jaws and Teeth of the Tadpole OSBORNE, P. L., AND A. J. MACLAHAN. 1985. The
(Rana pipiens). Ph.D. Dissertation, Brown Uni- effect of tadpoles on algal growth in temporary,
versity, Providence. rain-filled pools. Freshwat. Biol. 15:77-87.
1965. Morphogenesis of the horny jaws of PALMER, A. R., AND C. STROBECK. 1986. Fluctuat-
Rana pipiens larvae. Develop. Biol. 11:25-49. ing asymmetry: Measurement, analysis, patterns.
LYNCH, J. D. 1971. Evolutionary relationships, os- Ann. Rev. Ecol. Syst. 17:391-421.
teology, and zoogeography of leptodactyloid frogs. POTTHOFF, T. L., AND J. D. LYNCH. 1986. Inter-
Misc. Publ. Mus. Nat. Hist., Univ. Kansas (53):1- population variability in mouthparts in Scaphiopus
238. bombifrons in Nebraska (Amphibia: Pelobatidae).
MARSHALL,G. A., R. L. AMBORSKI, AND D. D. CUL- Prairie Nat. 18:15-22.
LEY. 1980. Calcium and pH requirements in the PRICE, R. M. 1982. Dorsal snake scale microder-
culture of bullfrog (Rana catesbeiana) larvae. Proc. matoglyphics: Ecological indicator or taxonomic
World Maricul. Soc. 11:445-453. tool. J. Herpetol. 16:294-306.
MERTENS,R. 1960. Die Larven der Amphibien und PYBURN,W. F. 1975. An unusual anuran larva from
ihre evolutive Bedeutung. Zool. Anz. 164:337-358. the Vaupes' region of southeastern Colombia. Pap.
MEYER,A. 1987. Phenotypic plasticity and heter- Avulsos Zool. 33: 231-238.
ochrony in Cichlasoma managuense (Pisces, Cich- RAFF, R. A. 1987. Constraint, flexibility, and phy-
lidae) and their implications for speciation in cich- logenetic history in the evolution of direct devel-
lid fishes. Evolution 41:1357-1369. opment in sea urchins. Develop. Biol. 119:6-19.
MICHALOWSKI, J. 1966. Studies on the relationship RAU, G. H., AND N. H. ANDERSON. 1981. Use of
of Bombina bombina (Linnaeus) and Bombina var- 13C/12C to trace dissolved and particulate organic
iegata (Linnaeus) II. Some taxonomic characters of matter utilization by populations of an aquatic or-
tadpoles of both species and of tadpoles obtained ganism. Oecologia 48:19-21.
from crosses under laboratory conditions. Acta Zool. SAVAGE,R. M. 1952. Ecological, physiological and
Cracoviensia 11:1-94. anatomical observations of some species of anuran
MITCHELL,S. L. 1983. Spatial Distribution of Eight tadpoles. Proc. Zool. Soc. London 122:467-514.
Species of Anuran Tadpoles. M.S. Thesis, Mississip- .1955. The ingestive, digestive and respi-
pi State University, Mississippi State. ratory systems of the microhylid tadpole, Hypo-
MORIN,P. J. 1981. Predatory salamanders reverse pachus aquae. Copeia 1955:120-127.
the outcome of competition among three species SCOTT, N. J., JR., AND R. D. JENNINGS. 1985. The
of anuran tadpoles. Science 212:1284-1286. tadpoles of five species of New Mexican leopard
1983. Predation, competition, and the com- frogs. Occas. Pap. Mus. Southw. Biol. 3:1-21.
position of larval anuran guilds. Ecol. Monog. 53: SEALE, D. B. 1980. Influence of amphibian larvae
119-138. on primary production, nutrient flux, and compe-
NACHTIGALL,W. 1974. Biological mechanisms of tition in a pond ecosystem. Ecology 61:1531-1550.
attachment: The comparative morphology and 1982. Obligate and facultative suspension
bioengineering of organs for linkage, suction and feeding in anuran larvae: Regulation in Xenopus
adhesion. Springer-Verlag, Berlin. and Rana. Biol. Bull. 162:214-231.
NICHOLLS, R. J. 1937. Taxonomic studies on the SEALE, D. B., AND N. BECKVAR. 1980. The com-
mouth parts of larval Anura. Illinois Biol. Monog. parative ability of anuran larvae (Genera: Hyla,
15:1-73. Bufo, and Rana) to ingest suspended blue-green
NISHIKAWA, K., AND R. WASSERSUG. 1989. Mor- algae. Copeia 1980:495-503.
phology of the caudal spinal cord in Rana (Ran- SEALE, D. B., E. RODGERS, AND M. E. BORAAS. 1975.
idae) and Xenopus (Pipidae) tadpoles. J. Comp. Effects of suspension-feeding frog larvae on lim-
Neurol.:In press. nological variables and community structure. Verh.
NOBLE,G. K. 1931. The Biology of the Amphibia. Int. Verein. Limnol. 19:3179-3184.
McGraw-Hill Book Co., New York. SEALE, D. B., AND R. J. WASSERSUG. 1979. Suspen-
NOLAND, R., AND G. R. ULTSCH. 1981. The roles sion feeding dynamics of anuran larvae related to
of temperature and dissolved oxygen in microhab- their functional morphology. Oecologia 39:259-272.
itat selection by the tadpoles of a frog (Rana pip- SEMLITSCH, R. D., AND J. P. CALDWELL. 1982. Ef-
lens) and a toad (Bufo terrestris). Copeia 1981: fects of density on growth, metamorphosis, and
645-652. survivorship in tadpoles of Scaphiopus holbrooki.
ODENDAAL, F. J., AND C. M. BULL. 1983. Water Ecology 63:905-911.
1989]
1989] HERPETOLOGICAL MONOGRAPHS 105

SHELTON, P. M. J. 1970. The lateral line system at 1980b. Genetic variation for larval specific
metamorphosis in Xenopus laevis (Daudin). J. Em- growth rate in the frog Hyla gratiosa. Growth 44:
bryol. Exp. Morphol. 24:511-524. 167-181.
SIN, G., AND L. GAVRILA. 1977. Study of the diet .1981. Control of larval growth variation in
of Rana ridibunda tadpoles. Stud. Cercet. Biol. 29: a population of Pseudacris triseriata (Anura: Hy-
93-98. lidae). Evolution 35:423-432.
SMITH, D. C. 1983. Factors controlling tadpole pop- 1983a. Variation in growth and survival of
ulations of the chorus frog (Pseudacris triseriata) Hyla gratiosa larvae in experimental enclosures.
on Isle Royale, Michigan. Ecology 64:501-510. Copeia 1983:232-237.
SMITH-GILL, S. J., AND D. E. GILL. 1978. Curvilin- 1983b. Variation in developmental pat-
earities in the competition equations: An experi- terns of larval anurans in temporary ponds. I. Per-
ment with ranid tadpoles. Am. Nat. 112:557-570. sistent variation within a Hyla gratiosa population.
SOKOL, O. M. 1975. The phylogeny of anuran lar- Evolution 37:496-512.
vae: A new look. Copeia 1975:1-23. TYLER, M. J. 1963. A taxonomic study of amphib-
STARRETT, P. H. 1973. Evolutionary patterns in ians and reptiles of the Central Highlands of New
larval morphology. Pp. 251-271. In J. L. Vial (Ed.), Guinea, with notes on their ecology and biology.
Evolutionary Biology of the Anurans. Contempo- Trans. Roy. Soc. S. Australia 86:105-130.
rary Research on Major Problems. Univ. Missouri VAN DIJK, D. E. 1972. The behavior of southern
Press, Columbia. African anuran tadpoles with particular reference
STEINWASCHER, K. 1978. Interference and exploi- to their ecology and related external morphological
tation competition among tadpoles of Rana utricu- features. Zool. Africana 7:49-55.
laria. Ecology 59:1039-1046. .1981. Material data other than preserved
1979a. Competitive interactions among specimens. Monit. Zool. Italia Suppl. 15:393-400.
tadpoles: Responses to a resource level. Ecology 60: VIERTEL, B. 1982. The oral cavities of Central Eu-
1172-1183. ropean anuran larvae (Amphibia) morphology, on-
1979b. Host-parasite interaction as a po- togenesis and generic diagnosis. Amphibia-Reptilia
tential population-regulating mechanism. Ecology 4:327-360.
60:884-890. 1984a. Filtration, eine Strategie der Nah-
STEINWASCHER, K., AND J. TRAVIS. 1983. Influence rungsaufnahme der Larven von Xenopus laevis,
of food quality and quantity on early larval growth Rana temporaria und Bufo calamita (Amphibia,
of two anurans. Copeia 1983:238-242. Anura). Verh. Gesell. Okologie (Berlin) 12:563-575.
STENECK, R. S., AND L. WATLING. 1982. Feeding 1984b. Suspension feeding of the larvae of
capabilities and limitation of herbivorous mollusc: Baleaphryne muletensis. Hist. Biol. Ferreret. Ciu-
A functional group approach. Mar. Biol. 68:299- tat de Mallorca Edit. Moll. 153-161.
319. . 1985. The filter apparatus of Rana tem-
TACHIBANA,T. 1978. The Merkel cell in the labial poraria and Bufo bufo larvae (Amphibia, Anura).
ridge epidermis of anuran tadpole. I. Fine struc- Zoomorphology 105:345-355.
ture, distribution and cytochemical studies. Anat. 1987. The filter apparatus of Xenopus lae-
Rec. 191:487-502. vis, Bombina variegata, and Bufo calamita (Am-
TAKAHAMA, H., F. SASAKI,AND K. WATANABE. 1988. phibia, Anura): A comparison of different larval
Morphological changes in the oral (buccopharyn- types. Zool. Jb. Anat. 115:425-452.
geal) membrane in urodelan embryos: Develop- VISSER, J. 1985. The fang-like teeth of the early
ment of the mouth opening. J. Morphol. 195:59- larvae of some Heleophryne. S. African J. Sci. 81:
69. 200-202.
TEST, F. H., AND R. G. MCCANN. 1976. Modifi- WAKE, M. H. 1982. Diversity within a framework
cation of foraging behavior by tadpoles of Bufo of constraints. Amphibian reproductive modes. Pp.
americanus in response to high densities of swim- 87-106. In D. Mossakowski and G. Roth (Eds.),
ming micro-organisms. Copeia 1976:576-578. Environmental Adaptation and Evolution. Gustav
THIBAUDEAU, D. G., AND R. ALTIG. 1988. Sequence Fischer, Stuttgart.
of ontogenetic development and atrophy of the oral WARREN, E. 1922. Observations on the develop-
apparatus of six anuran tadpoles. J. Morphol. 197: ment of the non-aquatic tadpole of Anhydrophryne
63-69. rattrayi Hewitt. S. African J. Sci. 19:254-163.
THRALL, J. H. 1972. Food, Feeding, and Digestive WASSERSUG, R. J. 1972. The mechanism of ultra-
Physiology of the Larval Bullfrog, Rana cates- planktonic entrapment in anuran larvae. J. Mor-
beiana Shaw. Ph.D. Dissertation, Illinois State Uni- phol. 137:279-288.
versity, Bloomington. 1980. Internal oral features of larvae from
TOFT, C. A. 1985. Resource partitioning in am- eight anuran families: Functional, systematic, evo-
phibians and reptiles. Copeia 1985:1-21. lutionary and ecological considerations. Misc. Publ.
TOWNSEND,D. S., AND M. M. STEWART. 1985. Di- Mus. Nat. Hist., Univ. Kansas (68):1-146.
rect development in Eleutherodactylus coqui (An- .1984. The Pseudohemisus tadpole: A mor-
ura, Leptodactylidae): A staging table. Copeia 1985: phological link between microhylid (Orton type 2)
423-436. and ranoid (Orton type 4) larvae. Herpetologica
TRAVIS, J. 1980a. Phenotypic variation and the out- 40:138-149.
come of interspecific competition in hylid tadpoles. In press. Locomotion in amphibian larvae
Evolution 34:40-50. (or "Why aren't tadpoles built like fishes?").
106 HERPETOLOGICAL MONOGRAPHS [No. 3

WASSERSUG, R. J., AND W. R. HEYER. 1983. Mor- 1983. Tadpole size and predation in the
phological correlates of subaerial existence in lep- Chihuahuan Desert. Southw. Nat. 28:470-471.
todactylid tadpoles associated with flowing water.
Canadian J. Zool. 61:761-769. Accepted: 25 October 1988
R. J., AND K. HOFF. 1985. The kine- Associate Editor: Raymond D. Semlitsch
WASSERSUG,
matics of swimming in anuran larvae. J. Exp. Biol.
119:1-30.
APPENDIX
WASSERSUG, R. J., ANDA. M. MURPHY. 1987. Aerial
respiration facilitates growth in suspension-feeding
anuran larvae (Xenopus laevis). Exp. Biol. 46:141- Developmental Guilds of
147. Endotrophic Anurans
WASSERSUG, R. J., AND W. F. PYBURN. 1987. The
"Live-bearing" is a confusing term. We restrict its
biology of the Pe-ret' toad, Otophryne robusta (Mi- use to situations where a froglet is birthed from the
crohylidae), with special consideration of its fos-
sorial larva and systematic relationships. Zool. J. oviducts, i.e., viviparous or ovoviviparous. To avoid
a new term for males that carry eggs and embryos,
Linnean Soc. 91:137-169.
WASSERSUG, R. J., AND K. ROSENBERG.1979. Sur- gravid is extended to either parent in the next four
face anatomy of branchial food traps of tadpoles: guilds associated with eggs or larvae. Guild names
A comparative study. J. Morphol. 159:393-426. are in boldface, and taxonomic listings are not always
exhaustive.
WASSERSUG,R. J., AND E. A. SEIBERT. 1975. Be-
havioral responses of amphibian larvae to variation 1. Viviparous: Froglet birthed from oviducts; nu-
in dissolved oxygen. Copeia 1975:86-103. trition provided by maternal oviducal materials
WATSON, G. F., AND A. A. MARTIN. 1973. Life after yolk exhaustion; placenta not required
history, larval morphology and relationships of (Wake, 1982); fertilization internal; develop-
Australian leptodactylid frogs. Trans. Roy. Soc. S. ment highly modified; fetus applies to advanced
Australia 97:33-45. unborns-Nectophrynoides occidentalis only.
WEBB, R. G., AND J. K. KORKY. 1977. Variation in 2. Ovoviviparous: Froglet birthed from oviducts;
tadpoles of frogs of the Rana tarahumarae group no extra-ova nutrition whether confined within
in western Mexico. Herpetologica 33:73-82. egg membranes or not; fertilization internal;
WEIGMANN,D. L. 1982. Relationships between development highly modified-Eleutherodac-
Production and Abundance of Ranid (Rana pip- tylus jasperi, Nectophrynoides tornieri, N. vi-
iens, R. catesbeiana, and R. clamitans) Tadpoles viparus only.
and the Effects of Tadpole Grazing on Algal Pro- 3. Paraviviparous: Froglet hatched from egg cap-
ductivity and Diversity and on Phosphorus, Nitro- sules at eventual site of deposition-in the stom-
gen and Organic Carbon. Ph.D. Dissertation, Mich- ach (Rheobatrachus), in the dorsal skin (Pipa
igan State University, Lansing. pipa), on the dorsal skin (some hemiphractine
WEYGOLDT,P. 1980. Complex brood care and re- hylids), or within a complete (Gastrotheca part)
productive behavior in captive poison-arrow frogs, or partial (Flectonotus) dorsal pouch; larva may
Dendrobates pumilio 0. Schmidt. Behav. Ecol. So- exhibit partial tadpole morphotype during de-
ciobiol. 7:329-332. velopment; female parent always involved.
WIENS, J. A. 1977. On competition and variable 4. Exoviviparous: Froglet not "birthed" from site
environments. Am. Sci. 65:590-597. of egg deposition as hatched embryo, 1) moves
WILBUR,H. M. 1972. Competition, predation, and of own accord into dorsal pouch with inguinal
the structure of the Ambystoma-Rana sylvatica port (Assa) or onto parent's dorsum (Leiopelma
community. Ecology 53:3-21. archeyi, L. hamiltoni, Sooglossus sechellensis)
. 1982. Competition between tadpoles of or 2) is swallowed into the vocal sac (Rhino-
Hyla femoralis and Hyla gratiosa in laboratory derma sp.); male parent usually involved. Be-
experiments. Ecology 63:278-282. cause 1) parental action involved, 2) a tadpole
WILBUR, H. M., P. J. MORIN, AND R. N. HARRIS. morphotype exists, and 3) a complex paternal-
1983. Salamander predation and the structure of larval nutritional relationship is involved (Goi-
experimental communities: Anuran responses. coechea et al., 1986), at least R. darwinii may
Ecology 64:1423-1429. have to be removed from this guild.
WILCZNSKA, B. 1981. The structure of the alimen- 5. Direct Developer: Froglet hatched from egg
tary canal and the dimensions of the mucosa in capsules of deposited eggs not intimately asso-
ontogenetic development of some Anura. Acta Biol. ciated with parent's body; reports of meroblasty
Cracoviensia Zool. 23:13-46. (de Villiers, 1929; Warren, 1922) warrant in-
WILTSHIRE, D. J., AND C. M. BULL. 1977. Potential vestigation; tail greatly reduced or enlarged and
competitive interactions between larvae of Pseu- vascularized; small gills present or not; spiracle
dophryne bibroni and P. semimarmorata (Anura: absent; oral structures of exotrophic tadpoles
Leptodactylidae). Australia J. Zool. 25:449-454. absent-Arthroleptis, Australopapuan micro-
WOODWARD, B. D. 1982a. Tadpole competition in hylids, Eleutherodactylus (most), Hylacto-
a desert anuran community. Oecologia 54:96-100. phryne, Myobatrachus, Oreophrynella, Platy-
.1982b. Tadpole interactions in the Chi- mantis, Sooglossus gardneri, Syrrhophus,
huahuan Desert at two experimental densities. Tomodactylus, and others.
Southw. Nat. 27:119-121. 6. Nidicolous: Egg or larva seldom intimately as-
19891
199 HERPEOLOGIAL MONOGRAPHS
HERPETOLOGICALMOOGRAPS 107
10

sociated with parent's body; eggs often small part functions, this group will surely be divided
and numerous; normal embryology results in further: total number of tooth rows, row ar-
free-living, non-feeding tadpole, continuum rangement (more on anterior labium, more on
from those having morphologicaltraits of exo- posterior labium, equal; one versus several rows
trophic tadpoles to those that have fins, oral broken medially adjacent to jaw sheaths), many
apparatus, pigment and/or spiracle variously variations in jaw sheath configurations and body
reduced to absent; tadpole remains in or near shape.
nest site until metamorphosis-Adenomera, 3a. Type 1: Belly flattened, snout rounded in lateral
Anodontohyla, Craspedoglossa, Eupsophus, view-Ascaphus, Bufo penangensis, Conraua,
Leiopelma hochstetteri, probably Mertenso- Hyla (many),Petropedetes,Scutiger(may have
phryne, Nectophrynoides malcolmi, Para- narrow anterior gap in marginal papillae typ-
cophyla, Pelophryne, Philoria, Phrynodon, ical of pelobatids), Trichobatrachus (Fig. 7D).
Platypelis, Plethodontohyla, Rana (few), 3b. Type 2: Anterior end of body markedly de-
Synapturanus, Vanzolinius, Zachaenus. pressed-Heleophryne, some peladryadinine
hylids.
EcomorphologicalGuilds of 4. Gastromyzophorous: Belly sucker present,
Exotrophic Anurans sucker cup papillate and/or with keratinized
Guild names are in boldface and noted by Arabic patches; complete marginal papillae or poorly
numerals;taxonomic listings are not always exhaus- developed posteriorly; fins low, dorsal fin often
tive. SectionI coverslotic forms,and section II covers terminates posterior to body; LTRF upwards
lentic forms. from 2/3; inhabit fastest, often turbulent water;
position maintenance by oral disc/belly sucker
Section I. Tadpoles inhabit flowing water continuous-Amolops, Atelopus (Figs. 7E, 8D).
systems-streams, rivers, creeks, seeps, 5. Psammonic: Long spiracular tube free from
rivulets;lotic is a generic term. depressed body; hypertrophied jaw serrations
A. Tadpolesinhabitflowingportionsof the system; lack basal sheath; labial teeth absent; "oral disc"
rheophilous(= rhyacophilous)is a generalterm; greatly reduced; eyes dorsal-Otophryne ro-
eyes small, non-bulging, and dorsal; oral disc busta (Pyburn, 1975; Wassersug and Pyburn,
usually large, pendant, and ventral;tail muscle 1987).
often massive;fins low, terminate at tail-body 6. Semiterrestrial: Semiterrestrial (Duellman and
junction or posteriorly,often in rounded tip. Trueb, 1986) seems more appropriate than sub-
1. Clasping:Marginalpapillae with anteriorgap; aerial (Wassersug and Heyer, 1983). May in-
tooth rows 5 (= 2/3) up to at least 8/8 but often volve two groups-Leptodactylus versus oth-
with more anterior rows than posterior rows; ers. Inhabit wet rock faces and rivulets, spend
eyes not as small as in following groups;inhabit time out of water; low tail fins on attenuate tail
slow water; position maintenance by oral disc often restricted to posterior third of tail; mouth
minor-Boophis (part),Natalobatrachus,Rana ventral with incomplete marginal papillae and
pustulosa and palmipes groups (Figs. 7A, 8A) 5 tooth rows; eyes often large and protrudent;
2. Adherent: Complete marginalpapillae numer- belly flattened; hindlegs often develop early-
ous, small,and closely packed, often in multiple Cycloramphus (part), Leptodactylus rugosus,
rows;LTRF (= labial tooth row formula) com- Nannophrys ceylonensis, Rana beddomei,
monly 2/3; inhabit faster water than Clasping Thoropa, Werneria (part) (Fig. 7G).
tadpoles; position maintenance via oral disc B. Tadpoles inhabit microhabitats of non-flowing
common to continuous; submarginal papillae portions (backwaters, burrows, eddies, pools) of
often large and in rows above A-1 and below lotic systems, habitats similar in many respects
last lower row; torrent tadpole is a collective to lentic systems; tadpoles resemble lentic forms.
term for this and next two guilds. 7. Benthic: See lentic-benthic tadpoles, II.12-
2a. Type 1: Belly flattened, body bluntly rounded Colostethusinguinalis, Hyla arenicolor,H. ca-
anteriorly-Hyla (many), Plectrohyla (all), daverina, Leptopelis sp. (Fig. 7H), Rana an-
Ptychohyla (part;Fig. 8G). golensis, R. fuscigula, Strongylopusgrayii (Fig.
2b. Type 2: Similar to above, but more slender; 7A).
tooth rows sometimes reduced to absent but 8. Fossorial: Vermiform burrowers with de-
never long;large areasof oraldisc withouttooth pressed body; eyes dorsal; low fins on long tail.
rows;jaw sheaths not robust-Hyla rufioculis, 8a. Type 1: LTRF 2/3, oral disc small; marginal
H. uranochroa (Figs. 7N, 8H), Mantidactylus papillae incomplete anteriorly; eyes small, not
femoralis, Ptychohyla (part), Taudactylus. bulging, and notedly parasagittal; often red be-
3. Suctorial: Complete marginal papillae small, cause of skin vascularity; stated that tadpoles
numerous,and usually in multiple rows;LTRF live among rocks of stream bottoms, but more
> 2/3-17/21; inhabit faster water than Ad- commonly buried within leaf mats in still-water
herent or Clasping tadpoles; position mainte- areas-Centrolene, Centrolenella (Fig. 70).
nance via oral disc continuous;submarginalpa- 8b. Type 2: Tooth rows reduced to absent; often
pillae often more numerous and not arranged found among interstices of small stones; mouth
as for adherent species; fleshy flap underlies structure suggests suctorial forms, but body-tail
hindleg anlagen in some (e.g., Ascaphus, He- configurations and behavior suggest placement
leophryne). Once more is known about mouth here-Leptobrachium (part).
 108 108 HERPETOLOGICAL MONOGRAPHS
HERPETOLOGICAL MONOGRAPHS [No.
[No. 3~~~~~~~~~~
3

9. Nektonic: Syn: pelagic (= open water). This and develop early; long tail with no fins; hides among
the analogous lentic guild (II.15) form mor- organic debris in tree holes-Indirana gundia
phological continuum-lotic-nektonic forms (Dubois, 1986).
have higher fins and more lateral eyes than llg. Type 7: Closely resembles most other phyllo-
lotic-benthic forms but never have fins as high medusine hylids (II.18) but with complete mar-
or eyes as lateral as most lentic-nektonic forms- ginal oral disc papillae-Agalychnis calcarifer,
Hyla miotympanum, Strongylopus fasciatus. Phyllomedusa fimbriata.
10. Neustonic: Enlarged oral disc directed dorsally 12. Benthic: Body depressed to globular-depressed;
(= umbelliform); eyes usually lateral, tail long oral disc ventral to anteroventral; eyes dorsal;
with low fins. Once development and function low fins do not extend beyond tail-body junc-
of the oral configurations are known, this guild tion; LTRF often 2/3.
may require division: presence/absence of ker- 12a. Type 1: Littoral, shallow, often weedy shoreline
atinized mouth parts, oral disc form and func- or shallow areas-Acris, Alytes, Bufo (most),
tion, and feeding ecology (see 11.16-) Colo- Colostethus (part), Dendrobates (part), Smilis-
stethus nubicola, Mantidactylus oviparus, ca (part).
Megophrys montana, Phyllomedusa guttata, 12b. Type 2: Profundal (= bottom without emergent
P. jandaia (Fig. 7B). vegetation or at least deeper); may move from
littoral to profundal areas for thermoregula-
Section II. Tadpoles inhabit non-flowing
tion-Bufo (part), Ptychadena, Rana (many).
systems-bromeliads, lakes, ponds, puddles,
tree holes; lentic is a generic term. 13. Carnivore: Configurations of body, tail, and eyes
usually similar to lentic-benthic forms (11.12),
11. Arboreal: As used by Lannoo et al. (1987) plus
although mouth often terminal.
all isolated pockets of water, elevated or not
13a. Type 1: Keratinized jaw sheaths present, often
(tree holes, terrestrial and arboreal bromeliads,
holes in logs, stumps, and bamboo, leaf axils, cuspate; tooth rows absent to >2/3; carnivory
in all individuals (Ceratophrys, Lechriodus,
fallen leaves, large snail shells, etc.); syn: phy-
totelmonous. Largely adopted from Lannoo et perhaps Hyla marmorata) or only some indi-
viduals at some sites in some years (Scaphiopus
al. (1987).
1 la. part); cannibalism may be included; prey may
Type 1: Elongate with attenuate body; tail: body be engulfed by one individual (Ceratophrys)
ratio >2; LTRF 1/1 or less; highly reduced gill
or rasped apart (alive or dead) by a group (Hyla
filters and filaments; musculoskeletal special-
izations for macrophagy; reduced pigment; in marmorata); jaw musculature often hypertro-
areas with low productivity-bromeliads and phied; body often widest anteriorly except in
leaf axils; all may be obligatory oophagous- hylids; most are active swimmers in the water
column; carnivorous/oophagous arboreal forms
Dendrobates (part), Hoplophryne, Osteopilus
placed in the arboreal guild because of adap-
brunneus, Philautus (part). tations for living in small bodies of water (Fig.
1 b. Type 2: Shorter with stout body; tail: body ratio 7Q).
<2; LTRF <2/2; reduced gill filters and fila-
13b. Type 2: LTRF 0/0, isolated jaw serrations with-
ments; reduced pigment; carnivorous and mac-
out a base; huge mouth without an oral disc;
rophagous-Anotheca spinosa, Hyla zeteki (Fig.
7P). body depressed; macrophagous-Lepidobatra-
llc. chus.
Type 3: Elongate with attenuate body; tail: body
ratio >1.7; LTRF >2/3; minimal reduction of 13c. Type 3: No keratinized mouth parts, soft mouth
internal oral features; no musculoskeletal spe- parts reduced and protrusible; inhabit water
cializations for macrophagy. Once functions of column, eyes lateral; macrophagous on small
tooth row configurations are known, this group crustaceans, etc.-Hymenochirus, Pipa carval-
hoi.
may be divided-those that increase the LTRF
above 2/3 by adding posterior rows and those 14. Macrophagous: LTRF 0/0 or 0/1; jaw sheaths
that add anterior rows-Hyla bromeliacia, H. present; terminal oral disc small to absent, with
dendroscarta, Mantidactylus (part), Phyllo- or without a few marginal papillae.
dytes (Fig. 71). 14a. Type 1: Benthic, body depressed; eyes dorsal-
lid. Type 4: Short with stout body; tail:body ratio Occidozyga (Fig. 7L), probably Cardioglossa.
<2; LTRF >2/3, some increase in anterior rows; 14b. Type 2: Nektonic, body cylindrical to depressed
gill filters and filaments normal to greater than in cross-section; eyes lateral; tail flagellum and
normal; darkly pigmented-Acanthixalus, cryptic coloration common; float motionless and
Nyctixalus, Theloderma. horizontally in water column, often among
lie. Type 5: Raised fleshy ring of unknown function vegetation-Afrixalus, Hyla leucophyllata, mi-
on head encompasses eyes and nares; otherwise crocephala, and parviceps groups (Figs. 7K, 8E).
resemble lentic-benthic forms (II.12)-Mer- 15. Nektonic: Body compressed; tail flagellum
tensophryne (morphologically belongs here, but common; fins commonly high and may extend
probably nidicolous, see above), Stephopaedes anteriorly to level of bulging, lateral eyes; tail
(Fig. 7R). muscle not massive; feed by rasping surfaces
llf. Type 6: Inhabit humid microhabitat on and with anteroventral mouth.
under bark of trees, feed on organic detritus 15a. Type 1: LTRF commonly 2/3-Hyla (many),
with jaw sheaths and labial teeth; large hindlegs Ololygon (most), Smilisca (part) (Fig. 7C).
1989] HERPETOLOGICALMONOGRAPHS 109
109

15b. Type 2: LTRF greater than 2/3-Phrynohyas, marginal papillae; eyes lateral; tail flagellum
Trachycephalus. present to lesser or greater degree; often sus-
16. Neustonic: Similar to II.12 but oral disc in- pended in midwater, often head-up; ventral fin
verted (= umbelliform) like lotic-neustonic may be considerably higher than dorsal fin,
(I.B.10)-Microhyla achatina (Figs. 7F, 8B). which stops near tail-body junction; schooling
17. Suspension Feeder: Keratinized mouth parts reported-Agalychnis (most), Kassina, Pachy-
absent; tail flagellum common; inhabit the water medusa, Phyllomedusa (most).
column; body usually strongly depressed, eyes Comments.-Behavioral variations sometimes
lateral; low fins similar in size or ventral larger.
17a. Type 1: Often lightly pigmented without bold suggest moving a taxon to another guild. For example,
the normally benthic tadpoles of Bufo and Rana
patterns; barbels present, no other prominent sometimes swim within the water column, whereas
oral features around terminal, horizontal, slit-
like mouth; includes only Orton (1953) Type I many nektonic species of Hyla spend considerable
time on the bottom when in shallow water. Tadpoles
tadpoles-Rhinophrynus, Silurana, Xenopus. of Kassina may be carnivorous at times (van Dijk,
17b. Type 2: Almost nonpigmented and transparent
other than eyes and gut to heavily pigmented; 1972). We suggest that such variations be noted but
that each taxon be placed in the one guild to which
oral disc derivatives variously shaped but sel-
it most commonly conforms.
dom prominent around terminal mouth; bar-
Other guilds and types will surely be recognized.
bels absent; often small; sometimes locomote by
The large genus Hyperolius with similar mouth parts
current from midventral spiracle-Chiasmo-
could not be categorized because of insufficient data.
cleis, Ctenophryne, Gastrophryne, Nelsono-
Tadpoles of Scarthyla ostinodactyla and the Ololy-
phryne, Hypopachus, Microhyla (part), Phry-
nomerus (Fig. 7J). gon rostrata group have unique oral configurations
that may warrant a separate lentic group.
18. Suspension-Rasper: Keratinized jaw sheaths and
labial teeth (usually 2/3) with anterior gap in