morphological diversity and evolution of egg and clutch structure in amphibians

32 Pages·2007·2.3 MB·English

HerpetologicalMonographs,21,2007,1–32
E2007byTheHerpetologists’League,Inc.
MORPHOLOGICAL DIVERSITY AND EVOLUTION OF EGG AND
CLUTCH STRUCTURE IN AMPHIBIANS
RONALD ALTIG1,3 AND ROY W. MCDIARMID2
1DepartmentofBiologicalSciences,MississippiStateUniversity,MississippiState,MS39762-5759,USA
2USGSPatuxentWildlifeResearchCenter,NationalMuseumofNaturalHistory,Washington,DC20560-0111,USA
ABSTRACT: Thefirstpartofthissynthesissummarizesthemorphologyofthejellylayerssurroundingan
amphibianovum.Weproposeastandardterminologyanddiscusstheevolutionofjellylayers.Thesecond
part reviews the morphological diversity and arrangement of deposited eggs—the ovipositional mode; we
recognize5morphologicalclassesincluding14modes.Wediscusssomeoftheoviductal,ovipositional,and
postovipositionaleventsthatcontributetothesemorphologies.Wehaveincorporateddatafromtaxafrom
throughouttheworldbutrecognizethatothertypeswillbediscoveredthatmaymodifyunderstandingof
thesemodes.Finally,wediscusstheevolutionarycontextofthediversityofclutchstructureandpresentafirst
estimateofitsevolution.
Keywords: Amphibia;Clutch;Eggs;Evolution;Jellylayers;Oviposition;Ovulation.
AMPHIBIANSsurelyhavethemostvariedand diversity of egg jellies and ovipositional
least known life histories of all terrestrial modes.
vertebrates.Typically,terrestrialadultsreturn Field identification of amphibian eggs is
to aquatic sites where courtship, egg de- difficult at best, in part because authors of
position, and fertilization occur. Eggs hatch published descriptions of eggs have used
into larvae that are mobile feeding forms. inconsistent and inexact terminology and
After a period of growth and development, misinterpreted certain structures. For exam-
theselarvaeortadpolesusuallyundergosome ple,thevitellinemembranesensulatoactually
form of metamorphosis and move onto land takes on three successive states with different
where they grow into reproductive adults. biochemical and physical properties and
Variations on this pattern most often repre- functions: coelomic, vitelline, and fertilization
sentevolutionarychangesindevelopmentalor membranes (Carotenuto, 2001; Gerton and
life history traits (e.g., paedomorphy and Hedrick, 1986; Takamune et al., 1987), but
pedotypy, egg placement, tadpole develop- most authors fail to distinguish among them.
ment) expressed during the ontogeny of an Inaddition,differentobservationaltechniques
individual prior to its becoming an adult. can lead to different results, so that research-
McDiarmid and Altig (1999) summarized the ers often disagree on traits as seemingly easy
biology of tadpoles, and we know even more to observe as the number of jelly layers
about the adult stage (e.g., Duellman and around an ovum.
Trueb,1986).Althoughtheresearcheffortsof While writing a key to the eggs of North
embryologists and cell biologists have pro- American amphibians, we realized that egg
identifications based on properly described
vided extensive information on ovum pro-
ovipositional modes were likely to be more
duction and early development, even if in-
accurate than those based on the more labile
volving a relative few taxa, nearly every facet
and poorly documented features of individual
ofthefieldbiologyofamphibianeggsispoorly
eggs.Clutchmorphologyiseasiertosee,more
documented. In this paper we focus on egg
readily definable, and usually less variable
structure, particularly the jelly layers, and
withinspecies.Bycombiningthesedefinitions
patterns of egg deposition. We also consider
with careful field observations (e.g., Pombal
the characteristics of the physical and bi-
and Haddad, 2005) made under well-docu-
ological environments in which the immobile
mentedenvironmentalconditions(e.g.,Wright
eggs are laid that likely account for the
and Wright, 1924), a good observer should be
able to identify the eggs of North American
3CORRESPONDENCE:e-mail,[email protected] amphibian taxa at least to genus with reason-
1
2 HERPETOLOGICALMONOGRAPHS [No.21
able confidence. We suspect that the same capsular chamber in salamanders and some
couldbedonewithotherfaunasaswell. frogs) events. Of greater import in essentially
Our initial goal in this review is to develop all cases is the lack of data or the lack of
a generalized framework for observed di- integration of those data across disparate
versity of amphibian eggs and ovipositional fields of research. Diagrams of eggs are
modes that will contribute to a broad un- schematic and usually drawn in lateral view
derstanding of the evolution of amphibian taken at the equator with the animal pole
reproduction.Webelievethattheoviposition- uppermost. Concentric circles drawn around
almodeisafactorthatcanprofitablyaugment the ovum (e.g., continuous, dashed, or stip-
our notions of breeding or reproductive pled lines or zones; Hoyt, 1960; McDiarmid
modes (e.g., Crump, 1974; Duellman and and Worthington, 1970) represent visually
Trueb,1986;HaddadandPrado,2005).Inthe perceived jelly layers. Even so, no adequate
process, we present a standardized terminol- description of the arrangement of jelly layers
ogy for egg jellies and clutch structures and around freshly oviposited eggs based on
discuss the evolution of these features. Al- observations using standardized techniques
though data acquired by cell biologists, de- (i.e., proper illumination, dissection, section-
velopmental geneticists, and particularly his- ing, staining) has been published.
tochemists will enhance the eventual We summarize the literature on amphibian
understanding of the biology of ova and egg egg morphology, especially that addressing
jellies, we do not delve into this extensive theproductionandevolutionofthejellylayers
literature. surrounding the amphibian ovum. We follow
Also, oneshouldbeconstantly aware ofthe withasynthetictreatmentofthemorphology,
integration of the data discussed herein with diversity, and evolution of the egg clutch. We
the concept of breeding mode that is being also propose a standardized terminology for
constantlyrevisedandexpanded(e.g.,Crump, describing eggs and egg clutches that we
1974; Duellman and Trueb, 1986). What we believe facilitates communication and ad-
present is a subset of the larger concept, vances the study of this important but
although this integration is left for a future neglected stage of the amphibian life cycle.
synthesis when egg biology is better under-
stood.
EGG MORPHOLOGYANDITS VARIATIONS
The morphological surveys of egg structure
Definitions
by Salthe (1963) and Salthe and Duellman
(1973) suggested that intriguing morphologi- We define ‘egg’ as an ovum (5 gamete
cal and ecological patterns exist within am- through gastrulation, stages 1–12; Gosner,
phibians. Variations in the surface morpholo- 1960; other versions in Duellman and Trueb,
gies of ovum coverings and the nature of the 1986 and McDiarmid and Altig, 1999, Chap-
arrangement and attachment of eggs in the ter 2) and its vitelline membrane of ovarian
environment as reported in other organisms origin surrounded by one to several oviduc-
(e.g., Mooi, 1990; Stiassny and Mezey, 1993; tally-producedjellylayer(s).Theinternallayer
Strathmann and Chaffee, 1984) surelyexist in is deposited first by an anterior region of the
amphibians,butcertainfactorsobfuscatetheir oviduct, and as the egg moves down the
interpretation in this group. For example, the oviductmorelayersaredepositedsequentially
number of jelly layers surrounding an ovum by more posterior oviductal regions.
and the number of secretory areas in an Various terms have been used to refer to
oviduct may vary independently (Greven, the individual jelly layers, and they and the
2003). In part, these variations can be entirecomplementoflayersasaunitgenerally
explained by oviductal (e.g., changes in lack discrete definitions; some of the terms
secretory regions in the oviduct; preoviposi- usedindifferentresearchfieldsareinaccurate
tional changes in jelly; D. M. Hardy and (e.g., capsule, envelope). We suggest that the
Hedrick, 1992), postovulatory (e.g., dissolu- entire assemblage of oviductal materials de-
tionofsomejellylayers)andpostovipositional posited around the ovum be referred to as
(e.g.,layersmeldingtogether;formationofthe ‘jelly’or‘jellylayers’regardlessofvariationsin
2007] HERPETOLOGICALMONOGRAPHS 3
their apparent physical structure (e.g., tough- revealed 5–6 layers, whereas dissections
ness,density,thickness,watercontent).Layer, (Salthe, 1963) and other techniques (Shaver,
zone,andmembranerefertounitswithinthis 1966; Shivers and James, 1970) indicated 3–5
assemblage. A ‘layer’ (often referred to in the layers. Descriptions of the eggs of Dicampto-
cell biology literature by the letter ‘J’ with don sp., probably based on visual examina-
a numerical subscript to denote position) is tion of intact eggs (Stebbins, 2003), indicated
amorphologicallydiscrete,easilyrecognizable 2 jelly layers, whereas visual and tactile
region of jelly surrounding the ovum; it has detection of density differences noted dur-
a discernible thickness along a radius and ing dissection (Nussbaum, 1969) showed 5
typically has a uniform and visually apparent layers.
optical density. Layers are numbered from More structure certainly exists in the jelly
interior to exterior (e.g., Daniel, 1937) in the layers than can be seen with simple micro-
order in which they are formed. The term scopy. Other observational techniques (e.g.,
‘layer’ does not encompass any specific differential interference contrast [DIC],
function or physical characteristic. phase-contrast and fluorescence microscopy
Daniel (1937) and others pointed out that and laser illumination) do not produce better
some layers have discernible but less discrete images. Laser confocal imaging and some
subdivisions. We designate these subdivisions techniques of electron microscopy provide
as ‘zones’ to suggest areas that are less easily better results but are costly and specimen
visualized than layers (see Steinke and Ben- preparation is laborious (e.g., Bonnell and
son, 1970), but we do not imply that all zones Chandler, 1998; Larabell and Chandler,
of nonstained eggs are visible. 2005). More refined zones, which can be
Finally, we use ‘membrane’ as a functional detected with histological and histochemical
termtodescribewhatappearstobeadelimit- techniques, likely will reveal patterns of
ing boundary between layers, exclusive of the biological, ecological, or phylogenetic interest
vitelline membrane. Data are currently in- (e.g.,HedrickandKatagiri,1988;Smithetal.,
sufficient to determine if membranes are 2002). Biochemical analyses show that egg
actual structures or if they are merely an jellies consist of ‘‘a fibrous glycoprotein
optical representation of the plane along superstructure that acts as a scaffold to which
which two layers of different densities abut. globular glycoproteins are bound’’ (Fig. 5E;
In some instances two layers apparently abut Bonnell and Chandler, 1998); which of these
without a visible membrane, thereby suggest- groups of molecules is structural and which is
ing that the densities of the two layers must biologically active is not known, but both are
reach some threshold before a visual mani- expected to have species-specific qualities
festation of the transition appears. If the (Maes et al., 1995). The three jelly layers of
‘membranes’ between layers are discrete theeggsofXenopuslaevisarecomposedofat
structures, we do not know if they are least nine glycoproteins (Yurewicz et al.,
independent of adjacent layers and produced 1975).
by a specific oviductal secretory region, the Thesekindsofdata,whileinteresting,areof
result of some postovipositional reaction of littleusetofieldbiologistswhousuallyrelyon
the jelly layers, or an outer or inner boundary visual impressions of structure viewed under
of a particular layer. incident,whitelight.Somedatasuggestthatit
maynotbepossibletoformulateadescriptive
Structure
model useful to all researchers. Accordingly,
Clear jelly layers certainly are not as field biologists, histochemists, and biologists
structurally uniform (e.g., Carroll et al., in other fields may find it expedient to use
1991) throughout their thickness as they their own sets of terms to describe observed
appear.TheeggofRanapipiens,forexample, variations in jelly traits.
has two, clear layers of jelly (Wright and In summary, jelly occurs in layers and
Wright, 1949:35). Studies (Steinke and Ben- a layer may be subdivided into less easily
son,1970)ofthetaxonofthesamenamewith discernible zones. Visually perceived layers
immunological and histochemical techniques typically would be the morphological trait
4 HERPETOLOGICALMONOGRAPHS [No.21
FIG.1.—Schematicdrawingsofanamphibianeggwithitscomponentparts,theoviductalregionsa–cthatproduce
thejellylayers,andthreehypotheticalpathsforaddingajellylayer.(A)Anovumwithtwomembranes(seetext)andtwo
jellylayersproducedinsequencebyoviductalsectionsaandbarrangedanteriorlytoposteriorly.(B)Ahypotheticalcase
where a new layer is added directly external to the vitelline membrane; new numbering of the jelly layers and the
arrangementoftheoviductalregionsareshown.(C)Acasewhereanewlayerisaddedexternally.(D)Acasewhere
anewlayerisaddedwithintheoriginalfirstlayer.Abbreviations:1–45jellylayersnumberedfromovumtosurface,ant
5anterior,IM5innermembrane,O5ovum,OM5outermembrane,post5posterior,VM5vitellinemembrane,
andVS5vitellinespace.
most useful for identification. For example, changeandneedformechanicalsupport,other
the basic egg structure consists of an ovum factors (e.g., homospecific sperm attraction
withavitellinemembranesurroundedbyjelly [Al-Anzi and Chandler, 1998], heterospecific
layers 1 and 2 (Fig. 1A, E) of oviductally- spermavoidance,heatconservation,andpred-
produced materials. Inner and outer mem- ator defense) likely have played roles in the
branes are labeled, although oviductal secre- evolution of the number, thickness, and
toryzonesfortheirproductionarenot,andno physical characteristics of the layers. To
zones are shown. appreciate layer homologies one must ulti-
mately understand the morphology of the
secretory regions of the oviduct and presum-
EVOLUTIONOF EGG JELLY ablyhavesomenotionofevolutionaryrelation-
The evolution of additional layers of jelly ships among taxa. Presumably, selection for
does not appear to result only from the differences in the number and characteristics
subdivision of ancestral layers. Thus, egg of the jelly layer has influenced oviductal
diameter often is greater in specieswith more morphology rather than the reverse, although
layers rather than in those with fewer layers, other factors that seemingly are not related to
even though the thickness of each layer, or at egg survival (e.g., Anderson et al., 2006) must
leastofsome,oftendecreasesasthenumberof be considered. Evidence suggests that envi-
layersincreases.Althoughthethicknessofjelly ronmental demands and phylogenetic con-
layersmustbemediated bythegeneraltrade- straints have influenced the evolution of egg
offs between the requirements for gas ex- morphology,althoughlittleisknownaboutthe
2007] HERPETOLOGICALMONOGRAPHS 5
selective factors involved, and details and We know of no proposed scheme of jelly
patterns of the morphology in most taxa are layer homologies for anurans, and phyloge-
still lacking (Salthe, 1963; Greven, 2003). netic patterns and possible selective factors
Hypothetical scenarios diagramed (Fig. 1B– contributing to their evolution are often
D, F–H; also Greven, 2002:fig. 11) illustrate discordant. For example, among North
potential modifications in oviductal secretory American frogs (Moore, 1940), many cool-
regions that would produce given changes in water breeders deposit clumps of eggs with
jelly layers; the review by Wake and Dickie either 2 or 4 jelly layers, whereas warm-water
(1998) provides informative discussions of breeders lay either clumps or films of eggs
oviductal anatomy relative to breeding mode. with 2 layers. ‘Wood frogs’ (West Coast
If we assume for heuristic purposes that endemics plus Rana sylvatica) produce
combinations of the following alternatives did clumps of eggs with 2–3 jelly layers in cool
not occur, then a third jelly layer might be water. Members of the Rana catesbeiana
added(1)betweenthevitellinemembraneand group breed in warm weather and deposit
thefirstjellylayer;Fig. 1B,F),(2)betweenthe eggs with 1–2 layers as clumps or surface
two original jelly layers), (3) external to the films. Members of the Rana pipiens complex
secondlayeroftheoriginalegg;Fig. 1C,G),or lay eggs with 2 jelly layers in clumps and
within an existing layer; Fig. 1D, H). If usually in cool water. Considering the four
a posterior (i.e., new layer added externally) optionsforlayerhomologies(Fig. 1B–D),one
oranterior(i.e.,newlayerinternally)regionof canaskiftheoutertwoortheinnertwolayers
the oviduct were involved and a phylogenetic of the wood frogs are homologous to the
scheme were known, one potentially could two layers of the other groups of North
track the products and thus hypothesize American ranids and which of the layers
homologies. Identification of other types of might reflect responses to temperature or
additions may require histochemical tech- ovipositional mode and which to phylogenetic
niques, andexamination ofchangesinoviduc- constraint.
tal regions. Which evolutionary pathway is A complete histochemical data set for an
most likely can be debated and will depend amphibian egg might agree with Salthe’s
largelyontheeasewithwhichhomologiescan (1963) morphological data. One still would
be assigned. Layer thickness also is likely not be sure of layer homologies, but his
influencedeitherbythelengthoftheoviductal hypotheses based on position and construc-
region, passage rate of the ovum, rate of jelly tion would have stronger support. Mapping
production, or differences in hygroscopic histochemical data onto appropriate clado-
qualities of the jelly. Although these parame- gramswouldlikelyprovidesomeinsightabout
ters are probably important, we have ignored the evolution of jelly layers, and information
thembecauseofthelackofdata. on the ecological functions of egg jellies
Salthe (1963) presented the only sugges- should reveal correlations useful to interpret-
tions of homologies of jelly layers in caudates. ing their evolution. Salthe (1963) suggested
His interpretations, based on 15 caudate that layer losses in plethodontids that lay
genera in 8 families suggest that (1) 8 jelly terrestrial eggs involve changes of internal
layers (as in Hynobius lichenatus) is the layers and that the tough outer layer is
primitive condition; (2) changes in the num- retained for protection. Evaluations of the
beroflayersoccuronlythrough(a)lossofthe hypotheses about jelly layer homologies will
most external layers (e.g., ambystomatids) or have to wait additional data.
(b) loss of more internal layers (i.e., elimina-
tion of layers somewhere within the series;
ANCILLARY SUBJECTSON EGGS
particularly plethodontids); and (3) eggs with
Characteristics of Jelly Layers
3layers ofundetermined homologies(as in of
Cryptobranchus)isthesimplestcondition.All Anyone who has handled eggs and espe-
of these ideas are based on the assumption ciallythosewhohavemanuallydejelliedthem
that layers in discernible positions and of are familiar with features such as elasticity,
similar construction are homologs. stickiness, toughness, turgidity, and wateri-
6 HERPETOLOGICALMONOGRAPHS [No.21
ness. Aquatic eggs usually are spherical when the jelly immediately after oviposition rather
submerged but sag when placed on a surface thanstructuraldifferencesperse.Thejellythat
inair.Thejelliesofmostterrestrialformsthat forms the flange surely surrounds the entire
do not lay suspended eggs have jelly with ovum,incontrasttobeingdepositedasaband,
sufficient tensile strength and turgidity to and is likely very watery. The pressure of the
remainsphericalinair.Theouterjellyofeggs water surrounding the egg as it sinks partway
of Ambystoma opacum and other terrestrial through the water surface likely pushes this
salamanders and frogs with direct develop- flimsyjellytotheair-waterinterfaceandforms
ment is tough relative to that of aquatic eggs. the flange. Different flange positions (e.g.,
Tougher membranes and increased turgidity equatorialinKaloulaspp.andnearthevegetal
of the enclosed fluids help to maintain the pole in Paradoxophyla palmata) may reflect
spherical shape of these large eggs in air and interspecific differences in the density of the
thereby allow proper development, oxygena- egg-jelly complex or in the hygroscopic qual-
tion, and protection from trampling by an itiesofthejelly.Li(1934)notedthattherimin
attendantparent.Theseterrestrialeggscanbe Kaloula borealis did not appear until about
grasped with minimal distortion and will 1 min after oviposition, and Taylor (1922)
bounce if dropped. If the outer layer is describedthegelatinousflangeinKalophrynus
removed, the remaining jellies spread out, pleurostigma as gradually widening to about
and the ovum usually ruptures. In some 6 mmdiameterafterextrusion.
aquatic eggs the most external visible layer is
Inclusions in Jelly
surrounded by a transparent, watery gel that
appears tolack adefiningexteriorsurface but The structure, origin, and function of
is crucial for flotation (see below). various crystalline inclusions in the jelly of
Asymmetries of jellies caused by tensile salamanders (e.g., Ambystoma maculatum,
differences,suchasthedroopingofeggjellies Salthe,1963;Ruthetal.,1993;Sirenlacertina
of terrestrial plethodontid eggs or the pentag- and Hynobius lichenatus, Salthe, 1963) need
onal appearance of jellies of eggs tightly further examination. For example, the egg
spaced in a film, are common. The observa- jellyofAmbystomamaculatumcanbeclearor
tion by Wright and Wright (1949) that the opaquewhite(L.M.HardyandLucas,1991).
innerjellylayeroftheeggsofRanaclamitans, White jelly gets its color from glycoprotein
which is not under any tensile forces, may be crystals that are produced with the jelly in
elliptical or pear-shaped needs verification. cells in the oviductal wall. Populations with
We assume that the conical ova that Wright clear-jellied,white-jellied,orbothtypesofegg
and Wright (1949) observed in Bufo alvarius masseshavebeenreportedfromnorthwestern
resulted from tension on the jelly string. Louisiana and adjacent areas of Texas and
There are cases of egg jelly asymmetries in Arkansas. The function of the crystals is
the absence of tension. The outer jellies of unknown, but they may reflect light and
some salamandrid eggs (Notophthalmus vir- thereby offer some protection to the de-
idescens; Bishop, 1943) are oval and attached veloping embryos or concealment from pred-
individually or in small groups to plants. Parts ators(L.M.Hardy,personalcommunication).
oftheouterjellylayersoftheeggsofsomeOld Similarly, the jelly of the eggs of the frog
World microhylids (e.g., Kaloula rugifera and Mantidactylus depressiceps (Mantellidae) is
K. macroptica, Fig. 2A, B, and Liu, 1950; milky white (RA, personal observation) al-
Kaloula borealis in Li, 1934; Kalophrynus though the cause of the color is unknown.
pleurostigma in Taylor, 1922; and Paradoxo- Whatappearasstriae,furrows,orcorrugations
phyla palmatain Glaw andVences,1992; RA, intheouterlayerofjellyofsomesalamanders
personal observation) are asymmetrical and (e.g., Hynobius lichenatus, Salthe, 1963) and
form a flange or rim which allows the eggs to frog egg masses (e.g., Cochranella pulverata
float. Asymmetry in jelly layers seems at odds [RWM, personal observation]) also deserve
with their mode of formation in the oviduct, attention.
but the asymmetry results from differences in Some ambystomatid eggs, notably those of
hygroscopic properties of specific portions of Ambystoma maculatum and A. gracile with
2007] HERPETOLOGICALMONOGRAPHS 7
FIG.2.—(A–C)Microhylid,(D–H)mantellid,and(I)arthroleptidfrogeggs.Asymmetricaleggjelliesof(A)Kaloula
rugifera(modifiedfromLiu,1950:fig.58)and(B)K.macroptica(modifiedfromLiu,1950:fig.60)causedbydifferential
hydration of specific parts of the jelly. (C) Part of a coherent film of Gastrophryne carolinensis eggs soon after
ovipositionshowingtheupperhemisphereofjellyprojectingabovethewatersurface(15downwardflexedmeniscusat
jelly margin, 2 5 glint on surface of ovum or vitelline membrane, and 3 5 trapezoidal reflection on curved, upper
hemisphere of outer egg jelly). Laminar array of a clutch of Guibemantis depressiceps eggs on a leaf (D) prior to
hydration(OV5ova;EJ5emptyjellies),(E)thesameclutchshowingalargeincreaseinvolumeafterhydration,and
(F)thesameclutchcutlongitudinallyafterhydration,stainedwithToluidineBlue,andviewedwithtransmittedlight(L
5 cut leaf; OS 5 ovumless stalk, OV 5 ova, and S 5 limits of one stalk with an ovum; stalks slightly darkened
electronically for better visibility). (G) Recently laid clutch of ‘Mantidactylus’ sp. on a leaf showing structure when
hydrationisnormallylessextensive,and(H)olderclutchof‘Mantidactylus’sp.attachedtothebarkofatreewithclear
jellyandalargeincreaseinjellywithhydration(embryosthatappeardeepinthejellyareactuallyonthesidesofthe
jelly).(I)TerrestrialclumpofArthroleptisschubotzifoundinleaflitter(photobyR.C.Drewes).
8 HERPETOLOGICALMONOGRAPHS [No.21
exceptionally dense jellies, have a symbiotic but possibly needs further study (Bragg,
greenalga(Oophilaamblystomatis)withinthe 1964).
innerjellylayersoftheegg(e.g.,Bachmannet
Hatching
al., 1986; Hammen, 1962).
Hatching occurs when a developmental
threshold is reached in concert with various
Capsular Chamber
biotic and environmental cues. Hypoxia (e.g.,
In salamanders and a few primitive frogs Petranka et al., 1982) likely is the proximal
(Salthe, 1963:165), the jelly layer abutting the trigger, but factors such as pathogens and
vitelline membrane dissolves soon after ovi- predators (e.g., Chivers et al., 2001; Touchon
position to form the capsular chamber; et al., 2006; Warkentin, 1995) or low pH
remnant debris, termed ‘white plac’, puddles (Dunson and Connell, 1982) can modify
at the bottom of the capsular chamber timing. Thumm and Mahony (2002) demon-
(Daniel, 1937). Eggs with a capsular chamber strated that the stage and degree of de-
are less confined than those without such velopment at hatching in Pseudophryne aus-
a chamber so that the ova lie slightly below tralis (Myobatrachidae) is variable. Hatching
centerintheremainingjelly;iftheeggmassis mechanisms seemingly are universal (Noble,
inverted,theovaimmediatelyturnsothatthe 1926;DuellmanandTrueb,1986)andfallinto
animal pole is uppermost. In eggs without two groups. In most amphibians enzymes
acapsularchambertheovaareconstrainedby from the frontal (5 hatching) glands on the
the jelly layers and take several minutes to head andsnout causeat leastpartial chemical
right themselves (Salthe, 1963). Embryos degradation of the jelly (e.g., Carroll and
usually exit the vitelline membrane long Hedrick, 1974; Urch and Hedrick, 1981),
before they hatch from the jelly (Salthe, which may often be augmented by simple
1963), and one can often see the vitelline writhing and pushing by the embryo to break
membrane crumpled up at the bottom of the throughthelayers(alsoBragg,1940;Gollman
inner jelly layer in advanced embryos. and Gollman, 1993; Lutz, 1944) or by rain
(Noble, 1926). In some direct developing
Functions of Egg Jelly
frogs an egg tooth on the snout or other part
Suggested functions of egg jellies (Greven, of the body mechanically ruptures the mem-
2002, 2003) include mechanical support for branes and jelly layers to allow hatching (e.g.,
theovum,attachmentofeggstoeachotheror J. D. Hardy, 1984; Noble, 1926).
a structure in the environment, enhancement
or prevention of entry by conspecific and Ovum Pigmentation
heterospecific sperm respectively (e.g., Bar- Even though the pigments in most ova are
bieri and Del Pino, 1970), prevention of melanic and not contained in cellular orga-
polyspermy, sperm capacitation, differential nelles, data on this pigmentation of maternal
protection from water molds Saprolegnia and origin are confusing. In large part this is
Achlys (Gomez-Mestre et al., 2006), pro- because there are no standards for describing
tection from contaminants (Marquis et al., color, intensity (e.g., dark, diffuse, pale),
2006), and protection from predators, patho- pattern,location,orthegradation fromdorsal
gens, and environmental stressors such as darkness to ventral paleness. In addition,
temperatureandUVlight(HunterandVogel, apparent changes in pigmentation with de-
1986; Itoh et al., 2002; McLaughlin and velopment or after preservation have rarely
Humphries, 1978; Ward and Sexton, 1981). been documented. Potential influences on
Incubation in birds helps to inhibit growth of ovum coloration include age of female, stage
bacteria and fungi on egg shells (Cook et al., of oogenesis, changes during development
2005); jelly layers may be a prepackaged way (i.e., when and how maternally-provided
toprotectovafromtheseperilsaftertheyexit pigment is supplanted during oogenesis by
from the relatively sterile oviduct. The in- that from the developing embryo), and egg-
fluenceofjellyonlightrefractionisthoughtto laying site (i.e., geographical and elevational
be insignificant (Cornman and Grier, 1941) variation). Likewise, color and pattern and
2007] HERPETOLOGICALMONOGRAPHS 9
their development in hatchlings (e.g., Altig, eggs deposited on the tops or undersides of
1972)thatareusefulforspeciesidentifications leaves) or provide infrared camouflage
(e.g.,hylidhatchlingscommonlyarestrikingly (Schwalm et al., 1977; also see Saito, 2001).
patterned, ranids usually unicolored) are Falchuk et al. (2002) have shown that
poorly documented. biliverdin IXa functions as a cytoplasmic de-
Nevertheless, some general correlates of terminant that is essential to normal embryo-
ovum pigmentation are apparent (Salthe, genesis in early stages of Xenopus laevis. The
1963). Eggs laid in open, exposed areas, destructionofbiliverdinviasomewavelengths
regardless of the specific site, taxon, or of UV light may provide a mechanism for
ovipositional mode, usually have melanic understanding their detrimental effects (e.g.,
pigment at the animal pole (e.g., Ambystoma, Blaustein et al., 1994). The omission of
Bufo, Hyla, and Rana of North America). As melanicpigmentinovasurelyconservessome
the number of embryonic cells increases and reproductive energy, and embryos developing
the cells move during gastrulation, the fertil- from nonpigmented eggs sometimes do not
ized ova become more uniformly pigmented develop pigment until well after hatching.
and paler because the ovum pigment is
dispersed among more cells until embryonic
pigment appears. Eggs laid in secluded sites CLUTCH MORPHOLOGY
(e.g., Amphiuma in burrows near lentic sites,
Definitions
Cryptobranchus, Dicamptodon, numerous
plethodontids, and Ascaphus hidden among ‘Clutch’ describes the total number of eggs
rocks in streams, some dendrobatid and deposited per ovulation event independent of
microhylid frogs among forest debris and the number or presence of a male(s), re-
phytotelmata, some centrolenids on leaves) productive or ovipositional mode, oviposition-
tendtobepaletononpigmentedregardlessof al behavior of parents, or number of ‘groups’
ovipositional mode. Biologists working in (5 aggregate of eggs produced in a single
temperate areas often assume that amphibian ovipositional bout; this term is useful when
eggs are pigmented because those the taxa one does not know the taxon or ovipositional
most commonly encountered are pigmented, mode or number of bouts [i.e., single egg
but in fact, many taxa in tropical regions have laying event]) that occurred. The number of
nonpigmented ova. ova ovulated sometimes exceeds the number
Several functions of egg pigmentation and of eggs oviposited. For example, if a female
adult behaviors associated with that trait have ovipositsherovulatedcomplementinmultiple
been proposed. Communal oviposition of bouts, the total number of eggs deposited
masses of dark eggs apparently enhances makes up the clutch (i.e., a group in this case
absorption of heat thereby increasing de- is not the clutch). A pair of Smilisca phaeota
velopmental rate in early breeding species of may deposit eggs among one to several
some North American ranids (Hassinger, puddles in a single night (RWM, personal
1970). Pigmentation in eggs can also provide observation); the eggs in each puddle com-
protection from deleterious effects of specific prise a group, are produced during a single
wavelengths of light or heat (Barrio, 1965; bout, and are only part of the clutch. Female
Jones, 1967). Biliverdin and lutein can pro- Ambystoma tigrinum may partition their
duce greenish to bluish ova in some phyllo- clutches into multiple discreet groups which
medusines (Pyburn, 1963; Marinetti and may be sired by one or more males; distances
Bagnara, 1983), some hyperoliids (Wager, between these groups in a single pond may
1965), some rhacophorids (Liu, 1950), and exceed 40 m (Gopurenko et al., 2006). Thus,
certain centrolenids (RWM, personal obser- the organization of the total clutch, and not
vation).Thesepigmentsarehousedintheyolk the number or manner in which eggs emerge
platelets (Barrio, 1965) and are from a differ- fromthefemaleinasinglebout,isimportant.
ent source than the melanic pigments. The If a female ovulates and oviposits more than
greenish color can enhance concealment of once a year or in multiple years, she has
frog eggs from some predators (e.g., greenish produced multiple clutches.
10 HERPETOLOGICALMONOGRAPHS [No.21
‘Ovipositional mode’ describes the mor- ent ages, and inadequate observational tech-
phology of an aggregation of deposited eggs niques. The mere presence of multiple,
(i.e., the clutch structure). Explanations of confusing reflections and refractions from
how egg aggregates are produced and postu- various water and egg surfaces and the
lates about the evolution of different struc- extreme transparency of egg jellies can in-
tures are largely lacking. It seems apparent terfere with accurate observations. For exam-
that oviductal (i.e., jelly formation), oviposi- ple, Alcala and Brown (1956) reported that
tional (i.e., parental behavior), and postovipo- Rana microdisca lays eggs on moist surfaces
sitional (e.g., jelly dissolution) processes are on stream banks in a ‘‘twice-coiled string.’’
involved. Differences in the rate or pattern of Thisnote,presumingcorrectidentificationsof
release of eggs from a female also can the adult, the eggs, and the ovipositional
influence clutch structure. Oviposition site, mode, is particularly vexing for the following
ovum number and size (Bernardo, 1996; reasons: a moist surface on a stream bank is
Pombal and Haddad, 2005; Salthe and Duell- a rare site for ranid egg deposition, to our
man, 1973; Summers et al., 2007), energy knowledge a string is unknown among ranid
content (e.g., Komoroski and Congdon, 2001; frogs,and ‘twice-coiled’isdifficult to envision
Komoroski et al., 1998), pigmentation, sub- (i.e.,singlestringcoileduponitselfversustwo
sequent characteristics of development, larval strings with simple coils that are intertwined)
ecomorphology,andothersuchfactorsarenot and possibly the result of disturbance.
directlyconsideredinthisreview.Inpart,this The notion that each species oviposits in
is because the functional roles of oviductal one mode is usually sound (see Williams and
(e.g., Greven, 2002, 2003) and especially Tyler, 1994 for an interesting exception;
ovipositional (e.g., Aronson, 1943) factors are Marsh and Borrell, 2001), but some oviposi-
not understood in sufficient detail. Likewise, tional modes seem to blend from one into
inaccurate, nonstandard terminology is a con- another. This apparent blending frequently is
stantly confounding problem. Livezey and the result of variations in the ovipositional
Wright (1947) and Wright and Wright (1949) behavior of the adults, disturbance at the
presented at least 14 terms, many of which deposition sites, comparisons of eggs of
were ill-defined, to describe some arrange- different ages, or a misinterpretation of the
mentofagroupofeggs.Severalotherworkers mode. For example, an investigator who
(e.g., Stebbins, 1951, 2003; Corkran and observes single eggs deposited close together
Thoms, 1996) have used similar terms in the orontopofeachother,butwiththepresence
same or different ways. A notable recent of single, disjunct eggs, might conclude (in-
exceptionisAnstis (2002)who clearly defined accurately in our opinion, see below) that the
the terms she used to describe the types of eggsare aclump. Eggsdeposited bydifferent
egg aggregations of frogs from southeastern taxa as surface films float by different means
Australia.Intheinterestofclarityandwithno and must be observed closely to determine
implication of discrete organization, we use whether the outer jellies are adherent or
‘tier(s)’ to describe a two- or three-dimen- coherent. Pairs of frogs that produce surface
sional arrangement of eggs and restrict ‘layer’ films must have adequate room to maneuver;
to descriptions of jelly morphology. movements in confined spaces sometimes
cause the eggs to sink. Species that deposit
Other Considerations
egg masses and those depositing single eggs
Accurate observations and descriptions of must have suitable substrates to which to
ovipositional events and the appearance of attachtheireggs.Eggsandclutchesdeposited
deposited groups of eggs are lacking for most under artificial conditions (e.g., laboratory
amphibians but deemed essential for under- containers,terrariums,plasticbags)maydiffer
standingamphibianreproductioninabroader from those deposited in the wild. As embryos
context. Difficulties stem from a general lack in clumps and masses approach hatching and
of knowledge of the natural history of jelly layers begin to break down, a group of
amphibians, problems with egg and species eggs may rise to the water surface. Any
identifications, comparisons of eggs of differ- environmental condition that produces lots