Life History of Plethodontids and Other …ronbonett.weebly.com/uploads/2/2/8/5/22856406/bonett...An Integrative Endocrine Model for the Evolution of Developmental Timing and Life

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

An Integrative Endocrine Model for the Evolution of Developmental Timing andLife History of Plethodontids and Other SalamandersAuthor(s): Ronald M. BonettSource: Copeia, 104(1):209-221.Published By: The American Society of Ichthyologists and HerpetologistsDOI: http://dx.doi.org/10.1643/OT-15-269URL: http://www.bioone.org/doi/full/10.1643/OT-15-269

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiriesor rights and permissions requests should be directed to the individual publisher as copyright holder.

An Integrative Endocrine Model for the Evolution of Developmental Timing

and Life History of Plethodontids and Other Salamanders

Ronald M. Bonett1

In recent years, many molecular endocrine mechanisms that regulate tissue morphogenesis have been detailed inlaboratory amphibian models. However, most of these pathways have not been examined across more closely relatedspecies to understand how deviations in endocrine pathways may have contributed to amphibian diversification. Thetiming of metamorphosis and maturation vary extensively across plethodontid salamanders (including directdeveloping, biphasic, and paedomorphic species), making them an ideal system for analyzing the evolution ofendocrine mechanisms in a phylogenetic context. Recent phylogenetic-based reconstructions concluded that ancestralplethodontids were likely direct developers, and free-living larval periods were independently derived multiple timeswithin this family. Furthermore, within one clade (Spelerpini) there have been multiple independent transitions frombiphasic (metamorphic) to paedomorphic developmental modes. This inspires the question: What endocrine/developmental mechanisms govern these extreme life history transitions? Prior endocrine models for the evolutionof direct development and paedomorphosis have been largely based on the ontogenetic timing of thyroid hormonerelease and/or thyroid hormone responsiveness of target tissues. Here I review endocrine pathways that influencemetamorphosis and maturation in laboratory amphibians (Clawed frog and Axolotl) and other species, and develop amodel that integrates prior thyroid hormone-based patterns with other endocrine axes. This integrated frameworkincorporates developmental shifts that result from plasticity or evolution in the timing of hatching, metamorphosis,and maturation, and can be used to test mechanistic changes that underlie life history variation of plethodontids andother salamanders.

AMPHIBIAN metamorphosis involves many concom-itant morphogenic changes in different tissue sys-tems. This indicates that common control

mechanisms, such as endocrine signals, may govern theseprocesses. Amphibians exhibit striking differences in thetiming of developmental events among species (heterochro-nies), which may ultimately be dictated by evolution and/orplasticity of endocrine control mechanisms (Dent, 1942,1968; Rose, 1996; Hanken et al., 1997; Denver, 2013; Elinson,2013; Johnson and Voss, 2013). A reoccurring conclusion isthat life history variation is correlated with ontogenetic shiftsin the timing of thyroid hormone (TH) release or THresponsiveness of target tissues. However, many otherendocrine and developmental pathways, particularly stressresponse and reproduction, are also known to influence lifehistory variation of amphibians, but these have not beenpreviously integrated with TH-based models.

Plethodontid salamanders may prove to be an enlighten-ing system to study the evolution of endocrine mechanismsgiven their extensive lineage and developmental diversifica-tion (Wake, 1966; Hanken, 1992; Ryan and Bruce, 2000;Chippindale et al., 2004; Bonett et al., 2014a, 2014b).Plethodontids exhibit three major developmental (life histo-ry) patterns: 1) biphasic: a free-living aquatic larval form thatmetamorphosis into a more terrestrial form; 2) paedomorphic:a permanently aquatic larval form that forgoes metamor-phosis; and 3) direct development: a precociously metamor-phosing form that lacks a free-living aquatic larval stage.Despite these categorizations, it may be more informative toview these ‘‘distinct’’ developmental modes as a continuumof morphogenic timing, with a primary difference being thelength of time spent in the larval form relative to hatchingand maturation (Ryan and Bruce, 2000; Bonett et al., 2014b).For example, direct development is the acceleration (speed-ing-up) of the time to metamorphosis to an age prior tohatching, whereas paedomorphosis is decelerating (slowing-down) the time to metamorphosis to an age after maturation

(see Bonett et al., 2014b for further discussion of thisconcept).

Ancestral plethodontids likely exhibited direct develop-ment, and therefore the derived biphasic lineages (speler-pines, desmognathines, and Hemidactylium) represent at leastthree independent decelerations in the durations of the larvalperiod (age at metamorphosis; Bonett et al., 2014b; Fig. 1).The lengths of larval periods within biphasic plethodontidsvary tremendously, from less than a few months to morethan a few years. Some of these differences represent furtherderived decelerations leading to very long larval periods (e.g.,Gyrinophilus porphyriticus or Desmognathus quadramaculatus),while other taxa likely represent secondarily derived acceler-ations (e.g., Eurycea quadridigitata). Even though ancestralspelerpines were likely biphasic, there have been severalindependent instances of paedomorphosis across this clade.In at least one clade (Edwards Plateau Eurycea), there waslikely a very recent reversal from paedomorphosis back to abiphasic (metamorphic) life cycle (Bonett et al., 2014a,2014b).

This review describes the general endocrine pathways thatmay underlie the development of plethodontids, based oninformation about metamorphosis and maturation derivedprimarily from studies of laboratory model species and otheramphibians. A model is presented which expands upon thepreviously proposed TH-based life history models by inte-grating endocrine-based plasticity of metamorphic timing, aswell as shifts in hatching and maturation time. This modelcan serve as a framework to test how changes in endocrinemechanisms regulate the plasticity and evolution of lifehistories of plethodontids and other salamanders.

GENERAL ENDOCRINE MECHANISMS REGULATINGMETAMORPHOSIS AND MATURATION

Several endocrine and environmental factors are known toregulate the timing of metamorphosis and maturation of

1 Department of Biological Science, University of Tulsa, Tulsa, Oklahoma 74104; Email: [email protected]: 20 March 2015. Accepted: 2 August 2015. Associate Editor: M. E. Gifford.� 2016 by the American Society of Ichthyologists and Herpetologists DOI: 10.1643/OT-15-269 Published online: 31 March, 2016

Copeia 104, No. 1, 2016, 209–221

amphibians. The hypothalamus plays a central role inmediating the influence of environmental factors on endog-enous hormone levels, which are important determinants ofhow phenotypes and life histories are manifested. Below Iprovide an overview of the relevant aspects of threeendocrine systems (thyroid, corticosteroid, and reproduc-tive), which individually and in concert are major drivers ofamphibian development and life histories.

Thyroid hormone and amphibian metamorphosis.—Amphibi-ans have historically been important models for ourunderstanding of thyroid hormone function, ever sinceGudernatsch (1912) precociously metamorphosed tadpolesof Rana temporaria by feeding them mammalian thyroidglands. Subsequently, metamorphosis was successfully in-duced with thyroid hormone for many amphibian larvae.These included treatments with both 3,5,3-triiodothyronine(T3) and the less active form 3,5,30,50-tetraiodothyronine (T4;thyroxine). Notable exceptions included some obligatelypaedomorphic salamanders (discussed under larval formpaedomorphosis below). Several studies also showed thatplasma levels of T3 and T4 increase during metamorphosis(e.g., Regard et al., 1978; Suzuki and Suzuki, 1981; Alberch etal., 1986; Norman et al., 1987), and ablation of the thyroidgland can prevent metamorphosis (Allen, 1916, 1918;Hoskins and Hoskins, 1919). Metamorphosis can be recov-ered in thyroidectomized larvae by treatment with TH (Allen,1932; Etkin, 1935; Hanaoka, 1966). Taken together, thesestudies demonstrate that thyroid hormone is necessary foramphibian metamorphosis.

The release of T3 and T4 from the thyroid gland is regulatedby the hypothalamic-pituitary-thyroid (HPT) axis (Fig. 2).Thyrotropin-releasing hormone (TRH) from the hypothala-mus causes the release of thyroid-stimulating hormone (TSH)from the anterior pituitary, which stimulates the thyroidgland to release T3 and T4 into the circulatory system.

However, TRH appears to only activate the release of TSH inadult metamorphosed amphibians (Darras and Kuhn, 1983;Denver, 1988; Jacobs et al., 1988). In contrast, hypothalamiccorticotropin-releasing factor (CRF; discussed with cortico-steroids below) appears to regulate the release of TSH in larvalamphibians (Denver and Licht, 1989; Denver, 1993, 1997,2009, 2013; Boorse and Denver, 2002). Plasma TSH levels areundetectable in early stages of tadpole development, butelevate prior to metamorphic climax (Dodd and Dodd, 1976).Treatment of larval amphibians and paedomorphic salaman-ders with TSH has been shown to stimulate thyroid activityand induce metamorphosis (Dent and Kirby-Smith, 1963;Norris et al., 1973; Norman and Norris, 1987). Furtherevidence of the central role of this axis comes fromexperiments that have inhibited metamorphosis by ablatingthe anterior pituitary (Kollros, 1961; Hanaoka, 1966).

At target tissues (e.g., tailfin, external gills, brain), T3 and T4

interact directly with the nuclear transcription factors TRa orTRb (thyroid hormone receptor alpha or beta, collectivelyTRs), which form heterodimers with retinoid X receptor(RXR) and bind to thyroid hormone response elements(TREs) in the genome (Ranjan et al., 1994; Buchholz et al.,2006; Das et al., 2009). Unliganded TRs are bound to co-repressor complexes, which inhibit gene expression (Buch-holz et al., 2003; Sato et al., 2007). Whereas, when T3 and T4

are bound to TRs they recruit co-activators, which can alterthe expression of genes regulated by thyroid hormone (Paulet al., 2005). T3 and T4 are known to upregulate anddownregulate hundreds of genes in target tissues of amphib-ians during metamorphosis (Brown et al., 1996; Denver et al.,1997; Das et al., 2006, 2009; Page et al., 2008, 2009). Amongthese, TRs are autoregulated by T3 and T4 and, at least forTRb, this upregulation is direct (Machuca and Tata, 1992;Ranjan et al., 1994; Das et al., 2009). In other words, T3 andT4 can further accentuate their effects by producing more

Fig. 1. Life history evolution of plethodontid salamanders. The plethodontid phylogeny and reconstruction are based on Bonett et al. (2014b). Thephylogeny was pruned to include relevant major lineages. Branches are colored according to categorical reconstruction of life histories: directdevelopment (yellow), primarily biphasic (red), and primarily paedomorphic (blue). Lineages with a mix of metamorphic and paedomorphicpopulations/species (e.g., Gyrinophilus and Eurycea tynerensis) are shown in purple.

210 Copeia 104, No. 1, 2016

receptors for the hormone to bind. Dominant negative TRbtransgenic tadpoles do not metamorphose (Buchholz et al.,2004), which further supports the necessity of TH andfunctional TRs. The expression patterns of TRs are related tometamorphic timing and life histories in Spadefoot Toads(Hollar et al., 2011) and Oklahoma Salamanders (Aran et al.,2014; discussed more below).

It should also be noted that TH activity can be altered intarget tissue via deiodinases, which are a class of enzymesthat can add or remove iodine to produce more or less activeforms of the hormone (Becker et al., 1997; Brown, 2005; St.Germain et al., 2009). For example, type-2 deiodinaseconverts T4 into the more active T3 with the removal of aniodine atom. The transcriptional actions of TH and associatedregulatory mechanisms ultimately govern tissue transforma-tion and metamorphosis of larval amphibians.

Corticosteroids and interactions with thyroid hormone.—Corti-costeroids are regulated through the hypothalamic-pituitary-interrenal (HPI) axis and are important mediators of‘‘environmental stress’’ in vertebrates (Denver, 2009, 2013;Carr, 2010; Crespi et al., 2013; Fig. 2). CRF from the

hypothalamus regulates adrenocorticotropic hormone(ACTH) secretion from the anterior pituitary, and ACTHstimulates the production and release of corticosteroids(glucocorticoids and mineralocorticoids) from the interrenalglands. Corticosteroids can regulate transcription in targettissues by binding to corticoid receptor (CRs, includingglucocorticoid or mineralocorticoid receptor, GRs or MRs,respectively) homodimers in the cytoplasm, which thentranslocate to the nucleus and bind to glucocorticoid ormineralocorticoid response elements (GREs or MREs) in thegenome (Denver, 2009, 2013). The pathways of corticoste-roids and their effects on larval and adult amphibians arecomplex and include alterations in activity, behavior,feeding, body weight, growth, reproduction, shape, and thetiming of metamorphosis (Glennemeier and Denver, 2002;Coddington and Moore, 2003; Denver, 2009, 2013; Fraker etal., 2009; Carr, 2010; Bliley and Woodley, 2012; Crespi et al.,2013; Reedy et al., 2014).

During early developmental stages, tadpoles treated withcorticosteroids or placed in stressful environments candecrease growth and development but increase tail depth(Glennemeier and Denver, 2002; Bonett et al., 2010;

Fig. 2. Consensus of general endocrine pathways for amphibian metamorphosis via the HPT and HPI axes. The pathway depicts CorticotropinReleasing Factor (CRF) from the hypothalamus as the regulator of both Adrenocorticotropic Hormone (ACTH) and Thyroid Stimulating Hormone(TSH) from the pituitary. These pituitary hormones stimulate the release of corticosteroids from the adrenocortical cells of the interrenal gland, and T3

and T4 from the thyroid gland, respectively. Gene expression is regulated in target tissues (e.g., tailfin) by these hormones through binding to theirrespective nuclear receptors: corticoid receptors (CRs, including GRs or MRs) or thyroid hormone receptors (TRs). Sufficient increases in thesehormones can drive the transformation of target tissues and metamorphosis, as depicted here with Eurycea tynerensis (see text for details).

Bonett—Endocrinology of salamander metamorphosis 211

Middlemis Maher et al., 2013). However, during amphibianmetamorphosis, corticosteroids increase endogenously (Kruget al., 1983; Carr and Norris, 1988; Denver et al., 1998;Chambers et al., 2011) and have been shown to alter theeffects of thyroid hormone on morphogenic changes(Frieden and Naile, 1955; Kobayashi, 1958; Kikuyama et al.,1983; Kuhn et al., 2004; Bonett et al., 2010). Mostimportantly, there is significant cross talk between the HPIand HPT axes at both the neuroendocrine level and in targettissues. For example, during metamorphosis the anteriorpituitary expresses CRF type 2 receptor, which allowshypothalamic CRF to directly stimulate the release of TSH(Manzon and Denver, 2004; Okada et al., 2007; see alsoabove). In target tissues, corticosterioids can enhance theeffects of thyroid hormone through several avenues. Tran-scriptional regulatory mechanisms driven by corticosterioidsinclude the regulation of genes that code for thyroidhormone converting enzymes (e.g., type-2 deiodinase;Bonett et al., 2010) or for transcription factors thatupregulate thyroid hormone receptor expression (e.g.,Kruppel-like factor 9, KLF9, formerly known as BTEB1;Bagamasbad et al., 2008; Bonett et al., 2009). More complexinteractions between corticosteroids and thyroid hormoneoccur at the level of gene regulation, including additive,synergistic, and inhibitory effects (Bonett et al., 2010;Kulkarni and Buchholz, 2012). For example, corticosteroidsand thyroid hormone have greater than additive (synergistic)effects on transcription of thyroid hormone receptors,deiodinases, and KLF9 (Kuhn et al., 2005; Bonett et al.,2010; Kulkarni and Buchholz, 2012; Bagamasbad et al.,2015). These regulatory effects may provide a link betweencorticosteroids and the rapid morphogenic changes in latestage larvae, which could expedite their departure from a‘‘stressful’’ larval environment such as a crowded, dryingpond.

Reproductive hormones.—Understanding the timing of repro-ductive development is critical for reconstructing theevolution of maturation with respect to other events (e.g.,metamorphosis) and also because reproductive hormonescan both stimulate and inhibit other endocrine pathways. Italso is a pivoting point for shifting resource allocationbetween growth/morphogenesis and reproduction (Stearns,1992).

Sex steroids (estrogens and androgens) can function in asimilar manner as corticosteroids, by binding to nucleartranscription factors (estrogen and androgen receptors) toregulate gene transcription. There is also significant crosstalkbetween the hypothalamic-pituitary-gonadal (HPG) axis andHPT axes. This topic has been extensively reviewed (Hayes,1997; Duarte-Guterman et al., 2014), and most research hasfocused on the effects of endocrine disruptors on reproduc-tive development and sex ratios. The most relevant aspectsfor this review are the influence of thyroid hormone ongonadal development (Flood et al., 2013) and sex steroids onthe inhibition of metamorphosis. Thyroid-less and thyroid-suppressed (goitrogen treated) tadpoles and larval salaman-ders can still develop gonadal germ cells (Allen, 1918;Wakahara, 1994; Yamaguchi et al., 1996; Kanki and Waka-hara, 1999; Rot-Nikcevic and Wassersug, 2004), and this maybe mediated through an increase in TSH (Kanki andWakahara, 1999). The ability of larval salamanders to developgonads without an increase of TH (which could stimulatemetamorphosis) provides the potential for larval formpaedomorphosis. At the same time, both estrogens and

androgens can inhibit tadpole metamorphosis (Gray andJanssens, 1990; Hogan et al., 2008), but data for salamandersare limited. If this mechanism is common among allamphibians, then the acceleration of maturation and theproduction of sex steroids into larval stages could perma-nently displace metamorphosis (Ryan and Semlitsch, 1998).

ENDOCRINE BASIS OF HETEROCHRONY INPLETHODONTIDS AND OTHER SALAMANDERS

Shifts in the timing of developmental events (heterochrony)such as hatching, metamorphosis, and maturation haveoccurred extensively among salamanders and probably mostextremely among plethodontids. These have resulted in threemajor developmental-based life histories: direct develop-ment, biphasic, and paedomorphic. Studies that examinethe endocrine basis of these developmental patterns inplethodontids and other amphibians are discussed immedi-ately below.

Direct development.—Direct development is widespread inplethodontids and has been considered an importantcontributor to their diversification, especially in the neo-tropics (Wake and Hanken, 1996). Direct development islikely the ancestral condition for the family and has beenmaintained in most major lineages except for spelerpines,one clade of desmognathines, and Hemidactylium (Bonett etal., 2014b; Fig. 1). The molecular mechanisms that underliedirect development are compelling, because somatic mor-phogenesis is accelerated so rapidly, and at such a small bodysize, that metamorphosis occurs completely within the egg(Dent, 1942; Wake and Hanken, 1996; Kerney et al., 2012;Bonett et al., 2014b). Surprisingly, there have been only a fewmolecular endocrine studies on direct developing pletho-dontids, although there is additional information that can bederived from the many studies on direct developing frogs.

Dent (1942) proposed, and partially tested, thyroidhormone-based mechanisms to explain direct developmentin plethodontids. He proposed that direct developers eitherhad precociously developed pituitary and thyroid glands (totrigger early metamorphosis) or that the tissues of directdevelopers had an increased sensitivity to thyroid hormone.Histological and morphological examination of Plethodoncinereus showed that the pituitary and thyroid gland developat relatively early embryonic stages (Dent, 1942). Further-more, he observed morphological changes to the thyroidgland that signified secretory activity, and were coincidentwith several morphogenic changes such as gill resorption.These changes in embryos of P. cinereus (between stages XXIIand XXIII) are similar to those that occur during earlymetamorphosis of Eurycea bislineata, which have a multi-yearlarval period (Wilder, 1925; Dent, 1942).

The mechanisms of direct development in the frogEleutherodactylus coqui have been investigated much moreextensively (reviewed in Elinson, 2013), and this speciesshares some important similarities with P. cinereus. Forexample, E. coqui also develop their thyroid system embry-onically, well before species with free-living tadpoles, andpeaks of thyroid activity are associated with metamorphicchanges (Hanken et al., 1997; Jennings and Hanken, 1998).CRF is important for metamorphosis in E. coqui and likelyoperates by directly stimulating TH release from the thyroidgland (Kulkarni et al., 2010; Elinson, 2013). However, whiletreatment of embryos of E. coqui with the goitrogenmethimazole prevents complete metamorphosis, many frog-

212 Copeia 104, No. 1, 2016

let features still develop (Callery and Elinson, 2000). Thissuggests that either thyroid hormone is provided maternallyin yolk, or other hormones (such as corticosterone) aredriving metamorphosis. The potential influence of cortico-steroids on metamorphosis of direct developing amphibiansmakes sense from a phylogenetic perspective, because alldirect developing lineages were originally derived frombiphasic ancestors, which likely had plastic larval periods.This may also explain the variation in metamorphic timingof biphasic species such D. fuscus that typically have anaquatic larval period of several months (Danstedt, 1975) butcan complete metamorphosis after less than one monthwhen raised strictly on wet surfaces without free-standingwater (Noble and Evans, 1932; see also Marks and Collazo,1998 for further discussion about D. aeneus). This shows thatsome biphasic plethodontids can effectively exhibit directdevelopment under some conditions.

Our experiments on hatchling Eurycea tynerensis, a speciesthat has minimally a several month larval period (or larvalform paedomorphosis), showed that treatment with exoge-nous thyroid hormone upregulates TRb in tail tissue andsignificantly advances nasal capsule development. Exoge-nous corticosterone alone has no gross developmentaleffects, but in combination with thyroid hormone has strongsynergistic effects on TRb expression and nasal capsuledevelopment (Jackson et al., unpubl.). However, other tissuesthat normally transform during metamorphosis of Eurycea,such as the hyobranchial apparatus, do not seem to besensitive to thyroid hormone or corticosterone treatment atsuch early stages of development (Jackson et al., unpubl.). Aswith direct developing frogs (Elinson, 2013), this indicatesthat other hormones may be necessary for the transforma-tion of some tissues, or local differences in tissue sensitivitydetermine whether or not metamorphosis can be completedprior to hatching.

Biphasic.—Biphasic life histories occur in spelerpines, oneclade of desmognathines, and Hemidactylium (Fig. 1). Perhapsthe most interesting aspect of biphasic life cycles ofplethodontids is that they are likely independently derivedfrom direct developing ancestors (Chippindale et al., 2004;Bonett et al., 2014b; Fig. 1). This suggests that not only wasthere potentially a major endocrine-based acceleration oflarval development in ancestral plethodontids (or earlier) toachieve direct development, but there may have beenmultiple endocrine-based decelerations in the ‘‘re-evolution’’of larval forms in desmognathines, spelerpines, and Hemi-dactylium. If these are indeed independent reversals to free-living aquatic larvae, then their origins may be based ondifferent mechanisms. These may have involved delays inthe development of the HPT axis or decreasing sensitivity oftissues to thyroid hormone and/or other endocrine signals.

Larvae of biphasic species/populations of spelerpines havebeen transformed using both T3 and T4 (Alberch et al., 1985;Rose, 1995a, 1995b, 1996; Aran et al., 2014). Hormonedosage and the age/stage of larvae influence the timing oftransformation of skeletal elements (Rose, 1995a, 1995b).Treatment of larvae from multiple metamorphic populationsof E. tynerensis with exogenous T3 upregulates TRa and TRbexpression in tailfin tissue (Aran et al., 2014).

Hickerson et al. (2005) found that metamorphic progress offirst year larvae of Desmognathus quadramaculatus was notinfluenced by a low dose of T4 (up to 4.8 nM) administeredover approximately two months. We found that similar doses(1 to 5 nM) of the more active hormone T3 were highly

influential on metamorphic changes of larval D. ocoee, D.santeetlah, and D. brimleyorum over periods of three to fivemonths (Robison et al., unpubl.).

Alberch et al. (1986) performed radioimmunoassays for T3

and T4 on a series of E. bislineata sampled from acrossmetamorphosis. They found the common pattern ofincreased concentrations of plasma T3 and T4 in metamor-phic individuals, but neither hormone was detectable insome specimens that were clearly in the process of meta-morphosis. This suggested that hormone levels may fluctuateduring metamorphosis, or that other hormones are used inthe process.

The lengths of larval periods of biphasic plethodontidsvary extensively among and within species (Petranka, 1998;AmphibiaWeb, 2015). Several ecological studies have shownthat factors such as temperature, hydroperiod, and streamorder can further influence metamorphic timing (Voss, 1993;Beachy, 1995; Camp et al., 2000; Freeman and Bruce, 2001;Bruce, 2005; Hickerson et al., 2005) but food availability doesnot (e.g., O’Laughlin and Harris, 2000; Hickerson et al.,2005). While variation and plasticity indicate that endocrineprocesses play a major role in regulating metamorphic timing(Rose, 2005; Denver, 2009, 2013; Crespi et al., 2013), therehave been relatively few published studies that simulta-neously evaluate the effects of environmental and endocrinefactors on biphasic plethodontids (Hickerson et al., 2005).

Temperature is known to influence the rate of larvalgrowth, development, and the timing of metamorphosis ofplethodontids, with low temperatures delaying metamor-phosis (Hickerson et al., 2005). An important endocrine facetof this effect, which has been generally underexplored insalamanders, is the potential for low temperatures to inhibitTH-induced metamorphosis, as seen in studies with larvae ofHynobius (Moriya, 1983). Integrative studies are needed toassess how environmental factors and natural stressorsinfluence TH regulated gene expression programs to controlmetamorphic timing of biphasic plethodontids.

Larval form paedomorphosis.—Shifts from biphasic to larvalform paedomorphic developmental modes have occurredmany times in salamanders, and nine of the ten salamanderfamilies include larval form paedomorphic species. Inplethodontids, larval form paedomorphosis is restricted toseveral independent instances in the Spelerpini, which areassociated with aquatic subterranean habitats and aridclimate regimes (Bonett et al., 2014a; Fig. 1). These includeisolated instances of paedomorphosis in primarily metamor-phic clades (e.g., E. cirrigera; McEntire et al., 2014), cladeswith highly variable life histories among populations (e.g., E.tynerensis, Bonett and Chippindale, 2004, 2006; Emel andBonett, 2011; Gyrinophilus, Niemiller et al., 2008), divergentpaedomorphic lineages (e.g., E. subfluvicola; Steffen et al.,2014), and primarily paedomorphic clades (Edwards PlateauEurycea; Chippindale et al., 2000). While most instances ofpaedomorphosis are derived, there is strong phylogeneticsupport that life history is reversible in spelerpines. Inparticular, the Edwards Plateau clade of Eurycea likelyincludes a derived instance of metamorphosis after severalmillion years of paedomorphosis (Bonett et al., 2014a,2014b).

Obligately paedomorphic salamanders, which never natu-rally metamorphose and also do not completely transformwith thyroid hormone treatment, include all species in thefamilies Amphiumidae (Kobayashi and Gorbman, 1962),Cryptobranchidae (Noble and Farris, 1929), Proteidae (Swin-

Bonett—Endocrinology of salamander metamorphosis 213

gle, 1922; Svob et al., 1973), and Sirenidae (Noble, 1924), aswell as the plethodontid species Eurycea rathbuni (Dundee,1957) and E. wallacei (Dundee, 1962). Analyses of thyroidhormone sensitivity in the obligately paedomorphic Protei-dae showed that their thyroid hormone receptors arefunctional and responsive to thyroid hormone (Safi et al.,2006). Therefore, thyroid hormone control has probablybeen decoupled from the process of transformation inancestrally metamorphic tissues (e.g., external gills; Safi etal., 2006; Vlaeminck-Guillem et al., 2006).

Facultatively paedomorphic salamanders, which can nat-urally metamorphose under some conditions or can becompletely transformed by stimulation of the HPT or HPIaxes, occur at varying frequencies in several families:Ambystomatidae (Collins, 1981; Semlitsch, 1985), Dicamp-todontidae (Nussbaum and Clothier, 1973), Hynobiidae(Sasaki, 1924), Salamandridae (Denoel et al., 2001), andPlethodontidae (Dent and Kirby-Smith, 1963; Aran et al.,2014; Bonett et al., unpubl.). Extensive genetic crosses andexperiments with species from the Ambystoma mexicanum/A.tigrinum clade have identified multiple allelic variants thatdiffer in metamorphic timing and thyroxine sensitivity (Vossand Shaffer, 1997; Voss and Smith, 2005; Voss et al., 2012;recently reviewed by Johnson and Voss, 2013). These allelicvariants appear to be a primary determinant of metamor-phosis in this group.

Approximately half of the populations in the western cladeof Eurycea tynerensis are paedomorphic (Emel and Bonett,2011). Most populations include only a single life historymode, and larval form paedomorphosis is associated withchert streambed substrate (Bonett and Chippindale, 2006;Emel and Bonett, 2011), which permits access to stable,subsurface aquatic environments (Treglia et al., unpubl.).There are a relatively small number of populations of E.tynerensis that are known to exhibit both paedomorphosisand metamorphosis. A few of the variable populations arefacultatively paedomorphic, based on the fact that paedo-morphic females layed fertilized eggs in the lab, but thencompleted metamorphosis in subsequent years. By compar-ison, other variable populations are composed of sympatric,but genetically distinct, paedomorphic and metamorphicpopulations (Bonett et al., unpubl.). In addition, somepaedomorphic populations of E. tynerensis have been main-tained in the lab (at the University of Tulsa) for several yearswithout spontaneous metamorphosis, whereas larvae frommetamorphic populations raised under the same conditionsalways metamorphose eventually. These observations suggestthat variability in metamorphic timing among populations isnot completely driven by environment. Larval E. tynerensisfrom metamorphic populations are more sensitive to exog-enous T3 treatment than larvae from paedomorphic popula-tions, with respect to both thyroid hormone receptor (TRa orTRb) upregulation and metamorphosis (Aran et al., 2014).Furthermore, larvae from paedomorphic populations alsodiffer in their sensitivity to T3, with extremely reducedsensitivity in some populations (Aran et al., 2014). Thisindicates that both facultative and obligate paedomorphosismay occur among populations of E. tynerensis occurringwithin a very narrow geographic area in the western OzarkPlateau.

Direct comparisons of circulating hormone levels betweenpaedomorphic and metamorphic plethodonids have notbeen conducted. Analyses of the assimilation of radioiodineby adult paedomorphic E. tynerensis showed that theirthyroid gland functioned normally, but they exhibited lower

levels of thyroid activity compared to other vertebrates(Dundee and Gorbman, 1960). Using similar methods, Dentand Kirby-Smith (1963) found that paedomorphic Gyrinophi-lus palleucus exhibited relatively low levels of thyroid activity,but activity was considerably higher in individuals thatshowed signs of spontaneous metamorphosis. These studiesdocumented general patterns of vertebrate endocrinology,but it is difficult to evaluate their evolutionary significancewithout direct comparisons to related metamorphic species.

INTEGRATIVE ENDOCRINE MODEL FOR SALAMANDER LIFEHISTORY EVOLUTION

The extensive variation in metamorphic timing amongamphibians and the pervasive influence of thyroid hormone,as well as changes in the thyroid gland itself, lead severalinvestigators to suggest that variations in the HPT axis maybe a driver of heterochronic evolution (Lynn, 1936, 1942;Dent, 1942, 1968; Rose, 1995a, 1995b, 1996; Hanken et al.,1997; Elinson, 2013; Johnson and Voss, 2013). Most recently,Johnson and Voss (2013) developed a salamander life historymodel (based largely on Ambystoma), which synthesizedcommonly identified differences associated with paedomor-phosis versus metamorphosis. This model showed thatcompared to metamorphosis, paedomorphosis is associatedwith greater habitat stability, lower TH responsiveness, andlarger body size (Johnson and Voss, 2013).

Here I provide an extension to previous TH-based lifehistory models by incorporating additional endocrine anddevelopmental processes that influence plasticity of meta-morphic timing and maturation (Fig. 3). This integratedmodel details where the three primary salamander lifehistory categories (direct development, biphasic, and paedo-morphic) occur in this spectrum and shows how plasticitycan result in variation in their expression. It also shows howsimilar life histories can be derived from different endocrineand developmental processes. The model can be used as aframework for both comparative endocrine analyses andendocrine manipulation experiments.

Both TH responsiveness and the timing/amount ofendogenous TH release can influence the timing of meta-morphosis; therefore, I considered both processes in thismodel (Fig. 3A). Timing of metamorphosis is generallypositively correlated with TH release and negatively correlat-ed with TH responsiveness. In other words, early TH releaseand high TH responsiveness are related to early metamor-phosis (e.g., direct development), and late TH release and lowTH responsiveness are related to late metamorphosis (e.g.,long larval periods and paedomorphosis). The evolution ofgenetic-based differences in ‘‘baseline’’ TH responsiveness/release could move a lineage along the diagonal betweendirect development and paedomorphosis, presumably alwayspassing through a biphasic life history.

At the same time, many other factors such as environ-mental stressors, temperature, and reproductive maturity(described above) can significantly influence the timing ofamphibian metamorphosis for a given genotype (or species).These factors typically work through molecular endocrineprocesses to influence metamorphic timing, but the magni-tude and direction of these effects can vary across species, lifehistories, and stages (detailed above). Nevertheless, theseeffects, individually or in combination, can push metamor-phic timing to its theoretical minimum or maximum for agiven genotype. I depict this plasticity along the ‘‘baseline’’

214 Copeia 104, No. 1, 2016

TH diagonal of the model, and show how plasticity of a givengenotype could result in a life history shift (Fig. 3A).

The model also shows two primary events that transect theontogeny of nearly all salamanders: hatching from anoviparous egg and maturation. Both of these events likelyhave ‘‘baselines’’ for a given genotype (or species), but mayalso be influenced by plasticity. Evolutionary or plasticity-based accelerations or decelerations of hatching and matu-ration could also result in major life history shifts (Fig. 3B, C).The following describes pathways to transitions betweenmajor life history categories, as well as permanent transitionsto paedomorphosis (obligate paedomorphosis) and the lackof transitions in some clades.

Transitions between direct development and biphasic.—Whenmetamorphosis is temporally proximal to hatching, thengenetic changes to baseline TH responsiveness/release orplasticity effects could shift a lineage from direct develop-ment to biphasic or vice versa through accelerations ordecelerations (Fig. 3A) in metamorphic timing. Also, geneticor plasticity effects that cause hatching time to shift before(Fig. 3B) or after (Fig. 3C) metamorphosis could also result intransitions between these life histories.

Transitions between biphasic and paedomorphic.—Whenmetamorphosis is proximal to maturation, genetic changesto baseline TH responsiveness/release or plasticity effectscould shift a lineage from biphasic to facultatively paedo-morphic, or the reverse. The transitions could occur throughdeceleration (neoteny) or acceleration of metamorphictiming (somatic morphogenesis). These transitions can also

be achieved by acceleration (progenesis; Fig. 3B) or deceler-ation (Fig. 3C) of maturation relative to metamorphosis(Alberch et al., 1979). Both mechanisms have been demon-strated in salamanders, sometimes even within a singlespecies (Denoel and Joly, 2000). The only phylogenetic test ofthese alternatives (neoteny vs. progenesis) was performed onEdwards Plateau Eurycea, where paedomorphosis was likelyderived through neoteny (Bonett et al., 2014b). However,given the extensive crosstalk between endocrine pathways,these processes may not be completely independent in allcases. For example, the loss of thyroid function should causea truncation of somatic development (neoteny), but maysimultaneously speed up reproductive development (progen-esis) as shown in frogs and salamanders (Gray and Janssens,1990; Wakahara, 1994; Hogan et al., 2008; see discussion ofreproductive hormones above).

Transitions to obligate paedomorphosis.—Obligate paedo-morphs are likely derived from facultatively paedomorphicancestors, but are instead infinitely paedomorphic (Fig. 3).They present an unusual case and may completely deviatefrom the model in several ways. Once TH is decoupled fromtissue transformation, the level or timing of TH release doesnot need to be low or late to prevent metamorphosis. Inother words, in terms of transformation, the TH responsive-ness of the decoupled tissues (e.g., external gills of Necturus;Safi et al., 2006; Vlaeminck-Guillem et al., 2006) will benonexistent, so the level and ontogenetic release time ofcirculating TH could fall anywhere on the spectrum com-pared to other salamander life histories. It should be notedthat some, but not all, tissues remain responsive in some

Fig. 3. Endocrine model for salamander life history plasticity and evolution. On all three graphs (A–C), the diagonal depicts the relationship betweenmetamorphic timing and thyroid hormone (TH) sensitivity and release, as suggested by many authors. The dashed lines depict the degree of plasticityin metamorphic timing for a given genotype. Horizontal lines indicate hatching (hat) and reproductive maturation (mat), and vertical lines bracket themajor life history categories (direct development, biphasic, paedomorphic) with respect to these events for a given genotype (position along thediagonal). Facultative and obligate paedomorphosis are also depicted separately, due to the nuances of obligate paedomorphosis (see text). Colorsindicate the most common environment of the individual during the life cycle with respect to hatching and metamorphosis: egg (yellow), aquatic,(blue), terrestrial (red), and plasticity between aquatic and terrestrial (purple). (A) Arrows between dashed lines indicate plasticity in metamorphosisfor a given genotype. (B) Down arrows in the right margin show the effects of acceleration of maturation and hatching on life history shifts (relative toA). (C) Up arrows in the right margin show the effects of deceleration of maturation and hatching on life history shifts (relative to A).

Bonett—Endocrinology of salamander metamorphosis 215

obligate paedomorphs, such as E. rathbuni and E. wallacei(Dundee, 1957, 1962), so the model should be specified as areference to either metamorphic timing of a particular tissueor a whole animal. Another unique attribute of obligatepaedomorphosis is that since metamorphosis never occurs,environmental effects (e.g., via ‘‘stress hormones’’) on thenon-responsive tissues should also be nonexistent. Environ-mental factors could, however, still affect tissues thatmetamorphose, such as the external gills of Amphiuma thatresorb just after hatching.

Lack of certain transitions in some clades.—Larval formpaedomorphosis evolved several times among the speler-pines, but is completely unknown for some spelerpine clades,as well as Desmognathus and Hemidactylium (Fig. 1). This issurprising given that many of these biphasic clades havelarvae that develop in permanently aquatic habitats. Con-sidering the origins of their biphasic life histories in thecontext of the life history model presented here may providesome clarity. Phylogenetic reconstructions show that biphas-ic Desmognathus and Hemidactylium (as well as spelerpines)are independently derived from direct developing ancestors(Chippindale et al., 2004; Bonett et al., 2014b). This meansthat they were derived from ancestors with highly acceler-ated metamorphic timing, and even though they currentlyexhibit free-living larval forms, the ontogenetic timingbetween metamorphosis and maturation may be too longto be bypassed through plasticity of either event. Forexample, Hemidactylium and many Desmognathus have larvalperiods of less than several months, but take minimally twoyears to reach maturation, so plasticity in these events wouldhave to be collectively shifted by more than a year formaturation to precede metamorphosis. It is also possible thata developmental mechanism prevents maturation fromoccurring without metamorphosis, such as in frogs. TheDesmognathus quadramaculatus/marmoratus clade is a uniqueexample that may be an informative test case. These specieshave long larval periods (two to three years) and minimallymature one year later, but always metamorphose eventhough D. marmoratus remain permanently aquatic as adults.Therefore, D. marmoratus accomplishes ‘‘ecological paedo-morphosis’’ by retaining their aquatic larval ecology withouta complete truncation of somatic development.

REQUIREMENTS FOR SYNTHESIS AND FUTURE RESEARCH

Much of the data on the timing of hatching, metamorphosis,and maturation for plethodontids was collected from the1950s through the 1980s. However, our knowledge ofphylogenetic relationships and species diversity were stillrudimentary at the time, limiting evolutionary interpreta-tions of life history data. Tremendous progress has beenmade on plethodontid phylogeny and species diversity overthe last decade, but now much more life history data areneeded to make accurate and comprehensive estimates of theevolution of developmental timing. An added complicationis the potential effect of plasticity in developmental timing,which must be separated from genetic-based, geographicvariation. Therefore, measurements of salamanders raisedunder common laboratory conditions are required to obtainbaseline estimates of developmental timing.

Several plethodontids have been treated with T3 or T4

(Kezer, 1952; Dundee, 1957, 1962; Dent and Kirby-Smith,1963; Rose, 1996; Hickerson et al., 2005; Aran et al., 2014),but more experiments are needed that include a common

dose and duration to make comparative evaluations ofsensitivity. There have been relatively few analyses of thyroidgland activity (Dent, 1942; Dent and Lynn, 1958; Dundeeand Gorbman, 1960; Dent and Kirby-Smith, 1963), and nostudies have directly compared related species that exhibitdifferent life histories. Also, there has been little publisheddata on transcriptional differences among plethodontid lifehistories (Aran et al., 2014; Jackson et al., unpubl.). Suchstudies are critical for establishing endocrine correlates formetamorphic timing (e.g., TH responsiveness and THrelease).

Similarly, several studies have examined a diversity ofenvironmental effects on metamorphic timing of differentspecies (Noble and Evans, 1932; Voss, 1993; Beachy, 1995;Camp et al., 2000; Freeman and Bruce, 2001; Hickerson et al.,2005), but there is very little data on the effects of stresshormones on metamorphic timing of plethodontids. Studiesare needed to systematically compare individual factorsamong species to establish the degree of the variance aroundbaseline metamorphic timing. In particular, it will beinformative to test species with baseline metamorphic timing(or TH sensitivity) that are near the transitions of life historycategories. This will help to establish the degree of plasticityrequired for major life history transitions and determinewhether some taxa are unable to make such transitions.

Finally, molecular endocrine experiments that examine theontogeny of hormone release and the effects of hormones ontranscription in target tissues are needed to determinewhether patterns of laboratory models are generalizableacross plethodontids. It will also ultimately permit phyloge-netic tests of how the evolution of endocrine pathways haveinfluenced plethodontid life histories.

ACKNOWLEDGMENTS

I would like to thank Matt Gifford for co-organizing thePlethodontid Conference and co-editing these proceedings. Iwould also like to thank C. Beachy, S. Martin, J. Phillips, M.Steffen, and A. Trujano for their comments on this reviewand fruitful discussions of the topic. This study was in partfunded by the National Science Foundation (DEB 1050322and OK EPSCoR IIA-1301789).

LITERATURE CITED

Alberch, P., E. A. Gale, and P. R. Larsen. 1986. Plasma T4

and T3 levels in naturally metamorphosing Euryceabislineata (Amphibia; Plethodontidae). General and Com-parative Endocrinology 61:153–163.

Alberch, P., S. J. Gould, G. F. Oster, and D. B. Wake. 1979.Size and shape in ontogeny and phylogeny. Paleobiology5:296–317.

Alberch, P., G. A. Lewbart, and E. A. Gale. 1985. The fate oflarval chondrocytes during the metamorphosis of theepibranchial in the salamander Eurycea bislineata. Journalof Embryology and Experimental Morphology 88:71–83.

Allen, B. M. 1916. Extirpation experiments in Rana pipienslarvae. Science 44:755–757.

Allen, B. M. 1918. The results of thyroid removal in thelarvae of Rana pipiens. Journal of Experimental Zoology 26:499–519.

Allen, B. M. 1932. The response of Bufo larvae to differentconcentrations of thyroxin. The Anatomical Record 54:45–64.

216 Copeia 104, No. 1, 2016

AmphibiaWeb. 2015. Information on amphibian biologyand conservation. Berkeley, California. http://amphibiaweb.org/ (Accessed 1 February 2015).

Aran, R. P., M. A. Steffen, S. D. Martin, O. I. Lopez, and R.M. Bonett. 2014. Reduced effects of thyroid hormone ongene expression and metamorphosis in a paedomorphicplethodontid salamander. Journal of Experimental Zoology322B:294–303.

Bagamasbad, P. D., R. M. Bonett, L. Sachs, N. Buisine, S.Raj, J. R. Knoedler, Y. Kyono, Y. Ruan, X. Ruan, and R. J.Denver. 2015. Deciphering the regulatory logic of anancient, ultraconserved nuclear receptor enhancer module.Molecular Endocrinology 29:856–872.

Bagamasbad, P., K. L. Howdeshell, L. M. Sachs, B. A.Demeneix, and R. J. Denver. 2008. A role for basictranscription element binding protein 1 (BTEB1) in theautoinduction of thyroid hormone receptor beta. Journalof Biological Chemistry 283:2275–2285.

Beachy, C. K. 1995. Effects of larval growth history onmetamorphosis in a stream-dwelling salamander (Desmog-nathus ochrophaeus). Journal of Herpetology 29:375–382.

Becker, K. B., K. C. Stephens, J. C. Davey, M. J. Schneider,and V. A. Galton. 1997. The type 2 and type 3iodothyronine deiodinases play important roles in coordi-nating development in Rana catesbeiana tadpoles. Endo-crinology 138:2989–2997.

Bliley, J. M., and S. K. Woodley. 2012. The effects ofrepeated handling and corticosterone treatment on behav-ior in an amphibian (Ocoee salamander: Desmognathusocoee). Physiology and Behavior 105:1132–1139.

Bonett, R. M., and P. T. Chippindale. 2004. Speciation,phylogeography and evolution of live history and mor-phology in the salamanders of the Eurycea multiplicatacomplex. Molecular Ecology 13:1189–1203.

Bonett, R. M., and P. T. Chippindale. 2006. Streambedmicrostructure predicts evolution of development and lifehistory mode in the plethodontid salamander, Euryceatynerensis. BMC Biology 4:6

Bonett, R. M., E. D. Hoopfer, and R. J. Denver. 2010.Molecular mechanisms of corticosteroid synergy withthyroid hormone during tadpole metamorphosis. Generaland Comparative Endocrinology 168:209–219.

Bonett, R. M., F. Hu, P. Bagamasbad, and R. J. Denver.2009. Stressor and glucocorticoid-dependent induction ofthe immediate early gene Kruppel-like factor 9: implica-tions for neural development and plasticity. Endocrinology150:1757–1765.

Bonett, R. M., M. A. Steffen, S. M. Lambert, J. J. Wiens, andP. T. Chippindale. 2014a. Evolution of paedomorphosis inplethodontid salamanders: ecological correlates and re-evolution of metamorphosis. Evolution 68:466–482.

Bonett, R. M., M. A. Steffen, and G. A. Robison. 2014b.Heterochrony repolarized: a phylogenetic analysis ofdevelopmental timing in plethodontid salamanders. Evo-Devo 5:27.

Boorse, G. C., and R. J. Denver. 2002. Acceleration ofAmbystoma tigrinum metamorphosis by corticotropin-re-leasing hormone. Journal of Experimental Zoology 293:94–98.

Brown, D. D. 2005. The role of deiodinases in amphibianmetamorphosis. Thyroid 15:815–821.

Brown, D. D., Z. Wang, J. D. Furlow, A. Kanamori, R. A.Schwartzman, B. F. Remo, and A. Pinder. 1996. Thethyroid hormone-induced tail resorption program duringXenopus laevis metamorphosis. Proceedings of the National

Academy of Sciences of the United States of America 93:1924–1929.

Bruce, R. C. 2005. Theory of complex life cycles: applicationin plethodontid salamanders. Herpetological Monograph19:180–207.

Buchholz, D. R., S. C. Hsia, L. Fu, and Y-B. Shi. 2003. Adominant-negative thyroid hormone receptor blocks am-phibian metamorphosis by retaining corepressors at targetgenes. Molecular and Cellular Biology 23:6750–6758.

Buchholz, D. R., B. D. Paul, L. Fu, and Y-B. Shi. 2006.Molecular and developmental analyses of thyroid hor-mone receptor function in Xenopus laevis, the Africanclawed frog. General and Comparative Endocrinology 145:1–19.

Buchholz, D. R., A. Tomita, L. Fu, B. D. Paul, and Y-B. Shi.2004. Transgenic analysis reveals that thyroid hormonereceptor is sufficient to mediate the thyroid hormonesignal in frog metamorphosis. Molecular and CellularBiology 24:9026–9037.

Callery, E. M., and R. P. Elinson. 2000. Thyroid hormone-dependent metamorphosis in a direct developing frog.Proceedings of the National Academy of Sciences of theUnited States of America 97:2615–2620.

Camp, C. D., J. L. Marshall, and R. M. Austin, Jr. 2000. Theevolution of adult body size in black-bellied salamanders(Desmognathus quadramaculatus complex). Canadian Jour-nal of Zoology 78:1712–1722.

Carr, J. A. 2010. Stress and reproduction in amphibians, p.99–116. In: Hormones and Reproduction of Vertebrates.Vol. 2. D. O. Norris and K. H. Lopez (eds.). Elsevier, SanDiego.

Carr, J. A., and D. O. Norris. 1988. Interrenal activity duringmetamorphosis of the Tiger Salamander, Ambystomatigrinum. General and Comparative Endocrinology 71:63–69.

Chambers, D. L., J. M. Wojdak, P. Du, and L. K. Belden.2011. Corticosterone level changes throughout larvaldevelopment in the amphibians Rana sylvatica andAmbystoma jeffersonianum reared under laboratory, meso-cosm, or free-living conditions. Copeia 2011:530–538.

Chippindale, P. T., R. M. Bonett, A. S. Baldwin, and J. J.Wiens. 2004. Phylogenetic evidence for a major reversal oflife-history evolution in plethodontid salamanders. Evolu-tion 58:2809–2822.

Chippindale, P. T., A. H. Price, J. J. Wiens, and D. M. Hillis.2000. Phylogenetic relationships and systematic revisionof central Texas hemidactyliine plethodontid salamanders.Herpetological Monographs 14:1–80.

Coddington, E., and F. L. Moore. 2003. Neuroendocrinol-ogy of context dependent stress responses: vasotocin altersthe effect of corticosterone on amphibian behaviors.Hormones and Behavior 43:222–228.

Collins, J. P. 1981. Distribution, habitats, and life historyvariation in the tiger salamander, Ambystoma tigrinum, ineast-central and southeast Arizona. Copeia 1981:666–675.

Crespi, E. J., T. D. Williams, T. S. Jessop, and B. Delehanty.2013. Life history and the ecology of stress: How doglucocorticoid hormones influence life-history variation inanimals? Functional Ecology 27:93–106.

Danstedt, R. T. 1975. Local geographic variation in demo-graphic parameters and body size of Desmognathus fuscus(Amphibia: Plethodontidae). Ecology 56:1054–1067.

Darras, V. M., and E. R. Kuhn. 1983. Effects of TRH, bovineTSH, and pituitary extracts on thyroidal T4 release in

Bonett—Endocrinology of salamander metamorphosis 217

Ambystoma mexicanum. General and Comparative Endo-crinology 51:286–291.

Das, B., L. Q. Cai, M. G. Carter, Y. L. Piao, A. A. Sharov, M.S. H. Ko, and D. D. Brown. 2006. Gene expressionchanges at metamorphosis induced by thyroid hormonein Xenopus laevis tadpoles. Developmental Biology 291:342–355.

Das, B., R. A. Heimeier, D. R. Buchholz, and Y-B. Shi. 2009.Identification of direct thyroid hormone response genesreveals the earliest gene regulation programs during frogmetamorphosis. Journal of Biological Chemistry 284:34167–34178.

Denoel, M., R. Duguet, G. Dzukic, M. L. Kalezic, and S.Mazzotti. 2001. Biogeography and ecology of paedomor-phosis in Triturus alpestris (Amphibia, Caudata). Journal ofBiogeography 28:1271–1280.

Denoel, M., and P. Joly. 2000. Neoteny and progenesis astwo heterochronic processes involved in paedomorphosisin Triturus alpestris (Amphibia: Caudata). Proceedings ofthe Royal Society B 267:1481–1485.

Dent, J. N. 1942. The embryonic development of Plethodoncinereus as correlated with the differentiation and func-tioning of the thyroid gland. Journal of Morphology 71:577–501.

Dent, J. N. 1968. Survey of amphibian metamorphosis, p.271–311. In: Metamorphosis, A Problem in DevelopmentalBiology. W. Etkin and L. I. Gilbert (eds.). Appleton-Century-Crofts, New York.

Dent, J. N., and J. S. Kirby-Smith. 1963. Metamorphicphysiology and morphology of the Cave SalamanderGyrinophilus palleucus. Copeia 1963:119–130.

Dent, J. N., and W. G. Lynn. 1958. A comparison of theeffects of goitrogens on thyroid activity in Triturusviridescens and Desmognathus fuscus. The Biological Bulletin115:411–420.

Denver, R. J. 1988. Several hypothalamic peptides stimulatein vitro thyrotropin secretion by pituitaries of anuranamphibians. General and Comparative Endocrinology 72:383–393.

Denver, R. J. 1993. Acceleration of anuran amphibianmetamorphosis by corticotropin-releasing hormone-likepeptides. General and Comparative Endocrinology 91:38–51.

Denver, R. J. 1997. Environmental stress as a developmentalcue: corticotropin releasing hormone is a proximatemediator of adaptive phenotypic plasticity in amphibianmetamorphosis. Hormones and Behavior 31:169–179.

Denver, R. J. 2009. Stress hormones mediate environment-genotype interactions during amphibian development.General and Comparative Endocrinology 164:20–31.

Denver, R. J. 2013. Neuroendocrinology of amphibianmetamorphosis, p. 195–227. In: Current Topics in Devel-opmental Biology. Vol. 103. Y-B. Shi (ed.). Academic Press,New York.

Denver, R. J., and P. Licht. 1989. Neuropeptide stimulationof thyrotropin secretion in the larval bullfrog: evidence fora common neuroregulator of thyroid and interrenalactivity in metamorphosis. Journal of Experimental Zool-ogy 252:101–104.

Denver, R. J., N. Mirhadi, and M. Phillips. 1998. Adaptiveplasticity in amphibian metamorphosis: response ofScaphiopus hammondii tadpoles to habitat desiccation.Ecology 79:1859–1872.

Denver, R. J., S. Pavgi, and Y-B. Shi. 1997. Thyroidhormone-dependent gene expression program for Xenopus

neural development. Journal of Biological Chemistry 272:8179–8188.

Dodd, M. H. I., and J. M. Dodd. 1976. The biology ofmetamorphosis, p. 467–599. In: Physiology of the Am-phibia. Vol. 3. B. Lofts (ed.). Academic Press, New York.

Duarte-Guterman, P., L. Navarro-Martın, and V. L. Tru-deau. 2014. Mechanisms of crosstalk between endocrinesystems: regulation of sex steroid hormone synthesis andaction by thyroid hormones. General and ComparativeEndocrinology 203:69–85.

Dundee, H. A. 1957. Partial metamorphosis induced inTyphlomolge rathbuni. Copeia 1957:52–53.

Dundee, H. A. 1962. Response of the neotenic salamanderHaideotriton wallacei to a metamorphic agent. Science 135:1060–1061.

Dundee, H., and A. Gorbman. 1960. Utilization of radioio-dine by thyroid of neotenic salamander, Eurycea tynerensisMoore and Hughes. Physiological Zoology 33:58–63.

Elinson, R. P. 2013. Metamorphosis in a frog that does nothave a tadpole, p. 259–276. In: Current Topics inDevelopmental Biology. Vol. 103. Y-B. Shi (ed.). AcademicPress, New York.

Emel, S. L., and R. M. Bonett. 2011. Considering alternativelife history modes and genetic divergence in conservation:a case study of the Oklahoma salamander. ConservationGenetics 12:1243–1259.

Etkin, W. 1935. The mechanism of anuran metamorphosis.I. Thyroxine concentration and the metamorphic pattern.Journal of Experimental Zoology 71:317–340.

Flood, D. E. K., J. I. Fernandino, and V. S. Langlois. 2013.Thyroid hormones in male reproductive development:evidence for direct crosstalk between the androgen andthyroid hormone axes. General and Comparative Endocri-nology 192:2–14.

Fraker, M. E., V. Cuddapah, S. A. McCollum, F. Hu, R.Relyea, J. Hempel, and R. J. Denver. 2009. Characteriza-tion of an alarm pheromone secreted by amphibiantadpoles that induces behavioral inhibition and suppres-sion of the neuroendocrine stress axis. Hormones andBehavior 55:520–529.

Freeman, S. L., and R. C. Bruce. 2001. Larval period andmetamorphosis of the three-lined salamander, Euryceaguttolineata (Amphibia: Plethodontidae), in the ChattoogaRiver watershed. American Midland Naturalist 145:194–200.

Frieden, E., and B. Naile. 1955. Biochemistry of amphibianmetamorphosis: I. Enhancement of induced metamorpho-sis by glucocorticoids. Science 121:37–39.

Glennemeier, K. A., and R. J. Denver. 2002. Small changesin whole-body corticosterone content affect larval Ranapipiens fitness components. General and ComparativeEndocrinology 127:16–25.

Gray, K. M., and P. A. Janssens. 1990. Gonadal hormonesinhibit the induction of metamorphosis by thyroidhormones in Xenopus laevis tadpoles in vivo, but not invitro. General and Comparative Endocrinology 77:202–211.

Gudernatsch, J. F. 1912. Feeding experiments on tadpoles. I.The influence of specific organs given as food on growthand differentiation. A contribution to the knowledge oforgans with internal secretion. Wilhelm Roux’ Archiv furEntwicklungsmechanik der Organismen 35:457–483.

Hanaoka, Y. 1966. The effect of thyroxine on the metamor-phosis in hypophysectomized and in thyroidectomized

218 Copeia 104, No. 1, 2016

frog tadpoles (Rana ornativentris). Journal of the Faculty ofScience, Hokkaido University 6:98–105.

Hanken, J. 1992. Life history and morphological evolution.Journal of Evolutionary Biology 5:549–557.

Hanken, J., D. H. Jennings, and L. Olsson. 1997. Mecha-nistic basis of life-history evolution in anuran amphibians:direct development. American Zoologist 37:160–171.

Hayes, T. B. 1997. Steroids as potential modulators of thyroidhormone activity in anuran metamorphosis. AmericanZoologist 37:185–194.

Hickerson, C-A. M., E. L. Barker, and C. K. Beachy. 2005.Determinants of metamorphic timing in the blackbelliedsalamander, Desmognathus quadramaculatus. SoutheasternNaturalist 4:33–50.

Hogan, N. S., P. Duarte, M. G. Wade, D. R. Lean, and V. L.Trudeau. 2008. Estrogenic exposure affects metamorphosisand alters sex ratios in the northern leopard frog (Ranapipiens): identifying critically vulnerable periods of devel-opment. General and Comparative Endocrinology 156:515–523.

Hollar, A. R., J. Choi, A. T. Grimm, and D. R. Buchholz.2011. Higher thyroid hormone receptor expression corre-lates with short larval periods in spadefoot toads andincreases metamorphic rate. General and ComparativeEndocrinology 173:190–198.

Hoskins, E. R., and M. M. Hoskins. 1919. Growth anddevelopment of Amphibia as affected by thyroidectomy.Journal of Experimental Zoology 29:1–69.

Jacobs, G. F., R. P. Michielsen, and E. R. Kuhn. 1988.Thyroxine and triiodothyronine in plasma and thyroids ofthe neotenic and metamorphosed axolotl Ambystomamexicanum: influence of TRH injections. General andComparative Endocrinology 70:145–151.

Jennings, D. H., and J. Hanken. 1998. Mechanistic basis oflife history evolution in anuran amphibians: thyroid glanddevelopment in the direct-developing frog, Eleutherodacty-lus coqui. General and Comparative Endocrinology 111:225–232.

Johnson, C. K., and S. R. Voss. 2013. Salamander paedo-morphosis: linking thyroid hormone to life history and lifecycle evolution, p. 229–258. In: Current Topics in Devel-opmental Biology. Vol. 103. Y-B. Shi (ed.). Academic Press,New York.

Kanki, K., and M. Wakahara. 1999. Precocious testiculargrowth in metamorphosis-arrested larvae of a salamanderHynobius retardatus: role of thyroid-stimulating hormone.Journal of Experimental Zoology 283:548–558.

Kerney, R. R., D. C. Blackburn, H. Muller, and J. Hanken.2012. Do larval traits re-evolve? Evidence from theembryogenesis of a direct-developing salamander, Pletho-don cinereus. Evolution 66:252–262.

Kezer, J. 1952. Thyroxin-induced metamorphosis of theneotenic salamanders Eurycea tynerensis and Eurycea neo-tenes. Copeia 1952:234–237.

Kikuyama, S., K. Niki, M. Mayumi, R. Shibayama, M.Nishikawa, and N. Shintake. 1983. Studies on corticoidaction on the toad tadpole tail in vitro. General andComparative Endocrinology 52:395–399.

Kobayashi, H. 1958. Effect of desoxycorticosterone acetateon metamorphosis induced by thyroxine in anurantadpoles. Endocrinology 62:371–377.

Kobayashi, H., and A. Gorbman. 1962. Thyroid function inAmphiuma. General and Comparative Endocrinology 2:279–282.

Kollros, J. J. 1961. Mechanisms of amphibian metamorpho-sis: hormones. American Zoologist 1:107–114.

Krug, E. C., K. V. Honn, J. Battista, and C. S. Nicoll. 1983.Corticosteroids in serum of Rana catesbeiana duringdevelopment and metamorphosis. General and Compara-tive Endocrinology 52:232–241.

Kuhn, E. R., B. De Groef, S. Van der Geyten, and V. M.Darras. 2005. Corticotropin-releasing hormone-mediatedmetamorphosis in the neotenic axolotl Ambystoma mex-icanum: synergistic involvement of thyroxine and corti-coids on brain type II deiodinase. General andComparative Endocrinology 143:75–81.

Kuhn, E. R., B. De Groef, S. V. H. Grommen, S. Van derGeyten, and V. M. Darras. 2004. Low submetamorphicdoses of dexamethasone and thyroxine induce completemetamorphosis in the axolotl (Ambystoma mexicanum)when injected together. General and Comparative Endo-crinology 137:141–147.

Kulkarni, S. S., and D. R. Buchholz. 2012. Beyond synergy:corticosterone and thyroid hormone have numerousinteraction effects on gene regulation in Xenopus tropicalistadpoles. Endocrinology 153:5309–5324.

Kulkarni, S. S., S. Singamsetty, and D. R. Buchholz. 2010.Corticotropin-releasing factor regulates the developmentin the direct developing frog, Eleutherodactylus coqui.General and Comparative Endocrinology 169:225–230.

Lynn, W. G. 1936. A study of the thyroid in embryos ofEleutherodactylus nubicola. The Anatomical Record 64:525–539.

Lynn, W. G. 1942. The embryology of Eleutherodactylusnubicola, an anuran which has no tadpole stage. Contri-butions to Embryology of the Carnegie Institution ofWashington 541:27–62.

Machuca, I., and J. R. Tata. 1992. Autoinduction of thyroidhormone receptor during metamorphosis is reproduced inXenopus XTC-2 cells. Molecular and Cellular Endocrinolo-gy 87:105–113.

Manzon, R. G., and R. J. Denver. 2004. Regulation ofpituitary thyrotropin gene expression during Xenopusmetamorphosis: negative feedback is functional through-out metamorphosis. Journal of Endocrinology 182:273–285.

Marks, S. B., and A. Collazo. 1998. Direct development inDesmognathus aeneus (Caudata: Plethodontidae): a stagingtable. Copeia 1998:637–648.

McEntire, K. P., T. W. Pierson, and J. C. Maerz. 2014.Eurycea cirrigera (Southern two-lined salamander) paedo-genesis. Herpetological Review 45:473.

Middlemis Maher, J., E. E. Werner, and R. J. Denver. 2013.Stress hormones mediate predator-induced phenotypicplasticity in amphibian tadpoles. Proceedings of the RoyalSociety B 280:20123075.

Moriya, T. 1983. The effects of temperature on the action ofthyroid hormone and prolactin in larvae of the salamanderHynobius retardatus. General and Comparative Endocrinol-ogy 49:1–7.

Niemiller, M. L., B. M. Fitzpatrick, and B. T. Miller. 2008.Recent divergence with gene-flow in Tennessee cavesalamanders (Plethodontidae: Gyrinophilus) inferred fromgene genealogies. Molecular Ecology 17:2258–2275.

Noble, G. K. 1924. The ‘retrograde metamorphosis’ of theSirenidae; experiments on the functional activity of thethyroid of the perennibranchs. The Anatomical Record 29:100.

Bonett—Endocrinology of salamander metamorphosis 219

Noble, G. K., and G. Evans. 1932. Observations andexperiments on the life history of the salamander,Desmognathus fuscus fuscus (Rafinesque). American Muse-um Novitates, Number 533, American Museum of NaturalHistory, New York.

Noble, G. K., and E. J. Farris. 1929. A metamorphic changeproduced in Cryptobranchus by thyroid solutions. TheAnatomical Record 42:59.

Norman, M. F., J. A. Carr, and D. O. Norris. 1987.Adenohypophysial-thyroid activity of the tiger salaman-der, Ambystoma tigrinum, as a function of metamorphosisand captivity. Journal of Experimental Zoology 242:55–66.

Norman, M. F., and D. O. Norris. 1987. Effects ofmetamorphosis and captivity on the in vitro sensitivityof thyroid glands from the tiger salamander, Ambystomatigrinum, to bovine thyrotropin. General and ComparativeEndocrinology 242:77–84.

Norris, D. O., R. E. Jones, and D. C. Cohen. 1973. Effects ofmammalian gonadotropins (LH, FSH, hCG) and gonadalsteroids on TSH-induced metamorphosis of Ambystomatigrinum (Amphibia: Caudata). General and ComparativeEndocrinology 20:467–473.

Nussbaum, R. A., and G. W. Clothier. 1973. Populationstructure, growth, and size of larval Dicamptodon ensatus(Eschscholtz). Northwest Science 47:218–227.

Okada, R., M. F. Miller, K. Yamamoto, B. De Groef, R. J.Denver, and S. Kikuyama. 2007. Involvement of thecorticotropin-releasing factor (CRF) type 2 receptor in CRF-induced thyrotropin release by the amphibian pituitarygland. General and Comparative Endocrinology 150:437–444.

O’Laughlin, B. E., and R. N. Harris. 2000. Models ofmetamorphic timing: an experimental evaluation withthe pond-dwelling salamander Hemidactylium scutatum(Caudata: Plethodontidae). Oecologia 124:343–350.

Page, R. B., J. R. Monaghan, J. A. Walker, and S. R. Voss.2009. A model of transcriptional and morphologicalchanges during thyroid hormone-induced metamorphosisof the axolotl. General and Comparative Endocrinology162:219–232.

Page, R. B., S. R. Voss, A. K. Samuels, J. J. Smith, S. Putta,and C. K. Beachy. 2008. Effect of thyroid hormoneconcentration on the transcriptional response underlyinginduced metamorphosis in the Mexican Axolotl (Ambysto-ma). BMC Genomics 9:78.

Paul, B. D., L. Fu, D. R. Buchholz, and Y-B. Shi. 2005.Coactivator recruitment is essential for liganded thyroidhormone receptor to initiate amphibian metamorphosis.Molecular and Cellular Biology 25:5712–5724.

Petranka, J. W. 1998. Salamanders of the United States andCanada. Smithsonian Institution Press, Washington, D.C.

Ranjan, M., J. Wong, and Y-B. Shi. 1994. Transcriptionalrepression of Xenopus TRB gene is mediated by a thyroidhormone response element located near the start site.Journal of Biological Chemistry 269:24699–25705.

Reedy, A., A. Edwards, C. Pendlebury, L. Murdaugh, R.Avery, J. Seidenberg, A. Aspbury, and C. Gabor. 2014. Anacute increase in the stress hormone corticosterone isassociated with mating behavior in both male and femalered-spotted newts, Notophthalmus viridescens. General andComparative Endocrinology 208:57–63.

Regard, E., A. Taurog, and T. Nakashima. 1978. Plasmathyroxine and triiodothyronine levels in spontaneouslymetamorphosing Rana catesbeiana tadpoles and in adultanuran amphibia. Endocrinology 102:674–684.

Rose, C. S. 1995a. Skeletal morphogenesis in the urodeleskull: II. Effect of developmental stage in TH-inducedremodelling. Journal of Morphology 223:149–166.

Rose, C. S. 1995b. Skeletal morphogenesis in the urodeleskull: III. Effect of hormone dosage in TH-inducedremodelling. Journal of Morphology 223:243–261.

Rose, C. S. 1996. An endocrine-based model for thedevelopmental and morphogenetic diversification in meta-morphic and paedomorphic urodeles. Journal of Zoology239:253–284.

Rose, C. S. 2005. Integrating ecology and developmentalbiology to explain the timing of frog metamorphosis.Trends in Ecology & Evolution 20:129–135.

Rot-Nikcevic, I., and R. J. Wassersug. 2004. Arresteddevelopment in Xenopus laevis tadpoles: how size con-strains metamorphosis. Journal of Experimental Biology207:2133–2145.

Ryan, T. J., and R. C. Bruce. 2000. Life history evolution andadaptive radiation of hemidactyliine salamanders, p. 303–325. In: The Biology of Plethodontid Salamanders. R. C.Bruce, R. G. Jaeger, and L. D. Houck (eds.). KluwerAcademic/Plenum Publishers, New York.

Ryan, T. J., and R. D. Semlitsch. 1998. Intraspecificheterochrony and life history evolution: decouplingsomatic and sexual development in a facultatively paedo-morphic salamander. Proceedings of the National Academyof Sciences of the United States of America 95:5643–5648.

Safi, R., V. Vlaeminck-Guillem, M. Duffraisse, I. Seugnet,M. Plateroti, A. Margotat, M. Duterque-Coquillaud, E. J.Crespi, R. J. Denver, B. Demeneix, and V. Laudet. 2006.Pedomorphosis revisited: thyroid hormone receptors arefunctional in Necturus maculosus. Evolution and Develop-ment 8:284–292.

Sasaki, M. 1924. On a Japanese salamander, in LakeKuttarush, which propagates like the axolotl. Journal ofthe College of Agriculture, Hokkaido Imperial University,Sapporo 15:1–36.

Sato, Y., D. R. Buchholz, B. D. Paul, and Y-B. Shi. 2007. Arole of unliganded thyroid hormone receptor in postem-bryonic development in Xenopus laevis. Mechanisms ofDevelopment 124:476–488.

Semlitsch, R. D. 1985. Reproductive strategy of a faculta-tively paedomorphic salamander Ambystoma talpoideum.Oecologia 65:305–313.

St. Germain, D. L., V. A. Galton, and A. Hernandez. 2009.Minireview: defining the roles of the iodothyroninedeiodinases: current concepts and challenges. Endocrinol-ogy 150:1097–1107.

Stearns, S. 1992. The Evolution of Life Histories. OxfordUniversity Press, Oxford, U.K.

Steffen, M. A., K. J. Irwin, A. L. Blair, and R. M. Bonett.2014. Larval masquerade: a new species of paedomorphicsalamander (Caudata: Plethodontidae: Eurycea) from theOuachita Mountains of North America. Zootaxa 3786:423–442.

Suzuki, S., and M. Suzuki. 1981. Changes in thyroidal andplasma iodine compounds during and after metamorpho-sis of the bullfrog, Rana catesbeiana. General and Compar-ative Endocrinology 45:74–81.

Svob, M., A. Musafija, F. Frank, N. Durovic, T. Svob, S.Cuckovic, and D. Hlaca. 1973. Response of tail fin ofProteus anguinus to thyroxine. Journal of ExperimentalZoology 184:341–344.

220 Copeia 104, No. 1, 2016

Swingle, W. W. 1922. Experiments on the metamorphosis ofneotenous amphibians. Journal of Experimental Zoology36:397–405.

Vlaeminck-Guillem, V., R. Safi, P. Guillem, E. Leteurtre, M.Duterque-Coquillaud, and V. Laudet. 2006. Thyroidhormone receptor expression in the obligatory paedomor-phic salamander Necturus maculosus. International Journalof Developmental Biology 50:553–560.

Voss, S. R. 1993. The relationship between stream order andlength of larval period in the salamander Eurycea wilderae.Copeia 1993:736–742.

Voss, S. R., D. K. Kump, J. A. Walker, H. B. Shaffer, and G. J.Voss. 2012. Thyroid hormone responsive QTL and theevolution of paedomorphic salamanders. Heredity 109:293–298.

Voss, S. R., and H. B. Shaffer. 1997. Adaptive evolution via amajor gene effect: paedomorphosis in the MexicanAxototl. Proceedings of the National Academy of Sciencesof the United States of America 94:14185–14189.

Voss, S. R., and J. J. Smith. 2005. Evolution of salamanderlife cycles: a major effect quantitative trait locus contrib-

utes to discrete and continuous variation for metamorphictiming. Genetics 170:275–281.

Wakahara, M. 1994. Spermatogenesis is extraordinarilyaccelerated in metamorphosis-arrested larvae of a salaman-der, Hynobius retardatus. Experientia 50:94–98.

Wake, D. B. 1966. Comparative osteology and evolution ofthe lungless salamanders, family Plethodontidae. Memoirsof the California Academy of Sciences 4:1–111.

Wake, D. B., and J. Hanken. 1996. Direct development inthe lungless salamanders: What are the consequences fordevelopmental biology, evolution and phylogenesis? In-ternational Journal of Developmental Biology 40:859–869.

Wilder, I. E. 1925. The Morphology of Amphibian Meta-morphosis. Smith College Press, Northampton, Massachu-setts.

Yamaguchi, M., S. Tanaka, and M. Wakahara. 1996.Immunohisto- and immunocytochemical studies on thedynamics of TSH and GTH cells in normally metamor-phosing, metamorphosed, and metamorphosis-arrestedHynobius retardatus. General and Comparative Endocrinol-ogy 104:273–283.

Bonett—Endocrinology of salamander metamorphosis 221