Development, Surface Exposure, and Embryo Behavior Affect Oxygen Levels in Eggs of the Red‐Eyed Treefrog,Agalychnis callidryas

Egg-Clutch Biomechanics Affect Escape-Hatching Behavior and Performance

Arboreal embryos of phyllomedusine treefrogs hatch prematurely to escape snake predation, cued by vibrations in their egg clutches during attacks. However, escape success varies between species, from ∼77% in Agalychnis callidryas to just ∼9% in A. spurrelli at 1 day premature. Both species begin responding to snake attacks at similar developmental stages, when vestibular mechanosensory function begins, suggesting that sensory ability does not limit the hatching response in A. spurrelli. Agalychnis callidryas clutches are thick and gelatinous, while A. spurrelli clutches are thinner and stiffer. We hypothesized that this structural difference alters the egg motion excited by attacks. Since vibrations excited by snakes must propagate through clutches to reach embryos, we hypothesized that the species difference in attack-induced hatching may reflect effects of clutch biomechanics on the cues available to embryos. Mechanics predicts that thinner, stiffer structures have higher free vibration frequencies, greater spatial attenuation, and faster vibration damping than thicker, more flexible structures. We assessed clutch biomechanics by embedding small accelerometers in clutches of both species and recording vibrations during standardized excitation tests at two distances from the accelerometer. Analyses of recorded vibrations showed that A. spurrelli clutches have higher free vibration frequencies and greater vibration damping than A. callidryas clutches. Higher frequencies elicit less hatching in A. callidryas, and greater damping could reduce the amount of vibration embryos can perceive. To directly test if clutch structure affects escape success in snake attacks, we transplanted A. spurrelli eggs into A. callidryas clutches and compared their escape rates with untransplanted, age-matched conspecific controls. We also performed reciprocal transplantation of eggs between pairs of A. callidryas clutches as a method control. Transplanting A. spurrelli embryos into A. callidryas clutches nearly tripled their escape success (44%) compared to conspecific controls (15%), whereas transplanting A. callidryas embryos into different A. callidryas clutches only increased escape success by 10%. At hatching competence, A. callidryas eggs are no longer jelly-encapsulated, while A. spurrelli eggs retain their jelly coat. Therefore, we compared the hatching response and latency of A. spurrelli in de-jellied eggs and their control, jelly-encapsulated siblings using manual egg-jiggling to simulate predation cues. Embryos in de-jellied eggs were more likely to hatch and hatched faster than control siblings. Together, our results suggest that the properties of parentally produced egg-clutch structures, including their vibration biomechanics, constrain the information available to A. spurrelli embryos and contribute to interspecific differences in hatching responses to predator attacks.

What Amphibians Can Teach Us About the Evolution of Parental Care

Parenting is considered a key evolutionary innovation that contributed to the diversification and expansion of vertebrates. However, we know little about how such diversity evolved. Amphibians are an ideal group in which to identify the ecological factors that have facilitated or constrained the evolution of different forms of parental care. Among, but also within, the three amphibian orders-Anura, Caudata, and Gymnophiona-there is a high level of variation in habitat use, fertilization mode, mating systems, and parental sex roles. Recent work using broad phylogenetic, experimental, and physiological approaches has helped to uncover the factors that have selected for the evolution of care and transitions between different forms of parenting. Here, we highlight the exceptional diversity of amphibian parental care, emphasize the unique opportunities this group offers for addressing key questions about the evolution of parenting, and give insights into promising novel directions of research.

The behavioural and physiological ecology of embryos: responding to the challenges of life inside an egg

Adaptations of post-hatching animals have attracted far more study than have embryonic responses to environmental challenges, but recent research suggests that we have underestimated the complexity and flexibility of embryos. We advocate a dynamic view of embryos as organisms capable of responding - on both ecological and evolutionary timescales - to their developmental environments. By viewing embryos in this way, rather than assuming an inability of pre-hatching stages to adapt and respond, we can broaden the ontogenetic breadth of evolutionary and ecological research. Both biotic and abiotic factors affect embryogenesis, and embryos exhibit a broad range of behavioural and physiological responses that enable them to deal with changes in their developmental environments in the course of interactions with their parents, with other embryos, with predators, and with the physical environment. Such plasticity may profoundly affect offspring phenotypes and fitness, and in turn influence the temporal and spatial dynamics of populations and communities. Future research in this field could benefit from an integrated framework that combines multiple approaches (field investigations, manipulative experiments, ecological modelling) to clarify the mechanisms and consequences of embryonic adaptations and plasticity.

Terrestrial reproduction and parental care drive rapid evolution in the trade-off between offspring size and number across amphibians

The trade-off between offspring size and number is central to life history strategies. Both the evolutionary gain of parental care or more favorable habitats for offspring development are predicted to result in fewer, larger offspring. However, despite much research, it remains unclear whether and how different forms of care and habitats drive the evolution of the trade-off. Using data for over 800 amphibian species, we demonstrate that, after controlling for allometry, amphibians with direct development and those that lay eggs in terrestrial environments have larger eggs and smaller clutches, while different care behaviors and adaptations vary in their effects on the trade-off. Specifically, among the 11 care forms we considered at the egg, tadpole and juvenile stage, egg brooding, male egg attendance, and female egg attendance increase egg size; female tadpole attendance and tadpole feeding decrease egg size, while egg brooding, tadpole feeding, male tadpole attendance, and male tadpole transport decrease clutch size. Unlike egg size that shows exceptionally high rates of phenotypic change in just 19 branches of the amphibian phylogeny, clutch size has evolved at exceptionally high rates in 135 branches, indicating episodes of strong selection; egg and tadpole environment, direct development, egg brooding, tadpole feeding, male tadpole attendance, and tadpole transport explain 80% of these events. By explicitly considering diversity in parental care and offspring habitat by stage of offspring development, this study demonstrates that more favorable conditions for offspring development promote the evolution of larger offspring in smaller broods and reveals that the diversity of parental care forms influences the trade-off in more nuanced ways than previously appreciated.

What selective pressures alter the tradeoff between offspring size and number? A phylogenetic comparative approach shows that amphibians with direct development and those that lay eggs in terrestrial environments have larger eggs and smaller clutches, while different care behaviours and adaptations vary in their effects on the tradeoff.

Extreme hypoxia and high lactate concentrations in early chicken embryos show that cutaneous oxygen uptake is limited by diffusion and metabolism is partially anaerobic

Respiratory gas exchange in avian embryos progresses through three stages inside the egg. During the first 3-5 days of incubation, the chicken embryo has no specialised respiratory organs and is not reliant on blood circulation. At this stage, it obtains oxygen mainly by diffusion through the eggshell, albumen, amniotic fluid and embryonic tissues. In the second stage, gas exchange relies on diffusion through the shell in the gas phase and convection by blood circulation through the chorioallantoic membrane and body. Day 19 starts the third stage, the transition from chorioallantoic to pulmonary gas exchange, which is complete when the chick hatches on day 20. Metabolism is thought to be aerobic throughout incubation, although the early embryo is covered by fluids (albumen and amniotic fluid) which would greatly resist oxygen diffusion. This study uses fibre-optic sensors to measure oxygen partial pressure (PO₂) near, and inside of, the embryo during days 3-5, and relates the data to total body lactate levels. The study shows that fluids surrounding the embryo greatly impede oxygen diffusion, with PO₂ becoming severely hypoxic near the embryo, occasionally almost anoxic inside it. Meanwhile, lactate rises to high levels, and the stored lactate can be later oxidised by the embryo when the chorioallantois takes over and metabolism becomes entirely aerobic.

How do red-eyed treefrog embryos sense motion in predator attacks? Assessing the role of vestibular mechanoreception

The widespread ability to alter timing of hatching in response to environmental cues can serve as a defense against threats to eggs. Arboreal embryos of red-eyed treefrogs, Agalychnis callidryas, can hatch up to 30% prematurely to escape predation. This escape-hatching response is cued by physical disturbance of eggs during attacks, including vibrations or motion, and thus depends critically on mechanosensory ability. Predator-induced hatching appears later in development than flooding-induced, hypoxia-cued hatching; thus, its onset is not constrained by the development of hatching ability. It may, instead, reflect the development of mechanosensor function. We hypothesize that vestibular mechanoreception mediates escape-hatching in snake attacks, and that the developmental period when hatching-competent embryos fail to flee from snakes reflects a sensory constraint. We assessed the ontogenetic congruence of escape-hatching responses and an indicator of vestibular function, the vestibulo-ocular reflex (VOR), in three ways. First, we measured VOR in two developmental series of embryos 3-7 days old to compare with the published ontogeny of escape success in attacks. Second, during the period of greatest variation in VOR and escape success, we compared hatching responses and VOR across sibships. Finally, in developmental series, we compared the response of individual embryos to a simulated attack cue with their VOR. The onset of VOR and hatching responses were largely concurrent at all three scales. Moreover, latency to hatch in simulated attacks decreased with increasing VOR. These results are consistent with a key role of the vestibular system in the escape-hatching response of A. callidryas embryos to attacks.

When and where to hatch? Red-eyed treefrog embryos use light cues in two contexts

Hatching timing is under strong selection and environmentally cued in many species. Embryos use multiple sensory modalities to inform hatching timing and many have spontaneous hatching patterns adaptively synchronized to natural cycles. Embryos can also adaptively shift their hatching timing in response to environmental cues indicating immediate threats or opportunities. Such cued shifts in hatching are widespread among amphibians; however, we know little about what, if anything, regulates their spontaneous hatching. Moreover, in addition to selection on hatching timing, embryos may experience benefits or suffer costs due to the spatial orientation of hatching. Amphibian eggs generally lack internal constraints on hatching direction but embryos might, nonetheless, use external cues to inform hatching orientation. The terrestrial embryos of red-eyed treefrogs, Agalychnis callidryas, hatch rapidly and prematurely in response to vibrational cues in egg-predator attacks and hypoxia if flooded. Here we examined A. callidryas’ use of light cues in hatching timing and orientation. To assess patterns of spontaneous hatching and the role of light cues in their diel timing, we recorded hatching times for siblings distributed across three light environments: continuous light, continuous dark, and a 12L:12D photoperiod. Under a natural photoperiod, embryos showed a clear diel pattern of synchronous hatching shortly after nightfall. Hatching was desynchronized in both continuous light and continuous darkness. It was also delayed by continuous light, but not accelerated by continuous dark, suggesting the onset of dark serves as a hatching cue. We examined hatching orientation and light as a potential directional cue for flooded embryos. Embryos flooded in their clutches almost always hatched toward open water, whereas individual eggs flooded in glass cups often failed to do so, suggesting the natural context provides a directional cue. To test if flooded embryos orient hatching toward light, we placed individual eggs in tubes with one end illuminated and the other dark, then flooded them and recorded hatching direction. Most embryos hatched toward the light, suggesting they use light as a directional cue. Our results support that A. callidryas embryos use light cues to inform both when and where to hatch. Both the spatial orientation of hatching and the timing of spontaneous hatching may affect fitness and be informed by cues in a broader range of species than is currently appreciated.

Fish embryos on land: terrestrial embryo deposition lowers oxygen uptake without altering growth or survival in the amphibious fish Kryptolebias marmoratus

Few teleost fishes incubate embryos out of water, but the oxygen-rich terrestrial environment could provide advantages for early growth and development. We tested the hypothesis that embryonic oxygen uptake is limited in aquatic environments relative to air using the self-fertilizing amphibious mangrove rivulus, Kryptolebias marmoratus, which typically inhabits hypoxic, water-filled crab burrows. We found that adult mangrove rivulus released twice as many embryos in terrestrial versus aquatic environments and that air-reared embryos had accelerated developmental rates. Surprisingly, air-reared embryos consumed 44% less oxygen and possessed larger yolk reserves, but attained the same mass, length and chorion thickness. Water-reared embryos moved their opercula ∼2.5 more times per minute compared with air-reared embryos at 7 days post-release, which probably contributed to the higher rates of oxygen uptake and yolk utilization we observed. Genetically identical air- and water-reared embryos from the same parent were raised to maturity, but the embryonic environment did not affect growth, reproduction or emersion ability in adults. Therefore, although aspects of early development were plastic, these early differences were not sustained into adulthood. Kryptolebias marmoratus embryos hatched out of water when exposed to aerial hypoxia. We conclude that exposure to a terrestrial environment reduces the energetic costs of development partly by reducing the necessity of embryonic movements to dispel stagnant boundary layers. Terrestrial incubation of young would be especially beneficial to amphibious fishes that occupy aquatic habitats of poor water quality, assuming low terrestrial predation and desiccation risks.

Polyandry, Predation, and the Evolution of Frog Reproductive Modes

Frog reproductive modes are complex phenotypes that include egg/clutch characteristics, oviposition site, larval development, and sometimes, parental care. Two evident patterns in the evolution of these traits are the higher diversity of reproductive modes in the tropics and the apparent progression from aquatic to terrestrial reproduction, often attributed to higher fitness resulting from decreased predation on terrestrial eggs and tadpoles. Here, we propose that sexual selection-and not only natural selection due to predation-favors terrestrial breeding by reducing the loss of fitness due to polyandry. To examine this novel selective mechanism, we reconstructed the evolution of reproductive diversity in two frog families (Hylidae and Leptodactylidae) and tested for concerted evolution of egg and tadpole development sites with specific mating behaviors. We found that oviposition and tadpole development sites are evolving independently, do not show the same diversity and/or directionality in terms of terrestriality, and thus may be diversifying due to different selective mechanisms. In both families, terrestrial egg deposition is correlated with amplexus that is hidden from competing males, and in hylids, testes mass was significantly larger and more variable in males with exposed amplexus that are vulnerable to polyandry. Our results indicate that intrasexual selection has been an underappreciated mechanism promoting diversification of frog reproductive modes.

External gill motility and striated muscle presence in the embryos of anuran amphibians

Anuran external gills were assessed for motility and striated muscle content in 16 species from seven families. Motility of three kinds was observed. Pulsatory movements related to heart beat rhythm were common. In embryos developing to a late stage before hatching, movements of the whole embryo were frequent, with gills rearranging as a consequence. The only clearly active movement, presumably muscle driven, was 'gill flicking', a posterior movement of the entire gill into the body either on one side only, or both together, followed by a return to the normal spread-out position. Some species may actively spread their gills when hanging from the water surface film, but we did not observe this. In some species, active gill movement developed over time, but we were not able to follow all species over such a developmental sequence. The relationship between active motility and muscle content was good in most cases. Observations on late stage embryos of the aromobatid Mannophryne trinitatis are presented for the first time. In one species, we noted spread external gills being used to adhere hatchlings to a surface.

Phylogenetic analyses reveal unexpected patterns in the evolution of reproductive modes in frogs

Understanding phenotypic diversity requires not only identification of selective factors that favor origins of derived states, but also factors that favor retention of primitive states. Anurans (frogs and toads) exhibit a remarkable diversity of reproductive modes that is unique among terrestrial vertebrates. Here, we analyze the evolution of these modes, using comparative methods on a phylogeny and matched life-history database of 720 species, including most families and modes. As expected, modes with terrestrial eggs and aquatic larvae often precede direct development (terrestrial egg, no tadpole stage), but surprisingly, direct development evolves directly from aquatic breeding nearly as often. Modes with primitive exotrophic larvae (feeding outside the egg) frequently give rise to direct developers, whereas those with nonfeeding larvae (endotrophic) do not. Similarly, modes with eggs and larvae placed in locations protected from aquatic predators evolve frequently but rarely give rise to direct developers. Thus, frogs frequently bypass many seemingly intermediate stages in the evolution of direct development. We also find significant associations between terrestrial reproduction and reduced clutch size, larger egg size, reduced adult size, parental care, and occurrence in wetter and warmer regions. These associations may help explain the widespread retention of aquatic eggs and larvae, and the overall diversity of anuran reproductive modes.

Plasticity of hatching in amphibians: evolution, trade-offs, cues and mechanisms

Many species of frogs and salamanders, in at least 12 families, alter their timing of hatching in response to conditions affecting mortality of eggs or larvae. Some terrestrially laid or stranded embryos wait to hatch until they are submerged in water. Some embryos laid above water accelerate hatching if the eggs are dehydrating; others hatch early if flooded. Embryos can hatch early in response to predators and pathogens of eggs or delay hatching in response to predators of larvae; some species do both. The phylogenetic pattern of environmentally cued hatching suggests that similar responses have evolved convergently in multiple amphibian lineages. The use of similar cues, including hypoxia and physical disturbance, in multiple contexts suggests potential shared mechanisms underlying the capacity of embryos to respond to environmental conditions. Shifts in the timing of hatching often have clear benefits, but we know less about the trade-offs that favor plasticity, the mechanisms that enable it, and its evolutionary history. Some potentially important types of cued hatching, such as those involving embryo-parent interactions, are relatively unexplored. I discuss promising directions for research and the opportunities that the hatching of amphibians offers for integrative studies of the mechanisms, ecology and evolution of a critical transition between life-history stages.

External gills and adaptive embryo behavior facilitate synchronous development and hatching plasticity under respiratory constraint

Plasticity in hatching timing allows embryos to balance egg- and larval-stage risks, and depends on the ability of hatching-competent embryos to continue developing in the egg. Hypoxia can slow development, kill embryos and induce premature hatching. For terrestrial eggs of red-eyed treefrogs, the embryonic period can extend ∼50% longer than development to hatching competence, and development is synchronous across perivitelline oxygen levels(PO2) ranging from 0.5–16.5 kPa. Embryos maintain large external gills until hatching, then gills regress rapidly. We assessed the respiratory value of external gills using gill manipulations and closed-system respirometry. Embryos without external gills were oxygen limited in air and hatched at an external PO2 of 17 kPa, whereas embryos with gills regulated their metabolism and remained in the egg at substantially lower PO2. By contrast,tadpoles gained no respiratory benefit from external gills. We videotaped behavior and manipulated embryos to test if they position gills near the air-exposed portion of the egg surface, where PO2 is highest. Active embryos remained stationary for minutes in gills-at-surface positions. After manipulations and spontaneous movements that positioned gills in the O2-poor region of the egg, however, they returned their gills to the air-exposed surface within seconds. Even neural tube stage embryos, capable only of ciliary rotation, positioned their developing head in the region of highest PO2. Such behavior may be critical both to delay hatching after hatching competence and to obtain sufficient oxygen for normal, synchronous development at earlier stages.

Plasticity in the timing of physiological development: Physiological heterokairy — What is it, how frequent is it, and does it matter?

The study of developmental sequences of physiological traits could be an important way of placing comparative developmental physiology (CDP) within the research agenda being forged by work on developmental plasticity. Here we focus on the concept of heterokairy defined by Spicer & Burggren in 2003 as changes in the timing of physiological development in an individual. The role of this concept in the future of the CDP is discussed. First we provide an historical perspective of the ideas that have led to the investigation of sequences in CDP. This is followed by a re-examination and clarification of the definition of physiological heterokairy before empirical case studies that (explicitly or implicitly) demonstrate physiological heterokairy are reviewed. We suggest that physiological heterokairy can be demonstrated through a wide range of invertebrate and vertebrate examples. However, care must be taken when inferring that heterokairy as a pattern is always the result of heterokairic processes as there is evidence that physiological heterokairy could result from the altered timing of both homologous or analogous physiological mechanisms. We conclude by discussing the potential link between heterokairy and heterochrony and suggest that the investigation of this link should be a major goal for workers in both CDP and developmental plasticity.

Oxygen, gills, and embryo behavior: mechanisms of adaptive plasticity in hatching

Many species alter the timing of hatching in response to egg or larval predators, pathogens, or physical risks. This plasticity depends on separation between the onset of hatching competence and physiological limits to embryonic development. I present a framework based on heterokairy to categorize developmental mechanisms and identify traits contributing to and limiting hatching plasticity, then apply it to a case of predator-induced hatching. Red-eyed treefrogs have arboreal eggs, and tadpoles fall into ponds upon hatching. Egg and tadpole predators select for earlier and later hatching, respectively. Embryos hatch up to 30% early in predator attacks, and later if undisturbed. They maintain large external gills throughout the plastic hatching period, delaying gill regression while development otherwise continues. Rapid gill regression occurs upon hatching. Prolonged embryonic development depends on external gills; inducing gill regression causes hatching. External hypoxia retards development, kills eggs, and induces hatching. Nonetheless, embryos develop synchronously and without hatching prematurely across a broad range of perivitelline PO2, from 0.5-12.5 kPa. Embryos exploit spatial variation of PO2 within eggs by positioning gills against patches of air-exposed surface. Respiratory plasticity and oxygen-sensitive behavior appear critical for the hatching plasticity that balances a predation risk trade-off across life stages.

To hatch and hatch not: similar selective trade-offs but different responses to egg predators in two closely related, syntopic treefrogs

Risk-sensitive hatching is adaptive for species facing a trade-off between egg-stage and post-hatching risks, and environmental variation in one or both stages. Such plasticity has been found in amphibians, fishes, reptiles and spiders, with red-eyed treefrogs (Agalychnis callidryas) being the best-studied case. We assessed hatching plasticity and egg- and larval-stage risks in a closely related, syntopic species, the gliding leaf-frog (Agalychnis spurrelli). We found a lower hatching response to egg-eating snakes in A. spurrelli (9-28% of embryos escaped) than in A. callidryas (59-80% escaped). Levels of snake predation were similarly high for clutches of both species monitored at a pond in Costa Rica, and in fish predation experiments early-hatched A. spurrelli tadpoles were more vulnerable than later hatchlings, as has been shown for A. callidryas. A. spurrelli thus face a risk trade-off similar to A. callidryas, and likely would benefit from predator-induced hatching; their lower responsiveness to snakes appears nonadaptive. A. spurrelli embryos showed a stronger hatching response (57% hatched in 1 h) to submergence underwater than to snake attacks even though submergence is a less frequent risk. This suggests they have a greater capacity for early hatching than is expressed in the context of snake attacks, but have much lower sensitivity to snake cues than to flooding cues. Development in A. spurrelli is accelerated compared to syntopic A. callidryas, and spontaneous hatching is earlier and more synchronous. This is congruent with predictions based on selection by egg predators in the absence of a strong escape hatching response.