Stage dependent plasticity and the evolution of a non-obligate cave dwelling life cycle in frogs

Wei Zhu; Liming Chang; Ningning Lu; Jianping Jiang; Ruoyao Ni; Bin Wang

doi:10.1186/s12915-026-02532-8

. 2026 Feb 4;24:64. doi: 10.1186/s12915-026-02532-8

Stage dependent plasticity and the evolution of a non-obligate cave dwelling life cycle in frogs

Wei Zhu ¹, Liming Chang ¹, Ningning Lu ¹, Jianping Jiang ¹, Ruoyao Ni ^1,^✉, Bin Wang ^1,^✉

PMCID: PMC12964613 PMID: 41634689

Abstract

Background

Plasticity is crucial for environmental adaptation in animals, yet its role in life cycle evolution remains poorly understood. Anurans exhibit a biphasic life cycle, with tadpole and frog following distinct evolutionary trajectories. The non-obligate cave-dwelling frog Oreolalax rhodostigmatus exemplifies this phenomenon: its tadpoles have well adapted to caves, while frogs depend on outside resources. Stage-dependent adaptive plasticity may explain this paradoxical life cycle, predicting that environmental responsiveness declines after metamorphic climax and that early-stage plasticity facilitates adaptations to resource-limited cave environments.

Results

We examined transcriptional plasticity of O. rhodostigmatus by comparing cave- and outside-dwelling tadpoles from the same population across four organs (liver, tail, skin, and hindlimb) at ten developmental stages, spanning pre-metamorphosis to metamorphic climax. Organ transcriptomes reflected both developmental and environmental effects. As predicted, transcriptional responses to environmental conditions in the liver, skin, and tail declined markedly after the onset of metamorphic climax, while the frog-specific organ, hindlimb, showed the weakest environmental responsiveness. Prior to the metamorphic climax, cave-dwelling individuals showed upregulation of fundamental cellular processes (e.g., RNA and protein synthesis) while downregulation of immune-related processes, consistent with a resource-allocation strategy in resource-limited environments. Moreover, increased expression of genes involved in bile biosynthesis and xenobiotic metabolism in the liver suggests enhanced nutrient absorption and utilization, while upregulation of glucagon- and insulin-resistance-related genes in the tail indicates improved tolerance to starvation.

Conclusions

Our findings support the hypothesis that stage-dependent plasticity explains the evolution of non-obligate cave-dwelling in anurans, highlighting the potential role of plasticity in shaping animal lifestyle evolution.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-026-02532-8.

Keywords: Adaptation, Insulin resistance, Starvation resistance, Troglomorphism

Background

Phenotypic plasticity, the ability of an organism to produce different phenotypes in response to environmental changes [1], plays a pivotal role in adaptive evolution [2]. It enables organisms to cope with variable and often unpredictable environments by modifying their morphology, physiology, or behavior without genetic changes [3]. This flexibility provides a short-term survival advantage, especially in habitats subject to fluctuating conditions such as temperature extremes, resource scarcity, or predation pressures [4–6]. Over evolutionary timescales, plasticity can also facilitate the emergence of novel traits and evolutionary innovations by exposing cryptic genetic variation to selection [7]. It serves as a buffer during environmental shifts, potentially reducing selective pressures on populations and allowing time for genetic adaptations to evolve [8]. While the role of plasticity in shaping specific phenotypic traits and life trajectories is increasingly acknowledged [9], whether and how it drives the evolution of animal lifestyles remains largely understudied [10].

Anurans are characterized by a biphasic life cycle, with an aquatic larval (tadpole) stage followed by an adult (frog) stage, both connected through the metamorphic period in which drastic morphological and physiological changes occur [11, 12]. This biphasic life cycle was already present in the early evolutionary history of stem-anurans and has remained stable for at least 161 million years [13]. Adult frogs and tadpoles exhibit uncoupled phenotypic evolution [14], and both of them have experienced independent expansion of morphospaces [15, 16]. Compared to post-metamorphic stages, tadpoles typically exhibit greater adaptive plasticity, encompassing both functional traits (e.g., tail fin depth and skin coloration) and developmental aspects related to life-history transitions [17–19]. The development and growth rates of tadpoles are highly plastic, responding to variations in food availability [20], drought degree [21], temperature [22], and predation pressure [23]. This plasticity allows anurans to optimize survival by adjusting the timing of the water-to-land transition, thereby enabling tadpoles to thrive in extreme environmental conditions. Furthermore, tadpoles exhibit a wide range of feeding strategies, including herbivory, carnivory, and omnivory, demonstrating greater dietary plasticity compared to adults [24–27], which typically have a more limited, predominantly carnivorous diet. The difference in plasticity between tadpoles and post-metamorphic individuals likely influences their environmental adaptability, potentially resulting in varying capacities for adaptation to extreme habitats.

Cave dwelling is an intriguing phenomenon in animals [28]. Vertebrate cave dwellers include cave fish (e.g., Astyanax mexicanus) [29–31], cave salamanders (e.g., Eurycea lucifuga) [32, 33], and the tadpoles of cave frogs (e.g., Oreolalax rhodostigmatus) [34, 35]. As a consequence of long-term adaptation to darkness, cave dwellers exhibit a series of convergent troglomorphic traits (troglobitic lifestyle), such as degenerated eyes, depigmentation, and increased resistance to starvation. Unlike cave fish and salamanders, O. rhodostigmatus does not sustain a fully troglobitic lifestyle throughout its life cycle but rather demonstrates a non-obligate cave-dwelling behavior [36]. During metamorphic climax, the tadpoles undergo rapid eye development and skin pigmentation, enabling a transition to a troglophilic or trogloxenic lifestyle at post-metamorphic stages. This adaptation allows them to venture outside the caves while still returning to breed within them. This raises an intriguing question: why, despite the tadpoles being well-adapted to the cave environment, do the post-metamorphic stages fail to evolve a troglobitic lifestyle? From an ecological standpoint, this phenomenon might be explained by environmental and resource constraints, where a troglophilic or trogloxenic lifestyle offers a selective advantage over a troglobitic one. Alternatively, from the perspective of cave adaptation mechanisms, it is possible that intrinsic factors related to the biphasic life cycle may limit the adaptation of post-metamorphic stages to cave environments.

Adaptive response appears to play a pivotal role in the cave adaptation of O. rhodostigmatus. Unlike cave fish and salamanders, the troglomorphic traits of O. rhodostigmatus tadpoles are not fixed. For example, exposure to light can induce pigmentation and stimulate partial eye development in these tadpoles [35, 36]. Moreover, the gut microbiota of O. rhodostigmatus displays remarkable functional plasticity, adapting to the nutrient-poor cave environment by enhancing cellulose degradation and short-chain fatty acid fermentation—processes critical for survival in caves [34]. These findings highlight adaptive plasticity as a key factor for O. rhodostigmatus to thrive in cave environments. We hypothesize that the reduced environmental responsiveness during the post-metamorphic stage constrains cave adaptation in O. rhodostigmatus, resulting in the stage-dependent cave-dwelling phenomenon.

To test this hypothesis, we examined the development- and environment-dependent gene expression variations of two O. rhodostigmatus tadpole cohorts that share the same genetic background but experience distinct environmental conditions [34]. One cohort is from caves (cave individuals; elevation = 1223 m, collected from > 500 m inside the cave) and displays typical troglomorphic traits, while the other comes from habitats outside the caves (outside individuals; elevation = 1104 m, approximately 700 m from the cave collection site; Fig. 1a) and is characterized by pronounced pigmentation and larger eye diameters. From an ecological perspective, these two cohorts belong to the same population, as their habitats are connected by the water system during the rainy season, with outside individuals having been washed downstream from the upstream cave-dwelling individuals. Both cohorts include tadpoles at various developmental stages (Gosner stages 25–45), encompassing pre-metamorphosis, pro-metamorphosis, and metamorphic climax (Fig. 1a). Notably, the cave-dwelling individuals remain strictly confined to the dark cave environment throughout these stages, providing an ideal comparison to their outside-dwelling counterparts at equivalent developmental stages (Fig. 1b). We propose two predictions: first, for the same organ, environment-induced gene transcriptional responses are more prominent during pre-metamorphic and pro-metamorphic stages than at the metamorphic climax. Moreover, frog-specific organs likely exhibit lower environmentally induced transcriptional responses compared to organs specific to tadpoles or those present in both life stages under similar conditions (Fig. 1c, d). Second, genes exhibiting environmentally induced responses during the tadpole stage are functionally closely associated with cave adaptation. We anticipate that these findings will deepen our understanding of extreme adaptations in amphibians and provide compelling evidence for the critical role of plasticity in shaping the evolution of animal lifestyles.

Fig. 1 — Experimental model and design. a Collection sites of the cave and outside tadpoles. The two sampling sites are approximately 700 m apart, and individuals representing different developmental stages were found at each site. For a more detailed topographic map, please refer to Additional file 1: Figure S1a. b An illustration depicting the environments where cave- and outside-dwelling tadpoles develop. Under normal conditions, *O. rhodostigmatus* deposits eggs inside caves, where tadpoles undergo metamorphosis (typical life history; red solid line), while adults forage outside the caves. Occasionally, seasonal floods wash tadpoles out of the caves, and these individuals complete metamorphosis outside the caves (atypical life history; yellow dashed line). The cave and outside environments differ significantly in light intensity and resource availability. c The four organs studied. We propose that the environmental responsiveness of gene expression varies among organs, with frog-specific organs exhibiting relatively low plasticity. d Main experimental indicators relevant to hypothesis testing. We propose that gene transcriptional responses vary across developmental stages, with tadpoles at metamorphic climax exhibiting reduced plasticity

Results

To verify that the differences between cave and outside individuals reflect plasticity rather than genetic divergence, we examined the genetic backgrounds of the two cohorts. Both nuclear and mitochondrial genes showed very low divergence (overall pi < 0.002; Additional file 1: Figure S1b) [34]. AMOVA indicated a marginally significant differentiation for mitochondrial genes (Fst = 0.0715, p = 0.0467), likely reflecting family structure among tadpoles rather than true population-level divergence, while nuclear genes showed no significant differentiation (Fst = 0.0343, p = 0.1775). Consistently, the NMDS scatter plot based on p-distance showed strong overlap between cave and outside individuals (Additional file 1: Figure S1c), and the STRUCTURE Q-plot revealed no clear genetic structuring (Additional file 1: Figure S1d).

Gene transcriptional variations with development and environment

The gene transcriptional profiles of the four organs were shaped by both developmental stages (Gosner stage) and environment (p < 0.05, two-factor PERMANOVA) (Table 1). Developmental stages had a greater impact on all organs, particularly in the hindlimb, compared to environment (Additional file 1: Figure S2).

Table 1.

Influences of development and environment on the gene transcriptional profile of O. rhodostigmatus tadpoles. The data were analyzed with two-factor PERMANOVA

Organ	Factor	Df	Sum of square	R²	F	p value
Hindlimb	Environment	1	0.130	0.058	5.160	0.004
	Stage	1	0.900	0.400	35.674	< 0.001
	Residual	48	1.210	0.539
	Total	50	2.248	1.000
Liver	Environment	1	0.263	0.091	8.461	< 0.001
	Stage	1	0.325	0.112	10.447	< 0.001
	Residual	73	2.273	0.785
	Total	75	2.897	1.000
Skin	Environment	1	0.114	0.041	3.475	0.004
	Stage	1	0.285	0.102	8.669	< 0.001
	Residual	73	2.400	0.856
	Total	75	2.802	1.000
Tail	Environment	1	0.183	0.052	4.775	0.005
	Stage	1	0.597	0.170	15.589	< 0.001
	Residual	72	2.758	0.785
	Total	74	3.515	1.000

Open in a new tab

Different organs reveal distinct variation patterns with developmental stages (Fig. 2a). By assessing the degree of variation (adjusted p < 0.05, PERMANOVA) in gene transcription profiles between consecutive stages, we can classify the developmental stages of each organ into several primary phases (Figs. 2b and S1). Gene transcription in the hindlimb showed the highest variability to developmental stages, with Gosner stages 34–45 categorized into five phases (S1, 34–37; S2, 38; S3, 39; S4, 40; and S5, 42–45). The skin was the least variable, with Gosner stages 25–45 divided into three phases (S1, 25–38; S2, 39–40; and S3, 42–45). The liver and tail also categorized Gosner stages 42–45 into four phases each (liver, S1, 25–33; S2, 34–37; S3, 38–40; and S4, 42–45; tail, S1, 25; S2, 26–36; S3, 37–40; and S4, 42–45). Accordingly, these four organs exhibited a synchronized transcriptional shift following the onset of metamorphic climax; however, they were notably asynchronous in transcriptional variations during the pro-metamorphic phases. The impact of the environment on transcriptomes was also analyzed. Results indicated that tadpoles from different environments exhibited significant transcriptional differences across organs, especially at stages 39 and 40 (Fig. 2b).

Two-way ANOVA was performed to screen the genes whose transcription varied with developmental phases or the environments (Fig. 2c). Among the four organs analyzed, the hindlimbs exhibited the most pronounced transcriptional variations (DEGs, adjusted p < 0.01 after BH correction), while their environment-dependent variations were minimal. In contrast, the liver and tail displayed the opposite pattern, while the skin demonstrated a comparable number of DEGs associated with both development and environment.

Environment-related gene transcriptional responses of each organ

Liver

The DEGs in the liver were divided into four clusters according to their variation patterns across developmental phases and environments (Fig. 3a, b). DEGs in cluster 1 showed increased transcription as metamorphosis progressed, with several genes being more robustly transcribed in outside individuals compared to cave individuals. These DEGs were primarily associated with oxidative phosphorylation, amino acid biosynthesis, arginine and proline metabolism (including the urea cycle), and porphyrin and chlorophyll metabolism (Fig. 3c, d). DEGs in cluster 2 exhibited decreased transcription as metamorphosis advanced, with lower expression levels observed in cave individuals (Fig. 3b). Key pathways highlighted in this cluster included xenobiotic metabolism by P450, fatty acid and lysine degradation, propanoate metabolism, insulin resistance, and bile acid biosynthesis and secretion (Fig. 3c, d). Notably, bile acid metabolism genes exhibited higher expression levels in cave individuals during pre- and pro-metamorphosis. However, their expression gradually declined in cave individuals, eventually reaching levels comparable to those of outside individuals by stage 40 and the metamorphic climax (Fig. 3e). DEGs in clusters 3 and 4 exhibited environment-dependent transcription. Cluster 3 DEGs were more strongly expressed in outside individuals during pre- and pro-metamorphosis but declined to levels similar to cave individuals after the onset of the metamorphic climax (Fig. 3b). These genes were primarily associated with immune responses, including TNF signaling, NF-kappa B signaling, and antigen processing (Fig. 3c, d). Cluster 4 DEGs were more highly expressed in cave individuals during pre- and pro-metamorphosis and were mainly involved in RNA processing and protein translation, such as RNA polymerase activity, RNA transport, ribosome biogenesis, aminoacyl-tRNA biosynthesis, protein processing, and N-glycan biosynthesis (Fig. 3c, d). Collectively, hepatic protein translation and immune responses showed prominent environmental responsiveness, while energy metabolism, bile acid metabolism, xenobiotic metabolism by P450, the urea cycle, and insulin resistance were influenced by both development and environment during pre- and pro-metamorphosis (Fig. 3d).

Fig. 3 — Gene transcriptional responses in the liver. a Liver-specific developmental phases. b Gene expression pattern of the DEGs. c Network presenting the significantly enriched KEGG pathways (adjusted p value < 0.001) by each DEG cluster. The nodes denote KEGG items, with their size standing for the DEG numbers. KEGG items that shared common DEGs were linked with edges, whose color and width denote the cluster and number of shared DEGs, respectively. The DEGs specific to a certain item were denoted as edges linking to itself. The red dashed box indicates the most representative differential pathways for this organ, and the expression patterns of their DEGs are further illustrated in e. d A four-quadrant coordinate plot, with the horizontal axis representing the degree of change in cellular processes during development and the vertical axis indicating the degree of change in response to the environment. The items in the upper right region (first quadrant) indicate processes that respond to both development and the environment; those in the lower right region (fourth quadrant) vary only with development, and those in the upper left region (second quadrant) respond exclusively to environmental changes; the red and blue colors of the dots indicate higher transcription levels in outside and cave individuals, respectively; the red and blue colors of the text represent increased and decreased transcription levels, respectively, as metamorphosis progresses. e Transcription pattern of DEGs involved in primary bile acid biosynthesis and bile secretion. The smooth line was fitted using the “locally estimated scatterplot smoothing” method, with the 95% confidence interval shown in gray

Tail

Differential analysis of tail transcriptional patterns focused on developmental phases S2–4, as no tail samples were collected from outside individuals at stage 25 (S1) (Fig. 4a). DEGs in the tail were grouped into four clusters, mirroring those identified in the liver, with each pair displaying similar trends. However, tail DEGs associated with developmental progression (clusters 1 and 2) showed limited responsiveness to environmental changes (Fig. 4b). DEGs in cluster 1 exhibited increased transcription with development and were enriched in processes related to protein degradation (e.g., proteasomes, lysosomes, and phagosomes) as well as apoptosis (e.g., apoptosis and endocytosis) (Fig. 4c, d), consistent with tail absorption during the metamorphic climax. In contrast, DEGs in cluster 2 displayed opposite transcriptional trends and were primarily associated with muscle functions (e.g., adrenergic signaling in cardiomyocytes, calcium signaling, and the oxytocin signaling pathway), highlighting the reduced role of the tail in locomotion. Similar to the liver, tail DEGs in clusters 3 and 4 exhibited elevated transcription in outside and cave individuals, respectively, during pre- and pro-metamorphosis. These levels subsequently declined to match those of their counterparts at the metamorphic climax (Fig. 4b). Cluster 3 DEGs were significantly enriched in immune response pathways (e.g., TNF signaling and NOD-like signaling) and cell proliferation processes (e.g., cell cycle and DNA synthesis) (Fig. 4c), suggesting a higher tail growth rate in outside individuals. In contrast, cluster 4 DEGs were associated with protein translation pathways (e.g., RNA transport, spliceosome, and ribosome biosynthesis) as well as glucagon signaling and insulin resistance (Fig. 4e). These genes exhibited higher transcription in the tails of cave individuals, particularly at stages 38–40, when the tadpoles had reached their maximum body size and ceased further growth. Overall, processes such as protein translation, immune responses, cell proliferation, and insulin resistance in the tail demonstrated environmental plasticity during pre- and pro-metamorphosis (Fig. 4d).

Fig. 4 — Gene transcriptional responses in the tail. a Tail-specific developmental phases. b Gene expression pattern of the DEGs. c Network presenting the significantly enriched KEGG pathways (adjusted p value < 0.001) by each DEG cluster. The nodes denote KEGG items, with their size standing for the DEG numbers. KEGG items shared common DEGs were linked with edges, whose color and width denote the cluster and number of shared DEGs, respectively. The DEGs specific to a certain item were denoted as edges linking to itself. The red dashed box indicates the most representative differential pathways for this organ, and the expression patterns of their DEGs are further illustrated in e. d A four-quadrant coordinate plot, with the horizontal axis representing the degree of change in cellular processes during development and the vertical axis indicating the degree of change in response to the environment (refer to the description in Fig. 3d). e Transcription pattern of DEGs involved in insulin resistance and glucagon signaling pathways. The smooth line was fitted using the “locally estimated scatterplot smoothing” method, with the 95% confidence interval shown in gray

Skin

The DEGs in the skin were grouped into four clusters, corresponding sequentially to those in the tail, with each pair exhibiting similar trends (Fig. 5a, b). Cluster 1 DEGs, which increased transcriptionally with development, were primarily involved in substrate and energy metabolism (e.g., thermogenesis, peroxisome, amino acid degradation, and fatty acid biosynthesis), xenobiotic metabolism via P450, vitamin and mineral metabolism (e.g., ubiquinone biosynthesis, nicotinate and nicotinamide metabolism, sulfur metabolism, and mineral absorption), and pigmentation (e.g., tyrosine metabolism) (Fig. 5c, d). The skin also showed transcriptional upregulation of protein synthesis and secretion pathways (e.g., ribosome, protein processing in the ER, N-glycan biosynthesis, and protein export) in both cave and outside individuals after the onset of the metamorphic climax (Fig. 5e), indicating a functional shift in the skin following metamorphosis. In contrast, cluster 2 DEGs, which decreased with development, were associated with endocytosis, insulin resistance, cell state regulation (e.g., pluripotency of stem cells and cellular senescence), and neuron development (e.g., axon guidance and the neurotrophin pathway). Cluster 3 DEGs, transcriptionally downregulated in the skin of cave individuals, were predominantly enriched in immune response pathways. Meanwhile, cluster 4 DEGs, transcriptionally upregulated in cave individuals, were primarily associated with protein translation (e.g., RNA transport, spliceosome, and aminoacyl-tRNA biosynthesis) (Fig. 5c, d). Together, these findings suggest that the immune response and protein translation in the skin exhibit plasticity in response to environmental differences between cave and outside conditions.

Fig. 5 — Gene transcriptional responses in the skin. a Skin-specific developmental phases. b Gene expression pattern of the DEGs. c Network presenting the significantly enriched KEGG pathways (adjusted p value < 0.001) by each DEG cluster. The nodes denote KEGG items, with their size standing for the DEG numbers. KEGG items shared common DEGs were linked with edges, whose color and width denote the cluster and number of shared DEGs, respectively. The DEGs specific to a certain item were denoted as edges linking to itself. The red dashed box indicates the most representative differential pathways for this organ, and the expression patterns of their DEGs are further illustrated in e. d A four-quadrant coordinate plot, with the horizontal axis representing the degree of change in cellular processes during development and the vertical axis indicating the degree of change in response to the environment (refer to the description in Fig. 3d). e Transcription pattern of DEGs involved in protein export. The smooth line was fitted using the “locally estimated scatterplot smoothing” method, with the 95% confidence interval shown in gray

Hindlimb

The DEGs in the hindlimb were also grouped into four major clusters. However, unlike the other three organs, the DEGs in the hindlimb exhibited minimal transcriptional responses to environmental changes, with variations primarily driven by developmental progression (Fig. 6a, b). DEGs in clusters 1 and 4, which showed increased transcription with development, were predominantly involved in substrate and energy metabolism (e.g., thermogenesis, starch and sucrose metabolism, purine metabolism, and fatty acid elongation), muscle contraction (e.g., adrenergic signaling in cardiomyocytes, calcium signaling pathway, and vascular smooth muscle contraction), protein synthesis (e.g., ribosome function and protein processing in the endoplasmic reticulum), and vitamin and cofactor metabolism (e.g., nicotinate and nicotinamide metabolism, as well as porphyrin and chlorophyll metabolism) (Fig. 6c, d). Focusing on DEGs associated with adrenergic signaling in cardiomyocytes, a key pathway involved in muscle function, we observed significant transcriptional upregulation of these genes during the pro-metamorphic stages (e.g., Gosner stages 39–40) (Fig. 6e). These findings suggest that the functionalization of the hindlimb as a locomotor organ is a gradual process, rather than being primarily established at the metamorphic climax. DEGs in clusters 2 and 3, which showed decreased transcription with development, were mainly associated with immune responses (e.g., NOD-like receptor signaling pathway and NF-kappa B signaling pathway), cell proliferation and differentiation (e.g., pluripotency of stem cells, cancer-related pathways, Hedgehog signaling pathway, and Wnt signaling pathway), and RNA processing (e.g., RNA transport, spliceosome, and mRNA surveillance pathway) (Fig. 6c, d).

Fig. 6 — Gene transcriptional responses in the hindlimb. a Hindlimb-specific developmental phases. b Gene expression pattern of the DEGs. c Network presenting the significantly enriched KEGG pathways (adjusted p value < 0.001) by each DEG cluster. The nodes denote KEGG items, with their size standing for the DEG numbers. KEGG items shared common DEGs were linked with edges, whose color and width denote the cluster and number of shared DEGs, respectively. The DEGs specific to a certain item were denoted as edges linking to itself. The red dashed box indicates the most representative differential pathways for this organ, and the expression patterns of their DEGs are further illustrated in e. d A four-quadrant coordinate plot, with the horizontal axis representing the degree of change in cellular processes during development and the vertical axis indicating the degree of change in response to environment (refer to the description in Fig. 3d). e Transcription pattern of DEGs involved in adrenergic signaling in cardiomyocytes. The smooth line was fitted using the “locally estimated scatterplot smoothing” method, with the 95% confidence interval shown in gray

Shared and distinct transcriptional variations across organs

Considering that the DEGs of the liver, tail, and skin can be classified into four shared variation patterns (clusters 1–4), we analyzed the shared and distinct functional changes across the three organs (Fig. 7a–d). The results revealed that development-related functional pathways (highlighted by DEGs of clusters 1 and 2) displayed significant organ-specific heterogeneity, with the skin exhibiting the highest functional diversity (Fig. 7a, b). In contrast, pathways associated with environmental responses (highlighted by DEGs of clusters 3 and 4) were highly similar across organs, particularly those related to immune responses and protein translation (Fig. 7c, d).

Fig. 7 — Similarity and dissimilarity in transcriptional responses among organs. a–d Comparison of transcriptional responses across the liver, tail, and skin, revealing similar variation patterns among DEG clusters. KEGG items are indicated in red and pink for those shared by three and two organs, respectively, while cyan highlights organ-specific items. e Metabolic pathways enriched by genes varying with developmental stages (DEG clusters 1 and 2 in the liver, tail, and skin, and clusters 1–4 in the hindlimb). Red and blue indicate increases and decreases in transcriptional levels during metamorphosis. f Comparison of transcriptional variations in metabolism-related genes among organs, with significant variation (adjusted p < 0.01) observed in at least one organ

Transcriptional shifts in metabolism were a shared feature across all four organs; however, the timing and patterns of these changes varied significantly among them (Fig. 7e). The hindlimb exhibited the most pronounced metabolic reorganization, marked by a comprehensive upregulation of substrate and energy metabolism (e.g., glycolysis/gluconeogenesis, fatty acid metabolism, TCA cycle, and oxidative phosphorylation) throughout metamorphosis (Fig. 7f), reflecting the progressive maturation of its locomotor functions. In contrast, the metabolic shifts in the other three organs were primarily confined to the onset of the metamorphic climax. The liver showed upregulation of oxidative phosphorylation and arginine metabolism (including the urea cycle), accompanied by downregulation of pathways related to fatty acid and lysine degradation. The skin predominantly exhibited upregulation of fatty acid metabolism and oxidative phosphorylation, while the tail was characterized by transcriptional downregulation of lysine degradation.

Discussion

Quantifying the relative environmental adaptive responses of tadpoles and post-metamorphic frogs is challenging due to the substantial differences in their habitats and the environmental pressures they encounter. However, cave adaptation in amphibians offers a compelling model to address this question, as both tadpoles and post-metamorphic juveniles are exposed to the same extreme cave environment. Additionally, the variations in gene expression patterns and functions provide a valuable framework for quantitatively assessing the adaptive plasticity [37].

Our findings demonstrated a general decline in environment-induced variations of organs specific to tadpoles, as well as those present in both life stages, as metamorphosis progressed—most notably after the onset of the metamorphic climax (Fig. 8a). This decline is unlikely to be attributed to environmental shifts, as O. rhodostigmatus tadpoles complete their whole metamorphosis within the caves. Moreover, the hindlimb, a frog-specific organ, showed the lowest environment-related gene transcriptional responses among the four organs, despite they experiencing the same environmental differences during metamorphosis. These findings supported our first hypothesis. Because cave and outside individuals share the same genetic pool, the environment-related transcriptional differences observed at the tadpole stage are more appropriately interpreted as plastic responses rather than intrinsic differentiation. This interpretation is supported by our previous study, in which cave and outside individuals were cohabited in the same tank for 2 months and exposed to different food gradients. In both groups, decreasing food availability led to a reduction in liver size and a corresponding shift in gut microbial communities, with no significant differences between cave and outside individuals [34].

Fig. 8 — Deduced mechanisms underlying the stage-dependent cave adaptation in *O. rhodostigmatus* tadpoles. a A schematic diagram denotes the stage-dependent gene transcriptional responses of the four organs. b A schematic diagram depicting the potential evolution mechanisms of stage-dependent cave-dwelling of *O. rhodostigmatus*. We propose that the strong transcriptional responses during larval stages serve as a prerequisite for their transition from an external habitat to a cave environment. However, the reduced environmental responsiveness following metamorphosis likely limits the ability of froglets and adults to fully adapt to a troglobitic lifestyle. This stage-dependent variation in plasticity drives metamorphosed individuals to leave the cave, resulting in a life cycle characterized by stage-specific cave-dwelling behavior. c A summary of the biological processes that varied with development and environmental conditions at the transcriptional level. The items in the upper right region indicate processes that respond to both development and environment, those in the lower right region vary only with development, and those in the upper left region respond exclusively to environmental changes. d A summary of the biological processes related to resource acquisition and consumption that exhibit environment-related variations. e, f Differences in resource allocation were observed between outside (e) and cave (f) individuals. In the cave environment, resource availability is lower and is potentially accompanied by reduced pathogen abundance. This enables cave tadpoles to maintain their energy budget by reallocating resources from immune-related functions to fundamental cellular processes

From an evolutionary perspective, cave adaptation in the tadpoles of O. rhodostigmatus is unlikely to be an ancestral trait. For example, the closely related species Leptobrachium leishanense (also in the family Megophryidae) inhabits mountain streams, where its tadpoles prefer to hide under leaf litter or within crevices but do not exhibit cave-dwelling behavior or develop troglomorphic traits [38]. The ancestors of O. rhodostigmatus likely had a similar lifestyle. Based on our hypothesis in this study, we speculate that O. rhodostigmatus tadpoles may have gradually occupied cave niches by leveraging their adaptive responses, whereas the post-metamorphic stages failed to fully adapt to the cave environment due to reduced environmental plasticity, resulting in stage-dependent cave-dwelling behavior (Fig. 8b). To further investigate this hypothesis, future studies should explore the gene expression plasticity of additional frog-specific and shared organs, such as the lungs and gut. Controlled laboratory experiments will be essential to validate the differences in the level of transcriptional responses between tadpoles and froglets in response to key environmental factors distinguishing cave and outside habitats, such as nutrient levels.

Our second hypothesis is that genes responsive to environmental changes are associated with cave adaptation. Resource scarcity is one of the most significant challenges faced by cave-dwelling organisms [39]. Enhancing resource utilization and reallocating resource investment to minimize expenditure are two key strategies to cope with this challenge [40, 41]. In terms of resource utilization, our previous research demonstrated that the gut microbiota of O. rhodostigmatus tadpoles can convert polysaccharides such as cellulose into short-chain fatty acids, which are metabolically accessible to the host, thereby partially alleviating resource scarcity [34]. In this study, we observed transcriptional upregulation of bile acid synthesis pathways in the liver of cave-dwelling individuals (Fig. 8c, d). These metabolites, upon being released from the gallbladder into the intestine, facilitate fat absorption and modulate gut microbiota composition. Moreover, the upregulation of liver genes involved in P450 and xenobiotic metabolism aligns with the significant enrichment of pathways related to secondary metabolite degradation (e.g., aromatic compounds and chloroalkane and chloroalkene degradation) and P450 enzymes in the gut metagenome of cave-dwelling individuals [34]. These findings imply an enhanced ability to utilize unconventional resources in cave environments (Fig. 8d). Regarding resource allocation, we found that genes involved in the synthesis of biomacromolecules such as RNA and proteins—fundamental cellular processes—exhibited higher transcription levels in cave-dwelling individuals. In contrast, immune-related genes showed lower transcription levels compared to outside-dwelling groups. Such transcriptional adjustments align with the distinctive cave environment, where reduced primary productivity likely coincides with a lower abundance of environmental microbes, including pathogens [42]. This reduced immune burden may lower the energy demands of O. rhodostigmatus tadpoles. Since biomacromolecule synthesis and immune responses are among the most energy-intensive biological processes [43, 44], the observed transcriptional shifts suggest a significant reallocation of resources, prioritizing metabolic maintenance to adapt to the nutrient-poor cave environment (Fig. 8e, f). Additionally, we observed upregulation of genes associated with glucagon signaling pathways and insulin resistance in the tails (Fig. 8d), the primary locomotory organ of tadpoles. Insulin resistance is a critical physiological adaptation to starvation in the cavefish A. mexicanus [45], linked to enhanced fat storage and reduced metabolic rate [46, 47]. Similarly, the transcriptional upregulation of glucagon signaling and insulin resistance in O. rhodostigmatus tadpoles under low-nutrient conditions may represent potential mechanisms for starvation tolerance (Fig. 8d). Taken together, these results suggest that the transcriptional responses of O. rhodostigmatus tadpoles are closely linked to their adaptation to the nutrient-poor cave environment. Collectively, our findings support both of our hypotheses and suggest that gene expression plasticity at different life stages may underlie the evolution of non-obligate cave-dwelling behavior of anurans.

Another key finding of this study is the developmental heterogeneity and allometry among organs. While the four organs examined exhibited synchronized transcriptional changes following the onset of the metamorphic climax, developmental variations before this event varied substantially among organs. These results suggest that developmental phase delineation should be tailored to each organ of interest. This challenges the traditional classification of developmental stages into three broad phases—pre-metamorphosis, pro-metamorphosis, and metamorphic climax (Fig. 2a)—a framework that, while simplifying experimental complexity, groups stages mainly based on similar morphological traits of individual organs [48–50]. Our findings underscore the value of comparative transcriptomic analyses across consecutive Gosner stages as a more nuanced approach to classifying developmental phases. This method may offer deeper functional insights compared to classifications based solely on morphological traits. For example, we found that core metabolic pathways, including glycolysis, the TCA cycle, and oxidative phosphorylation, are strongly shaped by developmental progression, with developmental influences outweighing environmental factors. These metabolic shifts could serve as robust references for defining developmental phases. Additionally, cellular processes with critical functional implications may serve as indicators of organ-specific developmental stages. In this study, protein modification and secretion pathways were particularly active in the skin, while adrenergic signaling played a significant role in the hindlimb. These insights enhance our understanding of the functional specialization of organs during the amphibian transition from aquatic to terrestrial life and reveal how different organs coordinate to facilitate this transition.

Our findings underscore the importance of considering development-dependent variations at the organ level when investigating the environmental adaptability and evolutionary potential of animals. Such an approach may provide critical insights into the mechanisms that constrain environmental adaptability, offering significant implications for advancing theoretical frameworks in evolutionary adaptation and guiding biodiversity conservation strategies. This study highlights the pivotal role of spatiotemporal plasticity in shaping the evolution of life history strategies and elucidates key processes driving the diversification of life cycle patterns in vertebrates. Future research should investigate why transcriptional responses vary across developmental stages in O. rhodostigmatus tadpoles. This will require precise control of environmental cues and examination of phenotypic responses across gradients, rather than relying solely on cave versus outside comparisons. A detailed analysis of traits such as development, pigmentation, and starvation tolerance will be essential to clarify the nature of the plasticity and to determine whether these traits represent discrete, environmentally induced polyphenism similar to that observed in Spea multiplicate [27].

Conclusions

Our results revealed significant variations in gene transcriptional responses during life stage transitions in anurans, as well as notable differences among organs. These findings could explain the evolution of the non-obligate cave-dwelling lifestyle of O. rhodostigmatus. The transcriptional responses observed in tadpoles reflect their adaptation to cave environments, including resource reallocation to prioritize fundamental cellular processes, upregulation of bile biosynthesis and xenobiotic metabolism to enhance environmental resource utilization, and insulin resistance to improve starvation tolerance. These results deepen our understanding of the mechanisms underlying cave adaptation in vertebrates and highlight the critical role of plasticity in shaping the evolution of life cycle patterns in animals.

Methods

Sample collection and information

The O. rhodostigmatus tadpoles were collected from a stream in a karst cave and outside (downstream) in Tongzi County, Guizhou Province, China (107.05 E 28.50 N), in 2019. The stream water in the cave had a relatively stable temperature of ~ 15 °C. Generally, O. rhodostigmatus tadpoles cannot leave the caves during their aquatic life stages. However, we found a group of O. rhodostigmatus tadpoles inhabiting and foraging in a shallow pool outside this cave. These outside-dwelling tadpoles had pigmented skin throughout their body, providing an opportunity to study the environment-induced transcriptional responses in O. rhodostigmatus. Despite their different appearance, the cave and outside tadpoles are not different in their genetic background for their nested phylogenetic relationship (Additional file 1: Figure S1 and [34]).

At the same collection time, individuals from both cave and outside cohorts were sampled across developmental stages ranging from Gosner 25 to 45, encompassing pre-metamorphosis, pro-metamorphosis, and metamorphic climax (Fig. 1a). The developmental stages of tadpoles were identified following the criterion raised by Gosner [51]. As some similar stages cannot be easily distinguished, we classify the stages into several periods with the following standard: Stage 25, free-swimming tadpoles without hindlimb bud; Stages 26–30, with hindlimb bud of varying length; Stages 31–33, with paddle-shaped hindlimb; Stages 34–36, formation of toe in hindlimb; Stage 37, bending hindlimb with individual toes; Stage 38, retractable hindlimb; Stage 39, with long hindlimb normally in “N” shape; Stage 40, developed and folded hindlimb (“W” shape); Stage 41, observation of forelimb beneath the abdomen skin; Stage 42, emergence of forelimb. Tadpoles in this study are at their Stages 25–42 (n = 76 for individuals, n = 278 for tissue samples). To simplify the experimental and analytic models, the Gosner stages of animals were usually divided into three primary phases according to morphological changes: pre-metamorphosis, pro-metamorphosis, and metamorphic climax. For instance, pre-metamorphosis (stages 25–30 or 25–36) shows no or rudimentary external hind limb buds; pro-metamorphosis (stages 31–41 or 36–41) shows more developed hindlimb; metamorphic climax (stages 42–45) is characterized by external forelimbs [49, 50]. The sample sizes per Gosner stage and group (cave- or outside-dwelling) are detailed in Table 2 and Additional file 2: Table S1.

Table 2.

Numbers of transcriptome samples for each organ across developmental stages. For each organ, every sample represents a distinct individual. In tadpoles at stages earlier than Gosner 34, the hindlimbs were either undeveloped or did not provide sufficient tissue for transcriptome sequencing

Environment	Organ	Gosner stages
Environment	Organ	25	26–30	31–33	34–36	37	38	39	40	42	43–45
Outside	Hindlimb	/	/	/	2	4	5	5	5	4	2
	Liver	1	6	5	3	4	5	5	5	4	2
	Skin	1	6	5	3	4	5	5	5	4	2
	Tail	0	6	5	3	4	5	5	5	4	2
Cave	Hindlimb	/	/	/	5	4	4	5	4	1	1
	Liver	6	5	0	6	4	4	5	4	1	1
	Skin	6	5	0	6	4	4	5	4	1	1
	Tail	6	5	0	6	4	4	5	4	1	1

Open in a new tab

In this study, tadpoles were euthanized using MS-222 and dissected to collect the liver, skin, tail, and hindlimb (Fig. 1c). The liver, shared by both tadpoles and frogs, plays a central role in metabolic regulation and energy storage. The skin, also common to both life stages, functions in defense and sensory perception. The tail, unique to tadpoles, serves as a locomotor organ and additionally acts as an energy reservoir during the metamorphic climax. The hindlimb, in contrast, is a locomotor organ exclusive to frogs. All animal procedures were performed according to protocols approved by the Animal Care Advisory Committee of the Chengdu Institute of Biology, Chinese Academy of Sciences, China (Permit number: CIBACUC2017031009 and CIB20191084).

Genetic background analysis

To illustrate that cave and outside individuals belong to the same genetic pool, we compared the differentiation patterns of both mitochondrial and nuclear genes. The sequence alignments of 13 mitochondrial genes (i.e., cox1, cox2, cox3, cytb, nd1, nd2, nd3, nd4, nd4l, nd5, and nd6) were extracted from the liver transcriptomes of O. rhodostigmatus individuals, while the partial sequences of ten nuclear genes (i.e., evpl, sacs, disp2, bdnf, cxcr4, fat1, ntf3, rag1, ncx1, and sall1) were obtained using PCR [34]. The p-distance, nucleotide diversity (π or π), and differentiation coefficient were calculated using MEGA 12 [52]. The calculation of Fst value and Analysis of Molecular Variance (AMOVA) were performed using pegas [53], a R package. Bayesian clustering analysis was performed using STRUCTURE software [54].

Transcriptomic analyses

The total RNA was extracted using TRIzol reagent (Invitrogen). The RNA quantity, purity, and integrity were assessed using Qubit 2.0 fluorometer, Nanodrop spectrophotometer, and Agilent 2100 Bioanalyzer respectively. The PolyA RNA sequencing libraries were prepared following enrichment, fragmentation, cDNA synthesis, adaptor ligation, and amplification steps. Then, the libraries were sequenced on an Illumina HiSeq 4000 platform (PE150 strategy), producing 6 GB of raw data for each sample. The adapters and low-quality reads (N base > 10%, or low-quality bases > 50%) were removed to obtain the clean data. The clean reads were aligned to the annotated O. rhodostigmatus genome using HISAT2 v2.1.0 [55]. Transcript assembly and expression quantification of coding genes were performed using StringTie v2.0.4 [56]. Gene expression quantification was performed using Bowtie2 [54] and RSEM [57]. Finally, the gene expression matrices, using transcripts per million (TPM) normalization method, were constructed for all the samples across developmental stages and environments. All the subsequent comparative analyses on the gene expression levels were based on these TPM matrices.

The raw sequencing data, including transcriptomes of the liver, skin, hindlimb, and tail, were deposited as part of our previous study [36] in the China National GeneBank DataBase (CNGBdb; https://db.cngb.org/) under accession number CNP0005050 [58]. In that study, we focused solely on the temporal expression variations in the skin, hindlimb, and tail, primarily examining global transcriptional changes around the metamorphic climax and the developmental dynamics of select pigmentation-related genes, with the aim of elucidating the spatiotemporal regulation of the pigmentary system. In contrast, the present study takes a completely different analytical perspective, aiming to uncover the transcriptional responses of different tissues in response to environmental changes across developmental stages.

Developmental phase classification based on transcriptomic similarity

A key challenge in developmental time-series analysis is the potential mismatch between morphologically defined developmental stages and the transcriptional states of specific tissues. Transcriptional shifts may occur asynchronously across tissues, and overly fine-scale staging can lead to false positives or false negatives in differential expression analysis, thereby reducing statistical robustness and obscuring the interpretation of gene expression trajectories. To address these issues, we grouped Gosner stages into a limited number of primary developmental phases for each tissue, based on similarities in their global transcriptional profiles. The classification procedure involved the following steps:

Dissimilarity matrix construction: For each tissue, the TPM expression matrix was transformed into a Bray–Curtis dissimilarity matrix.

Stage-to-stage distance calculation: Pairwise average dissimilarities between developmental stages were calculated. To assess the overall divergence among stages, permutational multivariate analysis of variance (PERMANOVA) was performed using the adonis function in the vegan package [59]. The pairwise dissimilarities were visualized using heatmaps.

Phase delineation: Based on both the magnitude and statistical significance of pairwise dissimilarities between adjacent stages, stages were clustered into broader phases. In principle, stages showing significant transcriptional divergence were not grouped into the same phase.

Validation of phase classification: To evaluate whether phase grouping better captured transcriptional variation than traditional stage-based grouping, we applied a two-way PERMANOVA model, incorporating developmental stage/phase and environment as factors. An increase in the model’s explanatory power when using phase-based classification was taken as evidence for the validity and robustness of the phase definitions.

Statistic analyses

To examine the influences of development (either Gosner stage or developmental phases) and environment on the transcriptomes, the TPM matrices were first transformed into Bray–Curtis distance matrices using vegdist, a function of the vegan package in R [59]. Then, adonis2 from the same package was used to perform PERMANOVA to examine the significance of the two factors on the variations in the distances between samples (“margin” = “by”, permutation = 9999). As the transcriptional variations of most genes did not deviate from normal distribution according to Kolmogorov–Smirnov tests, two-way ANOVA was performed to screen the genes whose transcriptional levels were affected by development, environment, or their interaction (differentially expressed genes, DEGs). Benjamini-Hochberg (BH) corrections were used to adjust the p values. The transcriptional patterns of DEGs were analyzed with timeclust, a function in TCseq packages (https://bioconductor.org/packages/TCseq). KOBAS-i was used for KEGG enrichment analyses on DEGs [60]. The networks were created using Cytoscape 3.5.1 [61]. Graphs in this study were drawn on GraphPad Prism 5 (v 5.01, GraphPad Software, San Diego, USA), ggplot2 [62], and MedPeer (www.medpeer.cn).

Supplementary Information

12915_2026_2532_MOESM1_ESM.docx^{(1MB, docx)}

Additional file 1: Figures S1-S2. FigS1-Genetic background analyses. FigS2- Explanation rate of developmental stages and environment on the variations of gene transcriptional profiles.

12915_2026_2532_MOESM2_ESM.xlsx^{(22.7KB, xlsx)}

Additional file 2: Table S1. Table S1-Sample information of this study.

Acknowledgements

We appreciate Zhonghao Luo, Yan Wang, and Shize Li from Chengdu Institute of Biology, CAS for their help in sample collection.

Authors’ social media handles

BlueSky: @zhuweibio.bsky.social.

BlueSky: @niruoyao.bsky.social.

Abbreviations

DEG: Differentially expressed gene
PERMANOVA: Permutational multivariate analysis of variance
TPM: Transcripts per million

Authors’ contributions

W.Z., B.W., and R.N. initiated and designed this study. W.Z., R.N., L.C., and N.L. processed data in this study. W.Z., R.N., L.C., and N.L. conducted experiments and wrote the manuscript. J.J. and B.W. contributed materials and edited the manuscript. All authors read and approved the final manuscript.

Funding

This study was funded by the Second Tibetan Plateau Scientific Expedition and Research Program (STEP) (Grant No: 2019QZKK05010203), the National Natural Science Foundation of China (Grant Nos: 32070426, 32270498, and 31900327), and the West Light Foundation of the Chinese Academy of Sciences (Grant No: 2021XBZG_XBQNXZ_A_006).

Data availability

The raw sequencing data were deposited in the China National GeneBank DataBase (CNGBdb; (https://db.cngb.org/)) under accession number CNP0005050 [58].

Declarations

Ethics approval and consent to participate

All animal procedures were performed according to protocols approved by the Animal Care Advisory Committee of the Chengdu Institute of Biology, Chinese Academy of Sciences, China (Permit number: CIBACUC2017031009 and CIB20191084).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Ruoyao Ni, Email: niry@cib.ac.cn.

Bin Wang, Email: wangbin@cib.ac.cn.

References

1.West-Eberhard MJ: Plasticity. Developmental Plasticity and Evolution. Edited by West-Eberhard MJ: Oxford University Press; 2003: 0.
2.Thompson JD. Phenotypic plasticity as a component of evolutionary change. Trends Ecol Evol. 1991;6(8):246–9. [DOI] [PubMed] [Google Scholar]
3.Miner BG, Sultan SE, Morgan SG, Padilla DK, Relyea RA. Ecological consequences of phenotypic plasticity. Trends Ecol Evol. 2005;20(12):685–92. [DOI] [PubMed] [Google Scholar]
4.Yampolsky LY, Schaer TMM, Ebert D. Adaptive phenotypic plasticity and local adaptation for temperature tolerance in freshwater zooplankton. Proc R Soc Lond B Biol Sci. 2014;281(1776):20132744. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Rion S, Kawecki TJ. Evolutionary biology of starvation resistance: what we have learned from Drosophila. J Evol Biol. 2007;20(5):1655–64. [DOI] [PubMed] [Google Scholar]
6.Barata C, Baird D, Soares A. Phenotypic plasticity in Daphnia magna Straus: variable maturation instar as an adaptive response to predation pressure. Oecologia. 2001;129(2):220–7. [DOI] [PubMed] [Google Scholar]
7.Agrawal AA. Phenotypic plasticity in the interactions and evolution of species. Science. 2001;294(5541):321–6. [DOI] [PubMed] [Google Scholar]
8.Price TD, Qvarnström A, Irwin DE. The role of phenotypic plasticity in driving genetic evolution. Proc R Soc Lond B Biol Sci. 2003;270(1523):1433–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Pigliucci M. Evolution of phenotypic plasticity: where are we going now? Trends Ecol Evol. 2005;20(9):481–6. [DOI] [PubMed] [Google Scholar]
10.Fabian D, Flatt T. Life history evolution. Nature Education Knowledge. 2012;3(10):24. [Google Scholar]
11.Wassersug RJ. The adaptive significance of the tadpole stage with comments on the maintenance of complex life cycles in anurans. Am Zool. 1975;15(2):405–17. [Google Scholar]
12.McDiarmid RW, Altig R. Tadpoles: the biology of anuran larvae. University of Chicago Press; 1999. [Google Scholar]
13.Chuliver M, Agnolín FL, Scanferla A, Aranciaga Rolando M, Ezcurra MD, Novas FE, et al. The oldest tadpole reveals evolutionary stability of the anuran life cycle. Nature. 2024;636(8041):138–42. [DOI] [PubMed] [Google Scholar]
14.Wollenberg Valero KC, Garcia-Porta J, Rodríguez A, Arias M, Shah A, Randrianiaina RD, et al. Transcriptomic and macroevolutionary evidence for phenotypic uncoupling between frog life history phases. Nat Commun. 2017;8(1):15213. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Sherratt E, Vidal-García M, Anstis M, Keogh JS. Adult frogs and tadpoles have different macroevolutionary patterns across the Australian continent. Nat Ecol Evol. 2017;1(9):1385–91. [DOI] [PubMed] [Google Scholar]
16.Roelants K, Haas A, Bossuyt F. Anuran radiations and the evolution of tadpole morphospace. Proc Natl Acad Sci U S A. 2011;108(21):8731–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Warkentin KM. Adaptive plasticity in hatching age: a response to predation risk trade-offs. Proceed Natl Acad Sci. 1995;92(8):3507–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Wilson RS, Kraft PG, Van Damme R. Predator-specific changes in the morphology and swimming performance of larval Rana lessonae. Funct Ecol. 2005;19(2):238–44. [Google Scholar]
19.Touchon JC, Warkentin KM. Thermally contingent plasticity: temperature alters expression of predator-induced colour and morphology in a Neotropical treefrog tadpole. J Anim Ecol. 2011;80(1):79–88. [DOI] [PubMed] [Google Scholar]
20.Zhu W, Zhang MH, Chang LM, Li C, Zhu WB, Xie F, et al. Characterizing the composition, metabolism and physiological functions of the fatty liver in Rana omeimontis tadpoles. Front Zool. 2019;16:42. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Tejedo M, Reques R. Plasticity in metamorphic traits of natterjack tadpoles: the interactive effects of density and pond duration. Oikos. 1994;71(2):295–304. [Google Scholar]
22.Álvarez D, Nicieza AG. Effects of temperature and food quality on anuran larval growth and metamorphosis. Funct Ecol. 2002;16(5):640–8. [Google Scholar]
23.Relyea RA. Getting out alive: how predators affect the decision to metamorphose. Oecologia. 2007;152(3):389–400. [DOI] [PubMed] [Google Scholar]
24.Carreira BM, Segurado P, Orizaola G, Gonçalves N, Pinto V, Laurila A, et al. Warm vegetarians? Heat waves and diet shifts in tadpoles. Ecology. 2016;97(11):2964–74. [DOI] [PubMed] [Google Scholar]
25.Montaña CG, Silva SDGTM, Hagyari D, Wager J, Tiegs L, Sadeghian C, et al. Revisiting “what do tadpoles really eat?” A 10-year perspective. Freshw Biol. 2019;64(12):2269–82. [Google Scholar]
26.Ramamonjisoa N, Iwai N, Natuhara Y. Post-metamorphic costs of carnivorous diets in an omnivorous tadpole. Copeia. 2016;104(4):808–15. [Google Scholar]
27.Pfennig D. The adaptive significance of an environmentally-cued developmental switch in an anuran tadpole. Oecologia. 1990;85(1):101–7. [DOI] [PubMed] [Google Scholar]
28.Romero A. Cave biology: life in darkness. Cambridge university press; 2009. [Google Scholar]
29.Jeffery WR. Regressive evolution in Astyanax cavefish. Annu Rev Genet. 2009;43(1):25–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Riddle MR, Aspiras AC, Gaudenz K, Peuss R, Sung JY, Martineau B, et al. Insulin resistance in cavefish as an adaptation to a nutrient-limited environment. Nature. 2018;555(7698):647–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Yang J, Chen X, Bai J, Fang D, Qiu Y, Jiang W, et al. The Sinocyclocheilus cavefish genome provides insights into cave adaptation. BMC Biol. 2016;14(1):1. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Salvidio S. Population dynamics and regulation in the cave salamander Speleomantes strinatii. Naturwissenschaften. 2007;94(5):396–400. [DOI] [PubMed] [Google Scholar]
33.Hutchison VH. The distribution and ecology of the cave salamander, Eurycea lucifuga. Ecol Monogr. 1958;28(1):2–20. [Google Scholar]
34.Zhu W, Chang L, Shi S, Lu N, Du S, Li J, et al. Gut microbiota reflect adaptation of cave-dwelling tadpoles to resource scarcity. ISME J. 2024;18(1):1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Zhu W, Liu L, Wang X, Gao X, Jiang J, Wang B. Transcriptomics reveals the molecular processes of light-induced rapid darkening of the non-obligate cave dweller Oreolalax rhodostigmatus (Megophryidae, Anura) and their genetic basis of pigmentation strategy. BMC Genomics. 2018;19(1):422. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Lu N, Zhu W, Tang CY, Yan C, Chen Q, Wu W, et al. Genome of a stage-dependent cave-dwelling frog reveals the genetic mechanism of an extremely divergent biphasic life cycle. Cell Rep. 2025;44(5):115640. [DOI] [PubMed] [Google Scholar]
37.Bilandzija H, Hollifield B, Steck M, Meng G, Ng M, Koch AD, et al. Phenotypic plasticity as a mechanism of cave colonization and adaptation. eLife. 2020;9:e51830. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Fei L, Ye CY, Jiang JP. Colored atlas of Chinese amphibians and their distributions. Chengdu, China: Sichuan Publishing House of Science & Technology; 2012. [Google Scholar]
39.Barr TC, Holsinger JR. Speciation in cave faunas. Annu Rev Ecol Evol Syst. 1985;16(1):313–37. [Google Scholar]
40.Bizjak Mali L, Sepčić K, Bulog B. Long-term starvation in cave salamander effects on liver ultrastructure and energy reserve mobilization. J Morphol. 2013;274(8):887–900. [DOI] [PubMed] [Google Scholar]
41.Hervant F: Starvation in subterranean species versus surface-dwelling species: crustaceans, fish, and salamanders. In: Comparative Physiology of Fasting, Starvation, and Food Limitation. Edited by McCue MD. Berlin, Heidelberg: Springer Berlin Heidelberg; 2012: 91–102.
42.Medley JK, Persons J, Biswas T, Olsen L, Peuß R, Krishnan J, et al. The metabolome of Mexican cavefish shows a convergent signature highlighting sugar, antioxidant, and ageing-related metabolites. Elife. 2022;11:e74539. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Martin LB, Scheuerlein A, Wikelski M. Immune activity elevates energy expenditure of house sparrows: a link between direct and indirect costs? Proc R Soc Lond B Biol Sci. 2003;270(1511):153–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Hou C. Energetic cost of biosynthesis is a missing link between growth and longevity in mammals. Proc Natl Acad Sci U S A. 2024;121(20):e2315921121. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Riddle MR, Aspiras AC, Gaudenz K, Peuß R, Sung JY, Martineau B, et al. Insulin resistance in cavefish as an adaptation to a nutrient-limited environment. Nature. 2018;555(7698):647–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Aspiras AC, Rohner N, Martineau B, Borowsky RL, Tabin CJ. Melanocortin 4 receptor mutations contribute to the adaptation of cavefish to nutrient-poor conditions. Proc Natl Acad Sci U S A. 2015;112(31):9668–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Hüppop K. Oxygen consumption of Astyanax fasciatus (Characidae, Pisces): a comparison of epigean and hypogean populations. Environ Biol Fishes. 1986;17(4):299–308. [Google Scholar]
48.Miyata K, Ose K. Thyroid hormone-disrupting effects and the amphibian metamorphosis assay. J Toxicol Pathol. 2012;25:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Sanchez E, Küpfer E, Goedbloed DJ, Nolte AW, Lüddecke T, Schulz S, et al. Morphological and transcriptomic analyses reveal three discrete primary stages of post embryonic development in the common fire salamander, Salamandra salamandra. J Exp Zool B Mol Dev Evol. 2018;330:96–108. [DOI] [PubMed] [Google Scholar]
50.Boatright-Horowitz SS, Simmons AM. Transient “deafness” accompanies auditory development during metamorphosis from tadpole to frog. Proc Natl Acad Sci U S A. 1997;94(26):14877–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Gosner KL. A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica. 1960;16(3):183–90. [Google Scholar]
52.Kumar S, Stecher G, Suleski M, Sanderford M, Sharma S, Tamura K. MEGA12: molecular evolutionary genetic analysis version 12 for adaptive and green computing. Mol Biol Evol. 2024;41(12):msae263. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Paradis E. pegas: an R package for population genetics with an integrated–modular approach. Bioinformatics. 2010;26(3):419–20. [DOI] [PubMed] [Google Scholar]
54.Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155(2):945–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Kim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat Biotechnol. 2019;37(8):907–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Pertea M, Pertea GM, Antonescu CM, Chang TC, Mendell JT, Salzberg SL. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nat Biotechnol. 2015;33(3):290–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Li B, Dewey CN. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 2011;12:323. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Lu N, Zhu W, Wang B: Chromosome-level assembly of Oreolalax rhodostigmatus genome and transcriptome reveals cave adaptation. CNGB https://db.cngb.org/data_resources/project/CNP0005050; 2025.
59.Dixon P. VEGAN, a package of R functions for community ecology. J Veg Sci. 2003;14(6):927–30. [Google Scholar]
60.Bu D, Luo H, Huo P, Wang Z, Zhang S, He Z, et al. KOBAS-i: intelligent prioritization and exploratory visualization of biological functions for gene enrichment analysis. Nucleic Acids Res. 2021;49(W1):W317–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Kohl M, Wiese S, Warscheid B. Cytoscape: software for visualization and analysis of biological networks. In: Hamacher M, Eisenacher M, Stephan C, editors. Data mining in proteomics: from standards to applications. Totowa, NJ: Humana Press; 2011. p. 291–303. [DOI] [PubMed] [Google Scholar]
62.Wickham H. Ggplot2: elegant graphics for data analysis. New York, NY, USA: Springer Publishing Company, Incorporated; 2009. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

12915_2026_2532_MOESM1_ESM.docx^{(1MB, docx)}

Additional file 1: Figures S1-S2. FigS1-Genetic background analyses. FigS2- Explanation rate of developmental stages and environment on the variations of gene transcriptional profiles.

12915_2026_2532_MOESM2_ESM.xlsx^{(22.7KB, xlsx)}

Additional file 2: Table S1. Table S1-Sample information of this study.

Data Availability Statement

The raw sequencing data were deposited in the China National GeneBank DataBase (CNGBdb; (https://db.cngb.org/)) under accession number CNP0005050 [58].

[CR1] 1.West-Eberhard MJ: Plasticity. Developmental Plasticity and Evolution. Edited by West-Eberhard MJ: Oxford University Press; 2003: 0.

[CR2] 2.Thompson JD. Phenotypic plasticity as a component of evolutionary change. Trends Ecol Evol. 1991;6(8):246–9. [DOI] [PubMed] [Google Scholar]

[CR3] 3.Miner BG, Sultan SE, Morgan SG, Padilla DK, Relyea RA. Ecological consequences of phenotypic plasticity. Trends Ecol Evol. 2005;20(12):685–92. [DOI] [PubMed] [Google Scholar]

[CR4] 4.Yampolsky LY, Schaer TMM, Ebert D. Adaptive phenotypic plasticity and local adaptation for temperature tolerance in freshwater zooplankton. Proc R Soc Lond B Biol Sci. 2014;281(1776):20132744. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Rion S, Kawecki TJ. Evolutionary biology of starvation resistance: what we have learned from Drosophila. J Evol Biol. 2007;20(5):1655–64. [DOI] [PubMed] [Google Scholar]

[CR6] 6.Barata C, Baird D, Soares A. Phenotypic plasticity in Daphnia magna Straus: variable maturation instar as an adaptive response to predation pressure. Oecologia. 2001;129(2):220–7. [DOI] [PubMed] [Google Scholar]

[CR7] 7.Agrawal AA. Phenotypic plasticity in the interactions and evolution of species. Science. 2001;294(5541):321–6. [DOI] [PubMed] [Google Scholar]

[CR8] 8.Price TD, Qvarnström A, Irwin DE. The role of phenotypic plasticity in driving genetic evolution. Proc R Soc Lond B Biol Sci. 2003;270(1523):1433–40. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Pigliucci M. Evolution of phenotypic plasticity: where are we going now? Trends Ecol Evol. 2005;20(9):481–6. [DOI] [PubMed] [Google Scholar]

[CR10] 10.Fabian D, Flatt T. Life history evolution. Nature Education Knowledge. 2012;3(10):24. [Google Scholar]

[CR11] 11.Wassersug RJ. The adaptive significance of the tadpole stage with comments on the maintenance of complex life cycles in anurans. Am Zool. 1975;15(2):405–17. [Google Scholar]

[CR12] 12.McDiarmid RW, Altig R. Tadpoles: the biology of anuran larvae. University of Chicago Press; 1999. [Google Scholar]

[CR13] 13.Chuliver M, Agnolín FL, Scanferla A, Aranciaga Rolando M, Ezcurra MD, Novas FE, et al. The oldest tadpole reveals evolutionary stability of the anuran life cycle. Nature. 2024;636(8041):138–42. [DOI] [PubMed] [Google Scholar]

[CR14] 14.Wollenberg Valero KC, Garcia-Porta J, Rodríguez A, Arias M, Shah A, Randrianiaina RD, et al. Transcriptomic and macroevolutionary evidence for phenotypic uncoupling between frog life history phases. Nat Commun. 2017;8(1):15213. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Sherratt E, Vidal-García M, Anstis M, Keogh JS. Adult frogs and tadpoles have different macroevolutionary patterns across the Australian continent. Nat Ecol Evol. 2017;1(9):1385–91. [DOI] [PubMed] [Google Scholar]

[CR16] 16.Roelants K, Haas A, Bossuyt F. Anuran radiations and the evolution of tadpole morphospace. Proc Natl Acad Sci U S A. 2011;108(21):8731–6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Warkentin KM. Adaptive plasticity in hatching age: a response to predation risk trade-offs. Proceed Natl Acad Sci. 1995;92(8):3507–10. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Wilson RS, Kraft PG, Van Damme R. Predator-specific changes in the morphology and swimming performance of larval Rana lessonae. Funct Ecol. 2005;19(2):238–44. [Google Scholar]

[CR19] 19.Touchon JC, Warkentin KM. Thermally contingent plasticity: temperature alters expression of predator-induced colour and morphology in a Neotropical treefrog tadpole. J Anim Ecol. 2011;80(1):79–88. [DOI] [PubMed] [Google Scholar]

[CR20] 20.Zhu W, Zhang MH, Chang LM, Li C, Zhu WB, Xie F, et al. Characterizing the composition, metabolism and physiological functions of the fatty liver in Rana omeimontis tadpoles. Front Zool. 2019;16:42. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Tejedo M, Reques R. Plasticity in metamorphic traits of natterjack tadpoles: the interactive effects of density and pond duration. Oikos. 1994;71(2):295–304. [Google Scholar]

[CR22] 22.Álvarez D, Nicieza AG. Effects of temperature and food quality on anuran larval growth and metamorphosis. Funct Ecol. 2002;16(5):640–8. [Google Scholar]

[CR23] 23.Relyea RA. Getting out alive: how predators affect the decision to metamorphose. Oecologia. 2007;152(3):389–400. [DOI] [PubMed] [Google Scholar]

[CR24] 24.Carreira BM, Segurado P, Orizaola G, Gonçalves N, Pinto V, Laurila A, et al. Warm vegetarians? Heat waves and diet shifts in tadpoles. Ecology. 2016;97(11):2964–74. [DOI] [PubMed] [Google Scholar]

[CR25] 25.Montaña CG, Silva SDGTM, Hagyari D, Wager J, Tiegs L, Sadeghian C, et al. Revisiting “what do tadpoles really eat?” A 10-year perspective. Freshw Biol. 2019;64(12):2269–82. [Google Scholar]

[CR26] 26.Ramamonjisoa N, Iwai N, Natuhara Y. Post-metamorphic costs of carnivorous diets in an omnivorous tadpole. Copeia. 2016;104(4):808–15. [Google Scholar]

[CR27] 27.Pfennig D. The adaptive significance of an environmentally-cued developmental switch in an anuran tadpole. Oecologia. 1990;85(1):101–7. [DOI] [PubMed] [Google Scholar]

[CR28] 28.Romero A. Cave biology: life in darkness. Cambridge university press; 2009. [Google Scholar]

[CR29] 29.Jeffery WR. Regressive evolution in Astyanax cavefish. Annu Rev Genet. 2009;43(1):25–47. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Riddle MR, Aspiras AC, Gaudenz K, Peuss R, Sung JY, Martineau B, et al. Insulin resistance in cavefish as an adaptation to a nutrient-limited environment. Nature. 2018;555(7698):647–51. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Yang J, Chen X, Bai J, Fang D, Qiu Y, Jiang W, et al. The Sinocyclocheilus cavefish genome provides insights into cave adaptation. BMC Biol. 2016;14(1):1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Salvidio S. Population dynamics and regulation in the cave salamander Speleomantes strinatii. Naturwissenschaften. 2007;94(5):396–400. [DOI] [PubMed] [Google Scholar]

[CR33] 33.Hutchison VH. The distribution and ecology of the cave salamander, Eurycea lucifuga. Ecol Monogr. 1958;28(1):2–20. [Google Scholar]

[CR34] 34.Zhu W, Chang L, Shi S, Lu N, Du S, Li J, et al. Gut microbiota reflect adaptation of cave-dwelling tadpoles to resource scarcity. ISME J. 2024;18(1):1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Zhu W, Liu L, Wang X, Gao X, Jiang J, Wang B. Transcriptomics reveals the molecular processes of light-induced rapid darkening of the non-obligate cave dweller Oreolalax rhodostigmatus (Megophryidae, Anura) and their genetic basis of pigmentation strategy. BMC Genomics. 2018;19(1):422. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Lu N, Zhu W, Tang CY, Yan C, Chen Q, Wu W, et al. Genome of a stage-dependent cave-dwelling frog reveals the genetic mechanism of an extremely divergent biphasic life cycle. Cell Rep. 2025;44(5):115640. [DOI] [PubMed] [Google Scholar]

[CR37] 37.Bilandzija H, Hollifield B, Steck M, Meng G, Ng M, Koch AD, et al. Phenotypic plasticity as a mechanism of cave colonization and adaptation. eLife. 2020;9:e51830. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Fei L, Ye CY, Jiang JP. Colored atlas of Chinese amphibians and their distributions. Chengdu, China: Sichuan Publishing House of Science & Technology; 2012. [Google Scholar]

[CR39] 39.Barr TC, Holsinger JR. Speciation in cave faunas. Annu Rev Ecol Evol Syst. 1985;16(1):313–37. [Google Scholar]

[CR40] 40.Bizjak Mali L, Sepčić K, Bulog B. Long-term starvation in cave salamander effects on liver ultrastructure and energy reserve mobilization. J Morphol. 2013;274(8):887–900. [DOI] [PubMed] [Google Scholar]

[CR41] 41.Hervant F: Starvation in subterranean species versus surface-dwelling species: crustaceans, fish, and salamanders. In: Comparative Physiology of Fasting, Starvation, and Food Limitation. Edited by McCue MD. Berlin, Heidelberg: Springer Berlin Heidelberg; 2012: 91–102.

[CR42] 42.Medley JK, Persons J, Biswas T, Olsen L, Peuß R, Krishnan J, et al. The metabolome of Mexican cavefish shows a convergent signature highlighting sugar, antioxidant, and ageing-related metabolites. Elife. 2022;11:e74539. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Martin LB, Scheuerlein A, Wikelski M. Immune activity elevates energy expenditure of house sparrows: a link between direct and indirect costs? Proc R Soc Lond B Biol Sci. 2003;270(1511):153–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Hou C. Energetic cost of biosynthesis is a missing link between growth and longevity in mammals. Proc Natl Acad Sci U S A. 2024;121(20):e2315921121. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Riddle MR, Aspiras AC, Gaudenz K, Peuß R, Sung JY, Martineau B, et al. Insulin resistance in cavefish as an adaptation to a nutrient-limited environment. Nature. 2018;555(7698):647–51. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Aspiras AC, Rohner N, Martineau B, Borowsky RL, Tabin CJ. Melanocortin 4 receptor mutations contribute to the adaptation of cavefish to nutrient-poor conditions. Proc Natl Acad Sci U S A. 2015;112(31):9668–73. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Hüppop K. Oxygen consumption of Astyanax fasciatus (Characidae, Pisces): a comparison of epigean and hypogean populations. Environ Biol Fishes. 1986;17(4):299–308. [Google Scholar]

[CR48] 48.Miyata K, Ose K. Thyroid hormone-disrupting effects and the amphibian metamorphosis assay. J Toxicol Pathol. 2012;25:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Sanchez E, Küpfer E, Goedbloed DJ, Nolte AW, Lüddecke T, Schulz S, et al. Morphological and transcriptomic analyses reveal three discrete primary stages of post embryonic development in the common fire salamander, Salamandra salamandra. J Exp Zool B Mol Dev Evol. 2018;330:96–108. [DOI] [PubMed] [Google Scholar]

[CR50] 50.Boatright-Horowitz SS, Simmons AM. Transient “deafness” accompanies auditory development during metamorphosis from tadpole to frog. Proc Natl Acad Sci U S A. 1997;94(26):14877–82. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Gosner KL. A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica. 1960;16(3):183–90. [Google Scholar]

[CR52] 52.Kumar S, Stecher G, Suleski M, Sanderford M, Sharma S, Tamura K. MEGA12: molecular evolutionary genetic analysis version 12 for adaptive and green computing. Mol Biol Evol. 2024;41(12):msae263. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Paradis E. pegas: an R package for population genetics with an integrated–modular approach. Bioinformatics. 2010;26(3):419–20. [DOI] [PubMed] [Google Scholar]

[CR54] 54.Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155(2):945–59. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Kim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat Biotechnol. 2019;37(8):907–15. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Pertea M, Pertea GM, Antonescu CM, Chang TC, Mendell JT, Salzberg SL. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nat Biotechnol. 2015;33(3):290–5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Li B, Dewey CN. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 2011;12:323. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Lu N, Zhu W, Wang B: Chromosome-level assembly of Oreolalax rhodostigmatus genome and transcriptome reveals cave adaptation. CNGB https://db.cngb.org/data_resources/project/CNP0005050; 2025.

[CR59] 59.Dixon P. VEGAN, a package of R functions for community ecology. J Veg Sci. 2003;14(6):927–30. [Google Scholar]

[CR60] 60.Bu D, Luo H, Huo P, Wang Z, Zhang S, He Z, et al. KOBAS-i: intelligent prioritization and exploratory visualization of biological functions for gene enrichment analysis. Nucleic Acids Res. 2021;49(W1):W317–25. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Kohl M, Wiese S, Warscheid B. Cytoscape: software for visualization and analysis of biological networks. In: Hamacher M, Eisenacher M, Stephan C, editors. Data mining in proteomics: from standards to applications. Totowa, NJ: Humana Press; 2011. p. 291–303. [DOI] [PubMed] [Google Scholar]

[CR62] 62.Wickham H. Ggplot2: elegant graphics for data analysis. New York, NY, USA: Springer Publishing Company, Incorporated; 2009. [Google Scholar]

PERMALINK

Stage dependent plasticity and the evolution of a non-obligate cave dwelling life cycle in frogs

Wei Zhu

Liming Chang

Ningning Lu

Jianping Jiang

Ruoyao Ni

Bin Wang

Abstract

Background

Results

Conclusions

Supplementary Information

Background

Fig. 1.

Results

Gene transcriptional variations with development and environment

Table 1.

Fig. 2.

Environment-related gene transcriptional responses of each organ

Liver

Fig. 3.

Tail

Fig. 4.

Skin

Fig. 5.

Hindlimb

Fig. 6.

Shared and distinct transcriptional variations across organs

Fig. 7.

Discussion

Fig. 8.

Conclusions

Methods

Sample collection and information

Table 2.

Genetic background analysis

Transcriptomic analyses

Developmental phase classification based on transcriptomic similarity

Statistic analyses

Supplementary Information

Acknowledgements

Authors’ social media handles

Abbreviations

Authors’ contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Footnotes

Contributor Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases