Altitudinal Variation of Limb Size of a High-Altitude Frog

Abstract

Geographical variation in morphological traits represents a significant ecological phenomenon. According to Allen’s rule, animals inhabiting cooler environments typically exhibit shorter extremities compared to those in warmer regions. While Allen’s rule has been extensively validated along latitudinal gradients, its applicability to altitudinal variation in morphological traits remains less well understood. In this study, we analyzed morphological data—including forelimb length and hindlimb length—from 492 individuals of Rana kukunoris from 26 populations to assess whether relative limb size in both sexes declines with altitude, aligning with Allen’s rule. However, this pattern exhibited geographical regional variation. In the northern and central regions, relative limb length consistently followed the predictions of Allen’s rule. In contrast, the southern region showed no significant altitudinal variation in limb proportions. These results highlight that the applicability of Allen’s rule may be modulated by regional environmental factors and altitude vertical differences, underscoring the need for a nuanced understanding of how local contexts influence morphological adaptations.

1. Introduction

Geographic variation in the morphological characteristics of animals is a widespread natural phenomenon to observe, influenced by the pressures of natural and sexual selection [1,2,3,4]. Environmental factors, such as altitude and latitude, serve as significant selective pressures that shape the morphological traits of animals [5,6,7]. To enhance their survival and optimize fitness, animals must adapt to their habitats through diverse strategies that encompass morphological variation [8,9,10]. Therefore, investigating geographic variation in morphological traits is crucial for understanding the environmental adaptability of organisms [11].

Two well-established biogeographical principles regarding morphological trait variation—Bergmann’s rule and Allen’s rule—have been recognized in the literature [12,13]. Bergmann’s rule addresses geographic variation in body size among populations of endothermic animals [14,15,16], positing that individuals in colder environments tend to exhibit larger body sizes compared to those in warmer climates. Similarly, Allen’s rule relates body part proportions to climate, with a particular focus on limb length in mammals [17]. This rule states that mammals in cold climates tend to have shorter appendages, including ears, tails, and limbs, compared to their counterparts in warmer climates, which aids in minimizing heat loss and conserving body heat in cold environments [18,19]. Both rules suggest that in colder regions, animals can maintain their body temperature by reducing the surface area-to-volume ratio [20]. While Bergmann’s rule has been extensively studied across various taxa, research on Allen’s rule remains relatively scarce [21,22,23,24,25]. To date, the validation of Allen’s rule has primarily focused on latitude [17,19,26], whereas the evaluation of altitudinal variation in appendage size has been less comprehensive [15,27]. Furthermore, the applicability of Allen’s rule to ectothermic species, particularly amphibians, remains to be validated within the context of altitude gradients [26,28,29,30]. Consequently, it is essential to select appropriate model species to investigate these issues.

Rana kukunoris, a high-altitude frog endemic to the Qinghai-Tibetan Plateau, ranges in altitude from 2000 to 4400 m [31], which is listed as least concern in the IUCN red list of threatened species [32]. Previous research on amphibians has primarily focused on life history traits and genetic differences [33,34,35]. Studies on morphological variation have mainly examined the geographical gradient in body size [31,36,37], while investigations into the geographic variation of relative limb size remain scarce [38,39]. The significant altitudinal differences across its distribution make R. kukunoris an excellent model for studying altitudinal variation in limb length and provide a valuable opportunity to test Allen’s rule. Previous studies have shown that there is a debate regarding the altitudinal variation in limb size of this species [38,39]. Therefore, it is necessary to test the hypothesis using this species across its entire distribution region.

In this study, we collected morphological data from 492 individuals of R. kukunoris across 26 populations from three regions to investigate the altitudinal variation in morphological characteristics (forelimb length and hindlimb length) along altitudinal gradients. Specifically, our objectives were to: (1) examine the altitudinal variation of relative limb size among different geographic populations; (2) assess the applicability of Allen’s rule at an altitudinal scale; and (3) explore potential regional and sexual differences in the application of Allen’s rule. This study will enhance our understanding of the adaptive potential of R. kukunoris in response to environmental changes through morphological variations.

2. Materials and Methods

2.1. Sampling Collection

We collected a total of 492 individuals (300 males and 192 females) from 26 populations across three regions, encompassing the entire distribution range of the species, between early March and late August 2023 (Figure 1, Supplementary Materials: Table S1 and Figure S1). Frogs were identified as sexually mature males based on the presence of nuptial pads on their foredigits, and as sexually mature females when they exhibited well-developed oocytes [33]. Upon completion of all experiments, all individuals were released at the same site where they were captured.

2.2. Data Acquisition

Morphometric measurements, encompassing body size (snout-vent length, SVL), right forelimb length (lower arm and hand length, LAHL), and right hindlimb length (the sum of thigh, tibia, tarsus, and foot lengths, HILL), were recorded for each individual using digital calipers with an accuracy of 0.01 mm. Previous research has shown a positive correlation between body size and age in R. kukunoris [31,40]. In our study, we opted not to determine individual ages through skeletochronology to minimize harm to the species. Instead, we utilized body length as a proxy indicator of age to control for the effect of age on limb length.

2.3. Statistical Analysis

To investigate the presence of altitudinal differences in LAHL (or HILL), we employed linear mixed models (LMMs) with LAHL (or HILL) as separate dependent variables. In these models, sex was treated as a fixed factor, SVL as a covariate to correct for the effect of age, and altitude as a covariate to test for the altitudinal effect on relative limb length. Geographic region was included as a random factor. Furthermore, we incorporated the interaction effects of altitude and sex into the model. To improve the parameter estimation accuracy and reliability of hypothesis testing, we utilized REML estimation for random effects and the Satterthwaite procedure to approximate the degrees of freedom [33]. Additionally, we conducted a separate linear regression analysis (LRA) to investigate the relationship between altitude and morphological traits (LAHL and HILL) for each region. The structure was consistent, with LAHL and/or HILL as dependent variables, sex as a fixed factor, and SVL and altitude as covariates, including the interaction effects of altitude and sex.

Before conducting the analyses, all data, including altitude, LAHL, HILL, and SVL, were log-transformed. We then fitted linear mixed models using the ‘lmer’ function installed in R (version 4.2.2). In this study, all probabilities were two-tailed, and values are reported as the mean ± standard deviation (mean ± SD).

3. Results

Overall, the relative limb length of males and females decreased with the increase of altitude gradients (LAHL/SVL: B = −0.087, t = −4.352; df = 405.6, p < 0.001; HILL/SVL: B = −0.096, t = −5.956; df = 411.6, p < 0.001). The change in forelimb length (B = −0.071, t = −0.317; df = 485.2, p = 0.751) and hindlimb length (B = −0.069, t = −0.376; df = 485.1, p = 0.707) between males and females was similar as altitude increased, and the interactive effect of altitude and sex was insignificant (both p > 0.8, Figure 2,

Table 1. The altitudinal change of the relative limb size of Rana kukunoris across all the distribution range (LAHL: Lower arm and hand length; HILL: the sum of thigh, tibia, tarsus, and foot lengths; SD: standard deviation; SE: standard error; df: degree of freedom).

Morphological Trait	Factor	Random		Fixed
		Variance	SD	Estimate	SE	df	t	p
LAHL	Geographic region	0.0005	0.021
	Intercept			0.128	0.177	401.0	0.723	0.470
	Altitude			−0.087	0.020	405.6	−4.352	<0.001
	Sex			−0.071	0.225	485.2	−0.317	0.751
	SVL			0.932	0.018	485.2	52.580	<0.001
	Altitude × sex			0.001	0.028	485.2	0.033	0.974
HILL	Geographic region	0.0003	0.018
	Intercept			1.408	0.144	404.2	9.775	<0.001
	Altitude			−0.096	0.016	411.6	−5.956	<0.001
	Sex			−0.069	0.182	485.1	−0.376	0.707
	SVL			0.951	0.014	485.2	66.073	<0.001
	Altitude × sex			0.003	0.023	485.1	0.149	0.882

).

Except for the southern region (LAHL/SVL: B = −0.194, t = −1.029, p = 0.306; HILL/SVL: B = 0.0204, t = 1.396, p = 0.166), the other two regions showed that the relative limb length of males and females decreased with the increase of altitude gradients (northern region: LAHL/SVL: B = −0.242, t = −3.473, p < 0.001; HILL/SVL: B = −0.134, t = −2.993, p = 0.003; central region: LAHL/SVL: B = −0.0.084, t = −3.831, p < 0.001; HILL/SVL: B = −0.094, t = −4.616, p < 0.001). For these three regions, the change in limb length was similar between males and females as altitude increased, and we also found that the interactive effect of altitude and sex was not obvious (both p > 0.4,

Table 2. The altitudinal variation in the relative limb size of Rana kukunoris across different geographic regions (LAHL: the lower arm and hand length; HILL: the sum of thigh, tibia, tarsus, and foot lengths; SE: the standard error).

Geographic Region	Morphological Trait	Factor	Estimate	SE	t	p
Northern region	LAHL	Intercept	1.586	0.587	2.701	0.008
		Altitude	−0.242	0.070	−3.473	0.001
		Sex	0.222	0.822	0.271	0.787
		SVL	0.862	0.032	26.632	<0.001
		Altitude × sex	−0.036	0.104	−0.342	0.733
	HILL	Intercept	1.740	0.378	4.606	<0.001
		Altitude	−0.134	0.045	−2.993	0.003
		Sex	−0.472	0.528	−0.894	0.373
		SVL	0.939	0.021	45.090	<0.001
		Altitude × sex	0.056	0.067	0.828	0.409
Central region	LAHL	Intercept	−0.011	0.198	−0.053	0.958
		Altitude	−0.084	0.022	−3.831	<0.001
		Sex	−0.276	0.292	−0.946	0.345
		SVL	0.966	0.023	42.444	<0.001
		Altitude × sex	0.026	0.037	0.725	0.469
	HILL	Intercept	1.316	0.184	7.155	<0.001
		Altitude	−0.094	0.020	−4.616	<0.001
		Sex	0.140	0.271	0.519	0.605
		SVL	0.967	0.021	45.805	<0.001
		Altitude × sex	−0.024	0.034	−0.710	0.479
Southern region	LAHL	Intercept	1.465	1.579	0.928	0.356
		Altitude	−0.194	0.189	−1.029	0.306
		Sex	−1.739	3.281	−0.530	0.598
		SVL	0.820	0.065	12.668	<0.001
		Altitude × sex	0.201	0.396	0.508	0.613
	HILL	Intercept	−0.904	1.225	−0.738	0.463
		Altitude	0.204	0.146	1.396	0.166
		Sex	−1.061	2.546	−0.417	0.678
		SVL	0.910	0.050	18.128	<0.001
		Altitude × sex	0.124	0.307	0.403	0.688

).

4. Discussion

Our findings revealed that R. kukunoris generally exhibit shorter relative appendages in high-altitude settings, which is consistent with Allen’s rule. However, this pattern was not uniform across the three geographical regions studied. In the northern and central regions, both males and females showed a significant reduction in relative limb lengths as altitude increased. Conversely, in the southern populations, no significant correlation was found between relative limb length and altitude. Simultaneously, we discovered that there were no sexual differences in the altitudinal variation of relative limb size.

Allen’s rule states that the appendages (including limbs, ears, tails, and snouts) of endothermic creatures are typically longer and more slender in warmer climates [12]. This rule has been widely validated in many mammals and birds, often manifesting as geographic variation in the size of body appendages along latitudinal or altitudinal gradients [24,41,42,43]. However, its applicability to ectothermic amphibians remains controversial [24,25,26,28,38,39]. For instance, Escalona et al. (2024) found that Neotropical treefrogs showed no significant reduction in relative limb length with decreasing environmental temperatures [28]. Moreover, Alho et al. (2011) found that Rana temporaria also failed to confirm this simple prediction as they found that there was insignificant linear latitudinal relationship in the appendages lengths [26]. The research results from Leung et al. (2021) demonstrated a significant decrease extremities of R. kukunoris with increasing elevation and thus supported Allen’s rule [38]. However, Yang et al. (2024) observed a converse altitudinal cline for R. kukunoris as they found an increasing forelimb length as altitude increased [39]. Here we found a positive relationship between limb length and altitude, aligning with the findings of Leung et al. and supporting Allen’s rule [38]. This relationship is likely closely linked to thermoregulatory benefits, as such morphological changes represent an adaptive evolutionary response aimed at minimizing surface area to reduce heat loss, thereby aiding in maintaining body temperature in cold environments [38,44].

Interestingly, our results generally supported Allen’s rule, as limb size decreased in colder climates [12]. However, the altitudinal cline of limb size showed regional differences. In the northern and central regions, the relative limb size of both males and females exhibited a significant negative correlation with altitude, whereas no such trend was observed in the southern region. A possible explanation for these differences may relate to the magnitude of vertical altitudinal variation of each region. The northern region spans an altitude difference of approximately 1000 m, while the central region covers a range of around 2000 m. In contrast, the southern region exhibits only a narrow altitudinal gradient of 500 m. Leung et al. (2021) also observed a significant decrease extremities of R. kukunoris with altitudinal difference of 1600 m across ten populations [38], a sampling site similar to the central population in our study. In the southern region, with an altitudinal difference of only 500 m, the relatively uniform environmental conditions may fail to reach the (altitudinal and/or temperature) threshold required to induce significant changes in limb proportions. Thus, the observed differences among these regions may arise from the magnitude difference of variations in altitudinal and environmental factors [38]. Nonetheless, these hypotheses remain speculative and require further investigation for validation.

Allen’s rule, derived from empirical generalizations across numerous animal taxa [14,17,45,46], often has exceptions but exhibits variable applicability to amphibians [28,47]. In this study, our findings supported Allen’s rule, as R. kukunoris tends to have shorter relative appendages in high-altitude environments. Unlike the results of Yang et al., who found sexual differences in the altitudinal variation of relative forelimb length of R. kukunoris [39], we did not observe sexual differences in the altitudinal variation of relative limb size of this species. Additionally, we observed potential regional disparities in morphological characteristics. Future research on the altitudinal patterns of morphological variation should broaden the sampling areas and thoroughly account for regional environmental differences, while also controlling for confounding factors such as latitudinal and longitudinal variations.

Overall, we measured the forelimb and hindlimb lengths of 492 R. kukunoris individuals from 26 populations and found that relative limb length decreased with increasing altitude, corroborating Allen’s rule. To gain deeper insights into the adaptive responses of R. kukunoris to varying environmental conditions, future research should explore the molecular mechanisms underlying the geographical variations in these morphological traits.