Received: 7 April 2021 Revised: 30 July 2021 Accepted: 2 August 2021 DOI: 10.1002/ar.24765 FULL LENGTH ARTICLE Postnatal ontogeny of the femur in fossorial and semiaquatic water voles in the 3D-shape space Ana Filipa Dura~o1 | Francesc Muñoz-Muñoz1 | Jacint Ventura1,2 1 Departament de Biologia Animal, de Biologia Vegetal i d'Ecologia, Facultat Abstract de Biociències, Universitat Autònoma Water voles of the genus Arvicola constitute an excellent subject to investigate de Barcelona, Campus de Bellaterra, to which extent function affects postnatal developmental growth of limb struc- Cerdanyola del Vallès, Spain 2 tures in phylogenetically close species. We performed a comparative analysis Àrea de recerca en petits mamífers, Museu de Ciències Naturals de Granollers of postweaning femur form changes between Arvicola sapidus (semiaquatic) “La Tela”, Barcelona, Spain and Arvicola scherman (fossorial) using three-dimensional landmark-based Correspondence geometric morphometrics. In both species, we observed greater femur robust- Ana Filipa Dur~ao and Francesc Muñoz- ness in juvenile individuals than in adult ones, probably due to the accommo- Muñoz, Departament de Biologia Animal, dation of high loads on the bone during initial locomotor efforts. Significant de Biologia Vegetal i d'Ecologia, Facultat de Biociències, Universitat Autònoma de interspecific differences were also found in the femur size and shape of adult Barcelona, Campus de Bellaterra, 08193 specimens, as well as in the postnatal allometric and phenotypic trajectories. Cerdanyola del Vallès, Spain. In terms of phenotypic variation, fossorial water voles show relatively wider Email: anafilipa.lopes@uab.cat (A. F. D.) and francesc.munozm@uab.cat (F. M.-M.) third and lesser trochanters, and greater femur robustness than A. sapidus, characters associated to the digging activity. In contrast, A. sapidus displays a Funding information slight increase of the greater trochanter in comparison with A. scherman, Comissionat per a Universitats i Recerca (Generalitat de Catalunya), Grant/Award which is seemingly an adaptive response for enhancing propulsion through Number: 2014-SGR-1241 the water. Results evidence that certain morphological traits and differences between A. sapidus and A. scherman in the allometric and phenotypic trajecto- ries of the femur are associated with their different locomotor mode. KEYWORDS allometry, Arvicola, locomotion, ontogenetic pattern, phenotypic trajectories 1 | INTRODUCTION in the architecture of the myoskeletal system (Candela & Picasso, 2008; Lessa et al., 2008; Salton & Sargis, 2009; Over the past decades, extensive work has been done on Sargis, 2002). Moreover, ecological factors, functional con- the role played by the locomotor behavior on the pheno- straints, or evolutionary adaptations can be associated typic variation of limb long bones in rodents (Dur~ao, with modifications of the postnatal developmental path- Muñoz-Muñoz, & Ventura, 2020; García-Esponda & ways of the corresponding bone structures (Adams & Candela, 2016; Hildebrand, 1985; Lehmann, 1963; Lessa, Nistri, 2010; Dur~ao et al., 2020; Dur~ao, Ventura, & Vassallo, Verzi, & Mora, 2008; Samuels & Valkenburgh, Muñoz-Muñoz, 2019; Esquerré, Sherratt, & 2008; Stein, 1993, 2000). Notably, it has been shown that Keogh, 2017; Gray, Sherratt, Hutchinson, & Jones, 2019; different locomotion modes may lead to profound changes Wilson & Sanchez-Villagra, 2010). In this context, a This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2021 The Authors. The Anatomical Record published by Wiley Periodicals LLC on behalf of American Association for Anatomy. Anat Rec. 2021;1–14. wileyonlinelibrary.com/journal/ar 1 2 DURÃO ET AL. comparison of the ontogenetic trajectories of limb long During swimming, A. sapidus uses fore and hind limbs to bones of closely related species with different locomotor move quickly in water and with hind limbs to generate types is an excellent way to determine to which extent propulsion. Thus, in A. scherman and A. sapidus, hind function affects growth patterns. Within rodents, water legs are employed differently, according to their domi- voles of the Palaearctic genus Arvicola constitute an nant type of locomotion. On the one hand, fossorial optimal subject to explore this issue since it comprises a water voles mainly use their hindlimbs to remove the soil high number of morphotypes and two main ecological from the burrow (see Airoldi et al., 1976). During this forms, semiaquatic, and fossorial. phase, the animal takes a flexed posture; the pelvis is The phylogenetic relationships of the representatives of tilted forward, the angle between the femur and the pel- Arvicola are currently under debate. Classically two vis diminishes, and the thigh is extended backward species within this genus were recognized (see, (Laville, 1989). On the other hand, as mentioned above, e.g., Reichstein, 1963, 1982): Arvicola sapidus, which com- semiaquatic water voles use hind legs to generate propul- prises semiaquatic populations, and Arvicola terrestris, with sion during swimming. Specifically, A. sapidus places first fossorial and/or semiaquatic representatives. In the taxo- the hind feet ventrally and then extends the hip and nomic review of this genus by Musser and Carleton (2005), knee. In this process, the femur is essential to make rapid three species were accepted: the southwestern water vole, movements and to exert force during the legs' stroke. A. sapidus Miller, 1908 (semiaquatic); the montane water Despite the important role that the hindlimbs play vole, Arvicola scherman (Shaw, 1801) (fossorial); and the in the different locomotion modes in Arvicola, little is European water vole, Arvicola amphibius (Linnaeus, 1758) known about the postnatal development of their bones (semiaquatic or mostly semiaquatic). Although A. sapidus is (Garde, 1992; Ventura, 1988, 1990, 1992). Surprisingly, widely accepted as a valid species, recent studies based on this observation can be also extended to rodents in gen- analyses of genetic data have given rise to controversy on eral (Elissamburu & Vizcaíno, 2004; Laville, 1989; the taxonomy of the formerly accepted A. terrestris (for Montoya-Sanhueza, Wilson, & Chinsamy, 2019; details, see Chevret et al., 2020; Kryštufek et al., 2015; Samuels & Valkenburgh, 2008; Stein, 1993). The pre- Mahmoudi et al., 2020; Pardiñas et al., 2017; Ventura & sent study is the third one of a series that aimed to Casado-Cruz, 2011 and references therein). As for the Ibe- determine the postnatal development of bone struc- rian populations analyzed in the present study, according to tures involved in locomotion in the two ecological these new results, fossorial water voles from the Pyrenees forms of Arvicola: fossorial and semiaquatic. The main could be attributed to A. amphibius sensu lato (Kryštufek goal of this research is to compare the postnatal ontog- et al., 2015), Arvicola monticola (Mahmoudi et al., 2020; eny of the femur in A. scherman (fossorial) and Pardiñas et al., 2017) or A. amphibius (Chevret et al., 2020). A. sapidus (semiaquatic) to detect possible phyloge- Since this point remains unsolved, we decided to follow the netic and functional signals in this process. We com- taxonomic arrangement by Musser and Carleton (2005) to pared the vectors of growth trajectories to estimate clearly distinguish between typical fossorial (A. scherman) whether evolved functional differences are associated and semiaquatic (A. sapidus) morphotypes. with different ontogenetic patterns. We hypothesize Fossorial water voles exhibit the so-called chisel tooth that these species share conserved shape changes digging, which involves the use of procumbent incisors to throughout their postnatal growth. However, consider- break the soil (Stein, 2000). The soil loosened by the inci- ing the corresponding type of locomotion, we also sors is removed from the tunnel through rapid and alter- expect significant interspecific differences both in the native movements of the fore and hind limbs, and the allometric and phenotypic trajectories. This hypothesis gallery walls are compacted by the rostrum (Airoldi, is supported by our previous results obtained for the Altrocchi, & Meylan, 1976; Laville, 1989; Laville, Casinos, mandible (Dur~ao et al., 2019) and humerus (Dur~ao Gasc, Renous, & Bou, 1989). This digging process allows et al., 2020) of these taxa. It is worth mentioning that A. scherman to build extensive and complex burrow sys- although there is available information on the tems in hard soils (Airoldi, 1976; Airoldi & De postweaning development of the femur in scratch dig- Werra, 1993; Laville, 1989). Conversely to fossorial water ger rodents (Echeverría, Becerra, & Vassallo, 2014; voles, A. sapidus lives in freshwater habitats where it Montoya-Sanhueza et al., 2019; Vassallo, 1998), here builds burrows on easily penetrable soils (Ventura, 2007). we give the first findings on this subject concerning a These rodents are excellent swimmers and divers (Mate, chisel-digger vole (A. scherman). The results obtained Barrull, Salicrú, Ruiz-Olmo, & Gosàlbez, 2013; Quéré & in A. scherman were compared, under a functional per- Le Louarn, 2011) albeit they do not show marked mor- spective, with those corresponding to a typical semi- phological characteristics associated with these activities. aquatic species of the same genus (A. sapidus). DURÃO ET AL. 3 2 | MATERIALS AND METHODS 2.2 | Femur form 2.1 | Study specimens 2.2.1 | Three-dimensional (3D) models We studied ontogenetic series of two Arvicola representa- Femur form variation was studied from 3D models tives: A. sapidus and A. scherman. The analyzed sample obtained by the photogrammetric technique, which permits (Table 1) consisted of 182 right femora from museum col- to obtain accurate and precise models of the external form lections, 89 belonging to A. sapidus from the Ebro Delta of small objects from photograph sets (Dur~ao, Muñoz- (Tarragona, Spain), and 93 to A. scherman from the Aran Muñoz, Martínez-Vargas, & Ventura, 2018). Each model Valley (Lleida, Spain). Due to the small number of juve- was constructed from a set of 33 photographs of each femur nile A. sapidus in this sample, we complemented it with using a Canon EOS 750D DSLR camera equipped with a femurs of juvenile individuals from Mélida (Navarra, Tamron 60 mm f/2.0 Di II LD macro lens and a polarizing Spain) collected in 1990, and Banyoles (Girona, Spain) filter mounted on a tripod. The models were obtained and Rieutort (Lozère, France) obtained in 1994. Results through the “turntable method,” which consists of keeping of previous studies showed no significant morphometric the camera fixed at a point and taking photographs of the differences between these populations (Garde, 1992; object while it rotates on a turntable in small intervals. All Ventura, 1988). photographs were taken inside a lightbox to even out the We analyzed specimens belonging to the same collections light and reduce gleams and harsh changes in shadows examined in previous comparative studies on the ontogeny during the rotation of the specimen. In order not to of the mandible (Dur~ao et al., 2019; Ventura & Casado- obstruct viewing angles and to obtain a complete digital Cruz, 2011) and humerus of these species (Dur~ao model, the photographs were taken at intervals of 15 et al., 2020). As in the case of these bones, those specimens (Falkingham, 2012). Thus, photographs were taken in two were already distributed into six classes of relative age complete rotations of 360 : the first rotation at the same according to criteria based on the moulting phase, sexual height as the bone (taking 24 photos at intervals of 15 ) stage, and skull morphology (for details, see Garde, Escala, & and the second rotation at the height of bone size, position- Ventura, 1993; Ventura & Gosalbez, 1992). The age intervals ing the lens at 20 from the horizontal plane (taking corresponding to each group are the following (Garde 9 photos at 40 intervals). Each photo was taken via remote et al., 1993; Ventura & Gosalbez, 1992): Class 0, 3 weeks trigger to reduce camera shake while capturing images and maximum; Class 1, between 3 and 5 weeks; Class 2, between get focused and sharped photos. Internal software parame- 6 and 10 weeks; Class 3, between 10 and 14 weeks; Class ters, such as total error and residual and precision values, 4, between 14 weeks and 1 winter; and Class 5, individuals were used to assess the quality of the models. For more that have lived more than 1 winter. Individuals from the last details about the protocol, see Dur~ao et al. (2018). two age classes were considered fully adults. Excepting the femurs of A. sapidus from Mélida, all material analyzed in the present study is provisionally 2.2.2 | Morphometric data housed in the collection from the Mammalian Biology Research Group of the Universitat Autònoma de Barce- The femur form was characterized using 3D landmark- lona (UAB, Bellaterra, Spain). Specimens from Mélida based geometric morphometrics. A set of 20 landmarks belong to the collection of the Museo de Ciencias of the (Figure 1) were digitized on the femur surface using Pho- Universidad de Navarra (UNAV, Pamplona, Spain). Cata- tomodeler Scanner software v.2015.1.1 (EOS Systems logue numbers of museum material are in Appendix. Inc., 2015). Landmark definition is available in T A B L E 1 Sample size for each Age class species, age class, and sex Species Sex 0 1 2 3 4 5 Total Arvicola sapidus F 2 10 9 3 9 4 89 M 3 10 8 10 9 12 Arvicola scherman F 2 5 10 10 10 10 93 M 4 2 10 10 10 10 Total 11 27 37 33 38 36 182 4 DURÃO ET AL. 2 with sex and age class as categorical factors and log- 18 1 17 transformed centroid size (LCS) as dependent variable. In 19 16 the case of Procrustes ANOVA, sex was introduced as cate- 15 20 14 14 gorical factor, LCS as continuous variable and Procrustes 3 3 coordinates as dependent variables. No sexual dimorphism 13 was found in any species (Supplementary Material - Tables S2–S5), which corroborates results obtained in other 4 4 studies (Garde, 1992; Ventura, 1988, 1990, 1992, 1993; Ven- tura & Casado-Cruz, 2011). Consequently, for each species 5 5 data from both sexes were analyzed together. 2.3.2 | Size and allometry 6 11 9 10 8 7 For both A. sapidus and A. scherman, the mean LCS and 12 10 the respective standard deviations for each age class were computed. The effects of species and age classes on femur F I G U R E 1 Femur image of Arvicola scherman showing the size were tested across the entire sample with a factorial position of the 20 anatomical landmarks used in the geometric ANOVA. To test for differences between species in each morphometric analyses, from left to right: cranial, medial, and caudal. See Table S1 for landmark definitions age class, a Tukey's HSD test was applied. To reduce the occurrence of a type-I error in sets of related tests, a sequential Bonferroni correction was conducted Supplementary Material S1. We set the scale for each (Rice, 1989). These analyses were carried out using model using the mean value of the three measurements STATISTICA v.12 software (StatSoft, 2014). of the linear distance between points 2 and 11. Measure- The statistical relationship between size and shape ments were obtained with a Mitutoyo 500-161-21 Digital (allometry) was assessed for each species separately by Caliper (Mitutoyo America Corporation, Aurora, IL) with a multivariate regression of Procrustes coordinates on LCS 0.01 mm resolution. To avoid interobserver error, all femur (Monteiro, 1999). Statistical significance of the results was digitalization was taken by the same person (A.F.D.). Subse- tested under the null hypothesis of no allometric relation- quent analyses were performed using MorphoJ v.1.07a ship through a permutation test with 10,000 iterations. (Klingenberg, 2011) and a routine written for R statistical These analyses were performed with MorphoJ v.1.07a. A framework v. 4.1.0 (R Development Core Team, 2016). Procrustes ANOVA model using LCS, species and their Femoral form was separated into size and shape compo- interaction as model effect (Goodall, 1991) was conducted nents, which were analyzed separately. Centroid size, defined to estimate the allometric relationship between species, by as the square root of the sum of the square distances of each means of the “procD.lm” function from the R package landmark from their centroid (Bookstein, 1991), was used as geomorph (Adams, Collyer, Kaliontzopoulou, & a proxy for femur size. The raw 3D landmarks coordinates Sherratt, 2017). The statistical significance was assessed were subjected to a generalized Procrustes alignment via distributions generated with resampling permutation (Rohlf & Slice, 1990) to remove nonshape information (size, with 10,000 iterations (Collyer, Sekora, & Adams, 2015). position, and orientation) (Dryden & Mardia, 1998; In this ANOVA, the significant interaction between spe- Goodall, 1991). Thus, the isometric effect of size on shape cies and LCS suggests interspecific differences in allome- was removed but not allometry. The resulting Procrustes try. Since this was the case (see the Results section), to coordinates were used as shape variables in further analyses. evaluate differences in ontogenetic allometric patterns, the attributes of the vectors of allometric coefficients, length (magnitude), and orientation (angle) were computed 2.3 | Statistical analyses applying the “pairwise” function from the package RRPP (Collyer & Adams, 2018). Vector length describes the 2.3.1 | Sexual dimorphism amount of shape changes per unit change of size and vec- tor orientation indicates the relative covariations of shape We assessed the presence and magnitude of sexual dimor- variables per unit change of size (Collyer & Adams, 2013). phism in each species on femur size and shape performing Statistical significance of the tests was via pairwise assess- a factorial analysis of variance (ANOVA) and a Procrustes ments of the similarity in slopes and through 10,000 ran- ANOVA, respectively. The factorial ANOVA was conducted domized residual permutations. DURÃO ET AL. 5 2.3.3 | Shape variation T A B L E 2 Results of the factorial ANOVA testing the effects of species, age class, and the interaction between them on femur size A principal component analysis (PCA) on Procrustes (logarithm of centroid size) aligned coordinates was performed to investigate the vari- Effect df SS MS F p ation among landmarks in the data set. To visualize the Species 1 3.951 3.951 1,566.2 <.001 shape differences associated with the major principal com- Age class 5 4.204 0.841 333.3 <.001 ponents (PCs), a warped 3D model was produced using plotRefToTarget function from R package geomorph Age class  species 5 0.044 0.009 3.5 .005 (Adams et al., 2017). To measure similarity and difference Residuals 170 0.429 0.003 between species, we calculated Procrustes distances from Abbreviations: df, degrees of freedom; F, F statistic; MS, mean squares; p, shape data and nonallometric shape data (allometric shape p-value; SS, sum of squares. removed) in each age class. The residuals obtained from the multivariate regression of Procrustes coordinates on 4.1 LCS for each age class of each species were used as non- allometric shape data (see Klingenberg, 2016). Sequential 4.0 Bonferroni corrections were applied. Femur shape differ- 3.9 ences across ontogeny between A. sapidus and 3.8 A. scherman were visualized by discriminant analysis 3.7 (DA) performed in two separate groups of individuals: 3.6 LCS juvenile and subadult (age classes 0–3), and adult (age classes 4 and 5). All the described analyses were conducted 3.5 with MorphoJ v.1.07a. The resulting coordinates of this 3.4 difference were imported and used in plotRefToTarget 3.3 function to visualize the differences between species. To compare ontogenetic trajectories between species, 3.2 a phenotypic trajectory analysis (PTA) on the Procrustes 3.1 A. sapidus A. scherman coordinates, using the “trajectory.analysis” function from 3.0 the RRPP, was conducted (Adams & Collyer, 2007, 2009; 0 1 2 3 4 5 Collyer & Adams, 2007, 2013). This function quantifies Age class the attributes of the phenotypic trajectories (size, orienta- tion, and shape) for each group and compares these attri- F I G U R E 2 Mean values of the logarithm of centroid size (LCS) butes between groups via permutation (Adams & of the femur across the ontogeny in Arvicola sapidus and Arvicola Collyer, 2009). We used species as group and age class as the scherman (whiskers, standard deviation) trajectory points. Attribute differences were estimated from sampling distributions generated from 10,000 random resid- ual iterations (Adams & Collyer, 2007, 2009; Collyer & throughout ontogeny in both species (Figure 2), Adams, 2007, 2013). To visualize the ontogenetic phenotypic although it was always larger in A. sapidus than in trajectories, a PC of fitted values from linear model fit over A. scherman (Table 3). Multivariate regression of Pro- multiple random permutations was computed (Adams & crustes coordinates on LCS for each species revealed Collyer, 2009). To visualize shape changes, a warped 3D that both species show significant shape changes, in model was plotted. To ensure that the low sample size of A. sapidus the size explained 38.52% of the shape varia- Classes 0 and 1 did not influence the PTA, the same analysis tion and in A. scherman 27.88% (Figure 3). was performed without these two age classes. Species, size, and their interaction had a significant effect on femur shape (Table 4). Thus, there was allometry in femur shape and the significant interaction term suggests 3 | R E SUL T S that the allometric slopes of these species were significantly different. Pairwise comparison of allometric trajectories con- 3.1 | Size and allometry firmed that allometric slopes significantly differed between species (θ = 26.02 ; p < .001). No significant pairwise differ- The factorial ANOVA indicated that femur size was sig- ence in trajectory length was detected (Δd = 0.017; nificantly influenced by species, age class, and the inter- p = .06), but A. sapidus displayed a longer vector than action of both factors (Table 2). Femur size increased A. scherman (dA. sapidus = 0.186; dA. scherman = 0.169). 6 DURÃO ET AL. T A B L E 3 Mean centroid size and Centroid size standard deviation of the femur in Age class Arvicola sapidus Arvicola scherman p-values* Arvicola sapidus and Arvicola scherman 0 35.49 ± 0.92 24.08 ± 0.84 <.001 in each age class and associated p-values 1 38.64 ± 0.46 28.09 ± 0.77 <.001 2 46.68 ± 0.50 32.25 ± 0.46 <.001 3 50.44 ± 0.57 36.38 ± 0.46 <.001 4 53.76 ± 0.48 40.27 ± 0.46 <.001 5 57.56 ± 0.51 42.09 ± 0.46 <.001 Mean 47.10 ± 0.24 33.86 ± 0.24 <.001 *After sequential Bonferroni correction. 0.08 accounted for differences between the two species, with 0.06 most A. sapidus specimens (excepting some adults) show- 0.04 ing positive scores and most A. scherman individuals (Regression score) Femur shape 0.02 (excepting certain individuals of each age class) negative 0.00 ones. Most individuals of A. sapidus (high PC2 values) –0.02 have a slight reduction of the femur head (lateral shift of –0.04 the fovea capitis) and a narrowing of the femur's neck, –0.06 slight shortening of the greater and lesser trochanters, –0.08 A. sapidus A. scherman marked shortening and narrowing of the third trochan- –0.1 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 ter, and distal expansion of both condyles (Figure 4c; LCS Figure S1). F I G U R E 3 Multivariate regression of shape onto the logarithm The DAs showed that both juveniles-subadult (age of centroid size (LCS) in Arvicola sapidus and Arvicola scherman classes 0–3) and adult (age classes 4 and 5) individuals of and their respective regression lines A. scherman showed a more robust femur than A. sapidus. This robustness is mainly due to the expansion of the lesser trochanter, a slight medial shift of the femur head, 3.2 | Variation in femur shape and relatively longer and prominent third trochanter, and phenotypic trajectories analyses medial expansion of both the medial epicondyle and medial condyle (Figures S2 and S3). In general, both spe- In the PCA of shape data, the first two PCs jointly cies showed similar morphological changes between explained 45.62% of overall variance (PC1 = 34.01%, juvenile-subadult and adult stages (Figures S2a,b and PC2 = 11.61%) (Figure 4a). Each of the remaining PCs S3a,b). In all age classes, both in shape data and in non- accounted for less than 6.7% of the variation (Table S6) allometric shape data, Procrustes distances between spe- and did not show a clear association with factors of inter- cies were significant (Table 5). These distances were est. The shape changes explained by PC1 were mainly always higher in nonallometric shape data. associated with age so that the youngest specimens Comparisons between the phenotypic trajectories showed positive values while the oldest ones showed neg- (Figure 5) revealed significant differences between species ative values. The PCA that characterizes the general in the direction (angle) of shape change (θ = 29.45 ; shape of the femur among specimens showed that youn- p < .001). However, neither the size (magnitude) ger individuals have high values of PC1 (Figure 4a). (Δd = 0.009; p = .36 [dA. sapidus = 0.129; dA. scherman = Increasing scores of PC1 involved a medial expansion of 0.138]) nor the shape of the phenotypic trajectories the femur's head (verifiable by a medial shift of the fovea (Dp = 0.21; p = .08) were significantly different between capitis), distal expansion of the lesser trochanter, lateral taxa. For PC1, both species trajectories run from lower expansion of the greater trochanter, shortening of the values as juveniles to higher values as adults. Along PC2 third trochanter, and a general expansion of the distal there is a separation between trajectories, showing epiphysis. Specifically, the latter change included a A. sapidus the highest scores. The variation detected in medial expansion of medial condyle and medial epi- shape data by a PTA without age classes 0 and 1 was simi- condyle, and a lateral expansion of lateral condyle and lar to that observed when considering all age classes lateral epicondyle (Figure 4b; Figure S1). The PC2 (Table S7). DURÃO ET AL. 7 T A B L E 4 Results of the Procrustes Effect df SS MS Rsq F Z p ANOVA testing the effects of logarithm of centroid size (LCS), species, and the LCS 1 0.135 0.135 0.254 73.899 5.319 <.01 interaction between them on shape Species 1 0.062 0.062 0.117 33.945 7.230 <.01 conducted on the entire sample LCS  species 1 0.008 0.008 0.016 4.615 4.732 <.01 Residuals 178 0.325 0.002 0.613 Total 181 0.530 Abbreviations: df, degrees of freedom; F, F statistic; MS, mean squares; p, p-value; Rsq, R-squared values; SS, sum of squares; Z, effect-size. (a) 0.06 Age 4 | DISCUSSION class Asap Asch 0 1 2 3 4 4.1 | Common ontogenetic phenotypic 0.03 5 changes in the femur of semiaquatic and PC 2 (11.61%) fossorial water voles 0.00 In concordance with our hypothesis, the PCA revealed that the general pattern of femur shape variation during –0.03 postnatal ontogeny is similar in A. sapidus and A. scherman. In particular, important shared morphologi- –0.06 cal changes were observed in the epiphyses. The first PC –0.09 –0.06 –0.03 0.00 0.03 0.06 0.09 showed that the proximal and distal epiphysis are rela- PC 1 (34.01%) tively larger in young individuals than in adult ones. (b) These features likely provide a more multiaxially mobile hindlimb to juvenile voles (Salton & Sargis, 2009). Results obtained in an experimental study on mice (Serrat, Lovejoy, & King, 2007) allow us to suggest that differ- ences in the epiphyses shape between juvenile and adult water voles could be associated to age-related differences in the expression levels of several growth factors. In addition, the femur is more robust in juvenile than in adult individuals. This is due, at least in part, to the increase in body mass that determines higher mechanical (c) loads on the limb long bones (García & da Silva, 2006). A similar growth pattern was reported for the postweaning ontogeny of the humerus (Dur~ao et al., 2020), according to which, regardless of the type of locomotion, both bones are more robust in young than in adult water voles. In rodents, limb long bones of juvenile individuals are gener- ally composed of faster deposition tissues and poorly min- eralized woven (Enlow, 1962; Montoya-Sanhueza & Chinsamy, 2018). In contrast, adult specimens are charac- terized by having a mature compact bone, formed by F I G U R E 4 (a) Scatterplot of the first two principal components lamellar bone that acts as a fiber-reinforced composite (PC1 and PC2) from the principal component analysis performed on material (Enlow, 1962; Montoya-Sanhueza & Chinsamy, Procrustes coordinates. (b, c) Shape changes associated with each 2018). Since the mechanical and structural properties of extreme of the first two principal components; gray, negative extreme the bone determine the organism's response to a mechani- of the axis; blue, positive extreme of the axis. Bones are represented cal load (Hart et al., 2017 and references therein), robust from left to right in medial, caudal, lateral, and cranial views limb long bones in young water voles might be a 8 DURÃO ET AL. T A B L E 5 Procrustes distance Procrustes distance between Arvicola sapidus and Arvicola Age class Shape data p-value* Nonallometric shape data p-value* scherman in each age class and 0 0.036 <.001 0.109 <.001 associated p-values 1 0.036 <.001 0.051 <.001 2 0.037 <.001 0.089 <.001 3 0.040 <.001 0.096 <.001 4 0.037 <.001 0.066 <.001 5 0.048 <.001 0.069 <.001 *After sequential Bonferroni correction. compensatory response to minimize the risk of fractures et al., 2020) (Table S7). No significant difference in the during initial locomotor efforts. As pointed out for other length of the vectors was found, which means that the mammal species (Carrier, 1983; Echeverría et al., 2014; amount of shape changes per size unit is equal in both Montoya-Sanhueza et al., 2019; Young, Férnandez, & taxa. The fact that both species have a similar magnitude Fleagle, 2010), this response in Arvicola is probably related of allometric shape changes in the femur may be due to to the weak bone matrix and lower muscular development the important involvement of the femur in corresponding characteristic of juvenile individuals. types of locomotion from an early age. Specifically, very young individuals of A. scherman quickly show an identi- cal digging behavior to adults (Airoldi et al., 1976). To our 4.2 | Differences in the postweaning knowledge, the age at which southern water voles begin growth of the femur between semiaquatic to swim has not been reported but personal observations and fossorial water voles (J.V.) suggest that this activity appears not too late after weaning. Although the mechanical action of the femur is 4.2.1 | Size change different in digging and swimming, functional pressure can be alike (Salton & Sargis, 2009; Samuels, Meachen, & The femur size of A. scherman is significantly smaller than Sakai, 2013; Samuels & Valkenburgh, 2008; Smith & that of A. sapidus across the whole postnatal ontogeny, Savage, 1956) contributing to a similar magnitude of the which is congruent with the body size relationship between femoral shape change during postnatal ontogeny in these both taxa. The relatively smaller body size of A. scherman species. The fact that the mandible and humerus play a appears to be an adaptive response to hypogeic life (Cubo, preponderant role in the digging process makes the func- Ventura, & Casinos, 2006; Dur~ao et al., 2019, 2020). Given tional pressure produce important morphological changes that the burrowing cost is proportional to the volume of soil in these bones in a short time. Since in A. sapidus these dug (Vleck, 1979), it is usually assumed that a decrease in structures are not subjected to the same functional pres- body size achieves a reduction in burrow cross-section, and sure, their age-related changes are not so rapid. Although, consequently a reduction of the locomotion cost. Since the in semiaquatic species, such as A. sapidus, the femur plays digging activity requires high energy, a decrease in body size an important role in swimming, in the tibia and metatar- becomes an advantage for fossorial animals (White, 2005) sals there are key changes associated with the generation allowing them to channel energy more efficiently. of propulsion during the power stroke and with paddling during swimming (Samuels & Valkenburgh, 2008). This could justify why, unlike the femur, in the mandible 4.2.2 | Shape change (Dur~ao et al., 2019) and the humerus (Dur~ao et al., 2020) the length of the vectors also differed significantly between The analysis of allometric trajectories revealed a signifi- species. cant difference in the direction of the vectors of allome- It has been suggested that time is needed for changes tric coefficients between A. sapidus and A. scherman, in the direction of allometric trajectories to take place which indicates that allometry is a significant component (Voje, Hansen, Egset, Bolstad, & Pélabon, 2013) due to the of the femoral shape variation. Significant interspecific high costs to growth dynamics (Gould, 1966). However, differences in the direction of the vectors of allometric several works have shown that ontogenetic trajectories coefficients have been previously reported in the mandi- themselves can evolve, being prone to selection and evolu- ble (Dur~ao et al., 2019) and the humerus (Dur~ao tionary change (Adams & Nistri, 2010; Gray et al., 2019; DURÃO ET AL. 9 (a) In all age classes, the shape of the femur differed sig- Age 0.04 class Asap Asch 0 1 nificantly between A. sapidus and A. scherman. Their 2 3 4 phenotypic trajectories showed a different direction, indi- 5 0.02 cating that not only the shape of the bone differs in each PC 2 (18.17%) age class, but also that shape changes occurring during 0.00 postnatal ontogeny are not concordant in the two species. The angle between phenotypic trajectories was quite sim- –0.02 ilar to that observed in the allometric trajectories, which may suggest that a significant amount of the variation in –0.04 femur shape is allometric. Similar results were also –0.075 –0.05 –0.025 0.00 0.025 0.05 0.075 PC 1 (68.66%) reported for the mandible of these species (Dur~ao (b) et al., 2019). In the humerus, in addition to the interspe- cific differences in the direction of the corresponding phenotypic trajectories, the length of these was also sig- nificantly different between both taxa, evidencing a larger shape change during the postnatal growth of this bone in A. scherman than in A. sapidus (Dur~ao et al., 2020). The relative amount of shape variation explained by the factor species was also greater in the humerus than in the femur (compare results from Table 4 with those of table 4 in Dur~ao et al., 2020). As results of the PCAs revealed, while in the humerus, differences between species were (c) mainly explained by PC1, which accounted for 31% of total shape variation (Dur~ao et al., 2020) in the femur dif- ferences between species were mostly explained by PC2, which accounted for less than 12% of the total shape vari- ation. These results seem to indicate that functional pres- sures are related to the type of locomotion and that, depending on the role of each bone, the shape changes occurring during ontogeny may be more or less pro- nounced. The more marked shape changes of the humerus during postnatal ontogeny are concurrent with the prominent action of this bone in the digging process, and with the fact that this is a biomechanically and ener- F I G U R E 5 Ontogenetic phenotypic trajectories of Arvicola getically very demanding activity (White, 2005). sapidus and Arvicola scherman derived from the phenotypic In semiaquatic animals, relatively major morphological trajectory analysis. (a) Shape trajectories projected onto the first changes in the distal bones of hindlimbs are expected, tak- two principal components (PC1 and PC2) (based on covariance ing into account their involvement in producing propulsive matrix of group means). White dots represent the beginning (age class 0) and the black dots the end (age class 5) of the trajectories. forces during swimming (Samuels & Valkenburgh, 2008). The lines define the trajectories and the larger dots to the mean of To test this hypothesis, it would be interesting, for example, each age class. (b, c) Shape changes associated with each extreme of to perform comparative morphological analyses by means the first two principal components; gray, negative extreme of the of geometric morphometrics on the postnatal ontogeny of axis; blue, positive extreme of the axis. Bones are represented from the tibia in fossorial and semiaquatic water voles, especially left to right in medial, caudal, lateral, and cranial views bearing in mind that certain evolutionary changes in the form of this bone have been reported in Arvicola by using Wilson & Sanchez-Villagra, 2010). Our results suggest that linear measurements (Cubo et al., 2006). ontogeny may be more flexible than previously thought and that locomotion mode is an important selective force modelling the allometric trajectory of the involved bones. 4.3 | Biomechanical consequences of The different functional pressures exerted on the femur in femur shape the two studied species may have led to phenotypic varia- tion and consequently to differences between the Locomotion imposes strong demands on the musculo- corresponding allometric trajectories. skeletal system, which may affect bone morphology and 10 DURÃO ET AL. the organization of attached musculature (Biewener & morphological features than we have observed in Patek, 2018). Our results on the femur of A. sapidus and A. scherman had already been described in other diggers, A. scherman show that the morphological distances such as species of the genus Tympanoctomys, scratch-dig- between these species were significantly different in all ger/chisel-tooth diggers (Pérez, Barquez, & Díaz, 2017), age classes. Distances were greater in the nonallometric and representatives of the family Bathyergidae (Sahd, shape data than shape data, which reveal that these taxa Bennett, & Hotzé, 2019). This trochanter provides the shared part of allometric shape changes. Several interspe- insertion site for the iliopsoas complex, the strongest cific differences in the adult femur shape were also flexor of the hip, that has an important function in stand- observed, the most conspicuous being the extension of ing, walking, and running, namely in the recovery phase. the greater and third trochanter, and the position of the In A. scherman, the well-developed lesser trochanter fovea capitis. could indicate a large insertion area for this complex. Its postero-medial protraction could be related to para- sagittal movements and a strong lateral rotation of the 4.3.1 | Fossorial locomotion femur considering that the medial position of the lesser trochanter increases the lever arm for the iliopsoas mus- The morphology of the third trochanter is highly variable cle (Argot, 2002; Taylor, 1976; Toledo, Bargo, & in the mammalian femur, in terms of its presence/absence, Vizcaíno, 2015). Arvicola scherman also shows a slight lateral expansion, and proximo-distal length (Salton & variation on the femur head angle in relation to Sargis, 2009). Both A. scherman and A. sapidus show this A. sapidus, being more medially in the latter species. tubercle but in the former species, it is wider and much These differences could be linked to the restricted planes distal. As described for example in the capybara (Hydro- of motion (Anemone & Covert, 2000; Salton & choerus hydrochaeris), in this part of the femur inserts the Sargis, 2009; Szalay & Sargis, 2001) and consequently gluteus superficialis muscle (García-Esponda & associated with differences in hip mobility between the Candela, 2016), which originates from the lumbodorsal two species. The femur of A. scherman exhibits a cranial- fascia and that helps flex the hip joint. As has been medial shift of the fovea capitis related to A. sapidus. In reported for several mammal species, the mass of this this dimple attach the teres femoris ligament, one of the muscle seems to have noticeable effects on the size, posi- large ligaments that connect the femoral head to the hip. tion, and shape of the third trochanter (Anemone & The relatively great extension of the fovea capitis in Covert, 2000; García-Esponda & Candela, 2016; Salton & A. scherman suggests a strong ligament and a greater sta- Sargis, 2009; Sargis, 2002; Smith & Savage, 1956). In partic- bilization of the hip and knee. The general morphology ular, a wider third trochanter is linked to the powerful of the lesser trochanter in A. scherman is consistent with development of the gluteus superficialis muscle (Smith & the ability for extension and flexion of the hip as well as, Savage, 1956). This feature may provide a mechanical abduction and external rotation of the femur. These advantage because it increases the efficiency of the movements are favored by the presence of a strong liga- moment arm as well as provides a great lever for gluteus ment that helps to stabilize the hip when vigorously kicks. superficialis muscle (Salton & Sargis, 2009; Sargis, 2002). A In turn, the medial position of the femur head in robust third trochanter also contributes to the hip stabili- A. scherman may limit hip mobility at parasagittal motion, zation during the extension of the leg and may be reduced while maintaining the capacity for extension and flexion. bending stresses. This trait has also been found in other These morphological characteristics in A. scherman are fossorial mammals (Salton & Sargis, 2009), such as the probably related to the digging process, specifically, to the Hispaniolan solenodon (Solenodon paradoxus), the mole- soil removal, in which the angle between the pelvis and like rice tenrec (Oryzorictes hova), the four-toed rice tenrec the femur diminishes, and the hip is extended backward (Oryzorictes tetradactylus), and the naked mole-rat (Hetero- to sweep loose soil farther away from the burrow. cephalus glaber). As suggested by Salton and Sargis (2009), Another distinctive feature of A. scherman with the sturdiness of this trochanter permits a powerful thigh respect to A. sapidus is the existence of a more robust extension with some abduction, action related to throwing femur, which is congruent with the results published in back the loose earth to kicking it out of the burrow. Since previous studies based on linear measurements (Cubo in A. scherman the hindlimbs play a relatively important et al., 2006; Ventura, 1988, 1990, 1992). This typical trait role in the digging process (Airoldi et al., 1976; of fossorial rodents (Biknevicius, 1993; Casinos, Laville, 1989, 1989b), a stronger flex of the hip could be Quintana, & Viladiu, 1993; Samuels & Valkenburgh, linked to the corresponding mechanical requirement. 2008; Wilson & Geiger, 2015) is related to withstand the The lesser trochanter of A. scherman is oriented bending and torsion forces placed on the limbs during postero-medially and is larger than in A. sapidus. Similar the digging process (Biknevicius, 1993; Stein, 2000). The DURÃO ET AL. 11 robustness of the femur in A. scherman likely results starting the power phase. This morpho-functional relation- from both the support that the bone needs for the devel- ship remains to be tested in A. sapidus. opment of muscles and as a response to the mechanical stress that hindlimbs are continuously subjected to. Compared to A. sapidus, A. scherman exhibits a wide 5 | CONCLUSIONS distal epicondyle which provides a great surface area for the insertion of the gastrocnemius and soleus muscles, Results obtained in the present study reveal a combination of plantar flexors of the foot (García-Esponda & phylogenetic and functional signals in the postnatal growth Candela, 2016). The increase in the epicondylar area may of the femur in A. sapidus and A. scherman. On the one aid in resisting the tendency of the body to be pushed hand, shared common traits between these taxa were found, backward while digging, as suggested by Samuels and which can be considered traces of their common origin. On Valkenburgh (2008) for other chisel-tooth digger rodents the other hand, we found evidence that certain variations in species (Bathyergids and Spalacids). the femur shape and particular differences in allometric and phenotypic trajectories between these taxa are associated with their different types of locomotion. Our study emphasizes the 4.3.2 | Aquatic locomotion importance of developing comparative studies in closely related species to assess to which extent functional and phylo- Swimming imposes challenges for semiaquatic species genetic factors affect the growth patterns of bones involved in because water is a dense and viscous medium that estab- locomotion. As for genus Arvicola in particular, it would be lishes a strong hydrodynamic demand on the musculo- interesting to extend this kind of studies to other semiaquatic skeletal system (Gillis & Blob, 2001). This requirement can or fossorial water vole populations, and even others that can lead to morphological changes that allow improving loco- change the type of locomotion according to the reproduction motor performance and stability, such as increased speed, cycle and/or the environmental conditions (Barreto & drag reduction, improved thrust output, and increased MacDonald, 2000; Kratochvíl & Grulich, 1961; Potapov, maneuverability (Fish, 2016; Fish & Stein, 1991 and refer- Nazarova, Muzyka, Potapova, & Evsikov, 2012; Stewart, ences therein). Different adaptations of the hindlimbs Jarrett, Scott, White, & McCafferty, 2019; Telfer et al., 2003). seem to have been developed by different quadrupedal paddling mammals with semiaquatic habits (Botton-Divet, ACKNOWLEDGMENTS Cornette, Fabre, Herrel, & Houssaye, 2016; García- We would like to thank Dr. David Galicia for access to Esponda & Candela, 2016; Stein, 1988). These morphologi- specimens at the Museo de Ciencias Naturales of the cal changes seem to be related to the amount of time spent Universidad de Navarra. We also thank Joan Jené for his by the animal in water, and to the activities performed assistance in figure production. The study was conducted there (Dunstone, 1979; Stein, 1988). Semiaquatic animals, in the framework of the PhD program in Biodiversity from due to their great dependence on land once compared Universitat Autònoma de Barcelona and was funded by to their terrestrial relatives, display few or absent Comissionat per a Universitats i Recerca, Generalitat de limb specializations (García-Esponda & Candela, 2016; Catalunya (grant number 2014-SGR-1241); A. F. D. Thewissen & Taylor, 2007). This moderate morphological received a PIF grant from UAB. variation appears to be associated with functional needs both on land and in water (Stein, 1989). As for Arvicola, CONFLICT OF INTEREST we observed that A. sapidus shows, in comparison with The authors declare no potential conflicts of interest. A. scherman, a slight medial-cranial extension of the greater trochanter. This tubercle provides the insertion AUTHOR CONTRIBUTIONS area for the gluteus piriformis and medius muscles, both Ana Filipa Dura~o: Formal analysis (lead); investigation extensors and abductors of the hip joint (García-Esponda & (lead); methodology (equal); visualization (lead); writing Candela, 2016). The modest expansion of the greater tro- – original draft (lead). Francesc Muñoz-Muñoz: chanter in A. sapidus could be related to the relative size Conceptualization (equal); methodology (equal); writing increase of the gluteus medius muscle, feature also found – review and editing (equal). Jacint Ventura: Conceptu- in other semiaquatic rodent species (Candela & alization (equal); resources (lead); supervision (lead); Picasso, 2008; Stein, 1988). This muscle is involved in writing – review and editing (equal). rapid movement toward the end of the stroke (Smith & Savage, 1956). As suggested by Stein (1988), the size DA TA AVAI LA BI LI TY S T ATE ME NT increase of the gluteus medius muscle allows decreasing The data generated by this study are available in the the resistance to water, maintaining the impulse after Zenodo repository (10.5281/zenodo.5155457). 12 DURÃO ET AL. ORCID Casinos, A., Quintana, C., & Viladiu, C. (1993). Allometry and adapta- ~ Ana Filipa Durao https://orcid.org/0000-0003-0859- tion in the long bones of a digging group of rodents (Ctenomyinae). 5371 Zoological Journal of the Linnean Society, 107, 107–115. Chevret, P., Renaud, S., Helvaci, Z., Ulrich, R. G., Quéré, J., & Michaux, J. R. Francesc Muñoz-Muñoz https://orcid.org/0000-0003- (2020). Genetic structure, ecological versatility, and skull shape differen- 2084-0368 tiation in Arvicola water voles (Rodentia, Cricetidae). Journal of Zoology Jacint Ventura https://orcid.org/0000-0001-7527-1532 Systematics and Evolutionary Research, 58, 1323–1334. Collyer, M. L., & Adams, D. C. (2007). Analysis of two-state multivari- R EF E RE N C E S ate phenotypic change in ecological studies. Ecology, 88, 683–692. Adams, D. C., & Collyer, M. L. (2007). Analysis of character diver- Collyer, M. L., & Adams, D. C. (2013). Phenotypic trajectory analy- gence along environmental gradients and other covariates. Evo- sis: Comparison of shape change patterns in evolution and ecol- lution, 61, 510–515. ogy. Hystrix, 24, 75–83. Adams, D. C., & Collyer, M. L. (2009). A general framework for the Collyer, M. L., Sekora, D. J., & Adams, D. C. (2015). A method for analysis of phenotypic trajectories in evolutionary studies. Evo- analysis of phenotypic change for phenotypes described by lution, 63, 1143–1154. high-dimensional data. Heredity, 115, 357–365. Adams, D. C., & Nistri, A. (2010). Ontogenetic convergence and Collyer, M. L., & Adams, D. C. (2018). RRPP: An r package for fitting evolution of foot morphology in European cave salamanders linear models to high-dimensional data using residual randomi- (Family: Plethodontidae). BMC Evolutionary Biology, 10, 216. zation. Methods in Ecology and Evolution, 9, 1772–1779. Adams, D. C., Collyer, M. L., Kaliontzopoulou, A., & Sherratt, E. Cubo, J., Ventura, J., & Casinos, A. (2006). A heterochronic inter- (2017). Geomorph: Software for geometric morphometric ana- pretation of the origin of digging adaptations in the northern lyses. R package version 3.0.5. water vole, Arvicola terrestris (Rodentia: Arvicolidae). Biological Airoldi, J. P. (1976). Le terrier de la forme fouisseuse du campagnol Journal of the Linnean Society, 87, 381–391. terrestre (Arvicola terrestris scherman Shaw) (Mammalia, Dryden, I., & Mardia, K. V. (1998). Statistical shape analysis. Chich- Rodentia). Zeitschrift für Säugetierkunde, 41, 23–42. ester, England: Wiley. Airoldi, J. P., Altrocchi, R., & Meylan, A. (1976). Le comportement Dunstone, N. (1979). Swimming and diving behavior of the mink fouisseur du campagnol terrestre, Arvicola terrestris scherman (Mustela vison Schreber). Carnivore, 2, 56–61. Shaw (Mammalia, Rodentia). Revue Suisse Zoologie, 83, 282–286. Dur~ao, A. F., Muñoz-Muñoz, F., Martínez-Vargas, J., & Ventura, J. Airoldi, J. P., & De Werra, D. (1993). The burrow system of the fos- (2018). Obtaining three-dimensional models of limb long bones sorial form of the water vole (Arvicola terrestris scherman Shaw) from small mammals: A photogrammetric approach. In C. (Mammalia, Rodentia): An approach using graph theoretical Rissech, L. Lloveras, J. Nadal, & J. M. Fullola (Eds.), Geometric methods and simulation models. Mammalia, 57, 423–433. morphometrics. Trends in biology, paleobiology and archaeology Anemone, R. L., & Covert, H. H. (2000). New skeletal remains of (pp. 125–138). Barcelona, Spain: SERP - Universitat Barcelona. Omomys (Primates, Omomyidae): Functional morphology of Dur~ao, A. F., Ventura, J., & Muñoz-Muñoz, F. (2019). Comparative the hindlimb and locomotor behavior of a Middle Eocene pri- post-weaning ontogeny of the mandible in fossorial and semi- mate. Journal of Human Evolution, 38, 607–633. aquatic water voles. Mammalian Biology, 97, 95–103. Argot, C. (2002). Functional-adaptive analysis of the hindlimb anat- Dur~ao, A. F., Muñoz-Muñoz, F., & Ventura, J. (2020). Three-dimensional omy of extant marsupials and the paleobiology of the Paleocene geometric morphometric analysis of the humerus: Comparative marsupials Mayulestes ferox and Pucadelphys andinus. Journal of postweaning ontogeny between fossorial and semiaquatic water Morphology, 256, 76–108. voles (Arvicola). Journal of Morphology, 281, 1679–1692. Barreto, G. R., & MacDonald, D. W. (2000). The decline and local Echeverría, A. I., Becerra, F., & Vassallo, I. (2014). Postnatal ontog- extinction of a population of water voles, Arvicola terrestris, in eny of limb proportions and functional indices in the subterra- southern England. Zeitschrift Für Säugetierkunde, 65, 110–120. nean rodent Ctenomys talarum (Rodentia: Ctenomyidae). Biewener, A. A., & Patek, S. N. (2018). Animal locomotion. Oxford: Journal of Morphology, 275, 902–913. Oxford University Press. Elissamburu, A., & Vizcaíno, S. F. (2004). Limb proportions and Biknevicius, A. R. (1993). Biomechanical scaling of limb bones and adaptations in caviomorph rodents (Rodentia: Caviomorpha). differential limb use in caviomorph rodents. Journal of Mam- Journal of Zoology, 262, 145–159. malogy, 74, 95–107. Enlow, D. H. (1962). A study of the post-natal growth and remo- Bookstein, F. L. (1991). Morphometric tools for landmark data: deling of bone. American Journal of Anatomy, 110, 79–101. Geometry and biology. New York, NY: University Press. EOS Systems Inc. (2015). Photomodeler Scanner (v2015.1.1). Van- Botton-Divet, L., Cornette, R., Fabre, A., Herrel, A., & Houssaye, A. couver, Canada: EOS Systems. (2016). Morphological analysis of long bones in semi-aquatic Esquerré, D., Sherratt, E., & Keogh, J. S. (2017). Evolution of extreme mustelids and their terrestrial relatives. Integrative and Com- ontogenetic allometric diversity and heterochrony in pythons, a parative Biology, 56, 1298–1309. clade of giant and dwarf snakes. Evolution, 71, 2829–2844. Candela, A. M., & Picasso, B. J. (2008). Functional anatomy of the limbs of Falkingham, P. L. (2012). Acquisition of high resolution 3D models Erethizontidae (rodentia, caviomorpha): Indicators of locomotor using free, open-source, photogrammetric software. Palaeontologia behavior in Miocene porcupines. Journal of Morphology, 269, 552–593. Electronica, 15, 1–15. Carrier, D. R. (1983). Postnatal ontogeny of the musculo-skeletal Fish, F. E., & Stein, B. R. (1991). Functional correlates of differences system in the black-tailed jack rabbit (Lepus californicus). Jour- in bone density among terrestrial and aquatic genera in the nal of Zoology, 201, 27–55. family Mustelidae (Mammalia). Zoomorphology, 110, 339–345. DURÃO ET AL. 13 Fish, F. E. (2016). Secondary evolution of aquatic propulsion in extinct ctenomyid and octodontid rodents. Biological Journal of higher vertebrates: Validation and prospect. Integrative and the Linnean Society, 95, 267–283. Comparative Biology, 56, 1285–1297. Mahmoudi, A., Maul, L. C., Khoshyar, M., Darvish, J., García, G. J. M., & da Silva, J. K. L. (2006). Interspecific allometry Aliabadian, M., & Kryštufek, B. (2020). Evolutionary history of of bone dimensions: A review of the theoretical models. Physics water voles revisited: Confronting a new phylogenetic model from of Life Reviews, 3, 188–209. molecular data with the fossil record. Mammalia, 84, 171–184. García-Esponda, C. M., & Candela, A. M. (2016). Hindlimb muscu- Mate, I., Barrull, J., Salicrú, M., Ruiz-Olmo, J., & Gosàlbez, J. (2013). lature of the largest living rodent Hydrochoerus hydrochaeris Habitat selection by southern water vole (Arvicola sapidus) in (Caviomorpha): Adaptations to semiaquatic and terrestrial riparian environments of Mediterranean mountain areas: A con- styles of life. Journal of Morphology, 277, 286–305. servation tool for the species. Acta Theriologica, 58, 25–37. Garde, J. M. (1992). Biología de la rata de agua Arvicola sapidus Monteiro, L. (1999). Multivariate regression models and geometric Miller, 1908 (Rodentia, Arvicolidae) en el sur de Navarra morphometrics: The search for causal factors in the analysis of (España). (PhD thesis). Universidad de Navarra, Navarra. shape. Systematic Biology, 48, 192–199. Garde, J., Escala, M. C., & Ventura, J. (1993). Determinaci on de la edad Montoya-Sanhueza, G., & Chinsamy, A. (2018). Cortical bone adaptation relativa en la rata de agua meridional, Arvícola sapidus Miller, 1908 and mineral mobilization in the subterranean mammal Bathyergus (Rodentia, Arvicolidae). Doñana, Acta Vertebrata, 20, 266–276. suillus (Rodentia: Bathyergidae): Effects of age and sex. PeerJ, 6, e4944. Gillis, G. B., & Blob, R. W. (2001). How muscles accommodate Montoya-Sanhueza, G., Wilson, L. A. B., & Chinsamy, A. (2019). movement in different physical environments: Aquatic vs ter- Postnatal development of the largest subterranean mammal (Bat- restrial locomotion in vertebrates. Comparative Biochemistry hyergus suillus): Morphology, osteogenesis and modularity of the and Physiology - Part A, 131, 61–75. appendicular skeleton. Developmental Dynamics, 248, 1101–1128. Goodall, C. (1991). Procrustes methods in the statistical analysis of Musser, G. G., & Carleton, M. C. (2005). Arvicola Lacépède, 1799; shape. Journal of the Royal Statistical Society, 53, 285–339. Arvicola amphibius (Linnaeus, 1758); Arvicola sapidus Miller, Gould, S. J. (1966). Allometry and size in ontogeny and phylogeny. 1908; Arvicola scherman (Shaw, 1801). In D. E. Wilson & D. M. Biological Reviews, 41, 587–640. Reeder (Eds.), Mammal species of the world. A taxonomic and Gray, J. A., Sherratt, E., Hutchinson, M. N., & Jones, M. E. H. (2019). geographic reference (pp. 963–966). Baltimore, MD: Johns Hop- Changes in ontogenetic patterns facilitate diversification in skull shape kins University Press. of Australian agamid lizards. BMC Evolutionary Biology, 19, 1–10. Pardiñas, U. F. J., Myers, P., Le on-Paniagua, L., Ord oñez-Garza, N., Hart, N. H., Nimphius, S., Rantalainen, T., Ireland, A., Siafrikas, A., & Cook, J. A., Kryštufek, B., … Patton, J. L. (2017). Family Newton, R. U. (2017). Mechanical basis of bone strength: Influence Cricetidae (true hamsters, voles, lemmings and new world rats of bone material, bone structure and muscle action. Journal of and mice). In D. E. Wilson, T. E. Lacher, Jr., & R. A. Musculoskeletal and Neuronal Interaction, 17, 114–139. Mittermeier (Eds.), Handbook of the mammals of the world. Vol. Hildebrand, M. (1985). Digging of quadrupeds. In M. Hildebrand, 7. Rodents II (pp. 204–535). Lynx Edicions: Barcelona, Spain. D. M. Bramble, K. F. Liem, & D. B. Wake (Eds.), Functional ver- Pérez, M. J., Barquez, R. M., & Díaz, M. M. (2017). Morphology of tebrate morphology (pp. 89–109). Cambridge, MA: Belknap Press. the limbs in the semi-fossorial desert rodent species of Klingenberg, C. P. (2011). MorphoJ: An integrated software package for Tympanoctomys (Octodontidae, Rodentia). ZooKeys, 710, geometric morphometrics. Molecular Ecology Resources, 11, 353–357. 77–96. Klingenberg, C. P. (2016). Size, shape, and form: Concepts of allom- Potapov, M. A., Nazarova, G. G., Muzyka, V. Y., Potapova, O. F., & etry in geometric morphometrics. Development Genes and Evo- Evsikov, V. I. (2012). Extrisic and intrinsic factors of regulations lution, 226, 113–137. of reproductive potential in the water vole (Arvicola amphibius) Kratochvíl, J., & Grulich, I. (1961). On the distribution and habitat population from western Siberia. In A. Triunveri & D. Scalise requeriments in the water-vole Arvicola terrestris, in Czechoslo- (Eds.), Rodents - Habitats, pathology and environmental impact vakia. Zoologicke Listy, 10, 265–280. (pp. 23–41). New York, NY: Nova Science Publishers. Kryštufek, B., Koren, T., Engelberger, S., Horvath, G. F., Purger, J. J., Quéré, J., & Le Louarn, H. (2011). Les rongeurs de France. Arslan, A., … Murariu, D. (2015). Fossorial morphotype does not  Faunistique et biologie. Versailles, France: Editions Quae. make a species in water voles. Mammalia, 79, 293–303. R Development Core Team (2016). R: A language and environment for Laville, E. (1989). Etude cinématique du fouissage chez Arvicola ter- statistical computing. Version 3.5.1. Vienna, Austria: R Foundation restris scherman (Rodentia, Arvicolidae). Mammalia, 53, 177–190. for Statistical Computing Retrieved from www.R-project.og Laville, E. (1989b). Etude morphofonctionelle comparative des Rice, W. R. (1989). Analyzing tables of statistical tests. Evolution, 43, structures osseuses impliquées dans le fouissage d'Arvicola ter- 223–225. restris scherman (Rodentia, Arvicolidae). Canadian Journal of Reichstein, H. (1963). Beitrag zur systematischen Gliederung des Zoology, 68, 2437–2444. Genus Arvicola Lacépède 1799. Zeitschrift für Zoologische Laville, E., Casinos, A., Gasc, J. P., Renous, S., & Bou, J. (1989). Les Systematik und Evolutionsforschung, 1, 155–204. mécanismes du fouissage chez Arvicola terrestris et Spalax Reichstein, H. (1982). Gattung Arvicola Lacépède 1799 Schermäuse. In ehrenbergi: étude fonctionelle et évolutive. Anatomischer J. Niethammer & F. Krapp (Eds.), Handbuch der Säugetiere Anzeiger, 169, 131–144. Europas Vol. 2 / 1 (pp. 209–252). Akademische Verlagsgesellschaft: Lehmann, W. H. (1963). The forelimb architecture of some fossorial Wiesbaden, Germany. rodents. Journal of Morphology, 113, 59–76. Rohlf, F. J., & Slice, D. (1990). Extensions of the Procrustes method Lessa, E. P., Vassallo, A. I., Verzi, D. H., & Mora, M. S. (2008). Evo- for the optimal superimposition of landmarks. Systematic Zool- lution of morphological adaptations for digging in living and ogy, 39, 40–59. 14 DURÃO ET AL. Sahd, L., Bennett, N. C., & Hotzé, S. H. (2019). Hind foot drum- santacrucian (Early Miocene) sloths (Xenarthra, Folivora) of ming: Morphological adaptations of the muscles and bones of Patagonia. The Anatomical Records, 298, 842–864. the hind limb in three African mole-rat species. Journal of Vassallo, A. I. (1998). Functional morphology, comparative behaviour, and Anatomy, 235, 811–824. adaptation in two sympatric subterranean rodents genus Ctenomys Salton, J. A., & Sargis, E. J. (2009). Evolutionary morphology of the (Caviomorpha: Octodontidae). Journal of Zoology, 244, 415–427. tenrecoidea (Mammalia) hindlimb skeleton. Journal of Mor- Ventura, J. (1988). Contribucion al conocimiento del género Arvicola phology, 270, 367–387. Lacépède, 1799, en el nordeste de la Peninsula Ibérica. (PhD the- Samuels, J. X., & Valkenburgh, B. V. (2008). Skeletal indicators of sis). Universidad de Barcelona, Barcelona. locomotor adaptations in living and extinct rodents. Journal of Ventura, J. (1990). Datos biométricos sobre los huesos largos y la Morphology, 269, 1387–1411. escapula de Arvicola sapidus Miller, 1908 (Rodentia, Samuels, J. X., Meachen, J. A., & Sakai, S. A. (2013). Postcranial Arvicolidae). Boletín de la Real Sociedad Española de Historia morphology and the locomotor habits of living and extinct car- Natural (Seccion Biologica), 86, 1–4. nivorans. Journal of Morphology, 274, 121–146. Ventura, J. (1992). Morphometric data on the scapula and limb long Sargis, E. J. (2002). Functional morphology of the hindlimb of bones of Arvicola terrestris (Linnaeus, 1758) (Rodentia, Tupaiids (Mammalia, Scandentia) and its phylogenetic implica- Arvicolidae). Revue Suisse Zoologie, 99, 629–636. tions. Journal of Morphology, 254, 149–185. Ventura, J. (1993). Crecimiento relativo de Arvicola terrestris mont- Serrat, M. A., Lovejoy, C. O., & King, D. (2007). Age- and site- icola (Rodentia, Arvicolidae). Miscellania Zoologica, 17, 237–248. specific decline in insulin-like growth factor-I receptor expres- Ventura, J. (2007). Arvicola terrestris (Linnaeus, 1758); Arvicola sion is correlated with differential growth plate activity in the sapidus Miller, 1908. In L. J. Palomo, J. Gisbert, & J. C. Blanco mouse hindlimb. The Anatomical Record: Advances in Integra- (Eds.), Atlas y Libro Rojo de los Mamíferos Terrestres de España tive Anatomy and Evolutionary Biology, 290, 375–381. (pp. 401–407). Madrid, Spain: Direcci on General para la Smith, J. M., & Savage, R. J. G. (1956). Some locomotory adapta- Biodiversidad -SECEM-SECEMU. tions in mammals. Zoological Journal of the Linnean Society, Ventura, J., & Gosalbez, J. (1992). Criterios para la determinaci on 42, 603–622. de la edad relativa en Arvicola terrestris monticola (Rodentia, StatSoft. (2014). Statistica (data analysis software system). Version Arvicolidae). Miscellania Zoologica, 16, 197–206. 12. Tulsa, OK: Author. Ventura, J., & Casado-Cruz, M. (2011). Post-weaning ontogeny of Stein, B. R. (1988). Morphology and allometry in several genera of the mandible in fossorial water voles: Ecological and evolution- semiaquatic rodents (Ondatra, Nectomys, and Oryzomys). Jour- ary implications. Acta Zoologica, 92, 12–20. nal of Mammalogy, 69, 500–511. Vleck, D. (1979). The energy cost of burrowing by the pocket Stein, B. R. (1989). Bone density and adaptation in semiaquatic gopher Thomomys bottae. Physiological Zoology, 52, 122–136. mammals. Journal of Mammalogy, 70, 467–476. Voje, K. L., Hansen, T. F., Egset, C. K., Bolstad, G. H., & Stein, B. R. (1993). Comparative hind limb morphology in Pélabon, C. (2013). Allometric constraints and the evolution of Geomyine and Thomomyine pocket gophers. Journal of Mam- allometry. Evolution, 68, 866–885. White, C. R. (2005). The allometry of burrow geometry. Journal of malogy, 74, 86–94. Zoology, 265, 395–403. Stein, B. R. (2000). Morphology of subterranean rodents. In E. A. Wilson, L. A. B., & Sanchez-Villagra, M. R. (2010). Diversity trends Lacey, J. L. Patton, & G. N. Cameron (Eds.), Life underground: and their ontogenetic basis: An exploration of allometric dispar- The biology of subterranean mammals (pp. 19–61). Chicago, ity in rodents. Proceedings of the Royal Society, 277, 1227–1234. CA: University of Chicago Press. Wilson, L. A. B., & Geiger, M. (2015). Diversity and evolution of femoral Stewart, R. A., Jarrett, C., Scott, C., White, S. A., & variation in Ctenohystrica. In P. G. Cox & L. Hautier (Eds.), Evolution McCafferty, D. J. (2019). Water vole (Arvicola amphibius) of the rodents: Advances in phylogeny, functional morphology, and devel- abundance in grassland habitats of Glasgow. The Glasgow opment (pp. 510–538). Cambridge, MA: Cambridge University Press. Naturalist, 27, 10–19. Young, J. W., Férnandez, D., & Fleagle, J. G. (2010). Ontogeny of Szalay, F. S., & Sargis, E. J. (2001). Model-based analysis of postcra- long bone geometry in capuchin monkeys (Cebus albifrons and nial osteology of marsupials from the Palaeocene of Itaboraí Cebus apella): Implications for locomotor development and life (Brazil) and the phylogenetics and biogeography of Metatheria. history. Biology Letters, 6, 197–200. Geodiversitas, 23, 139–302. Taylor, M. E. (1976). The functional antomy of the hindlimbs of some African Viverridae (Carnivora). Journal of Morphology, SU PP O R TI N G I N F O RMA TI O N 148, 227–254. Additional supporting information may be found in the Telfer, S., Dallas, J. F., Aars, J., Piertney, S. B., Stewart, W. A., & online version of the article at the publisher's website. Lambin, X. (2003). Demographic and genetic structure of fosso- rial water voles (Arvicola terrestris) on Scottish islands. Journal of Zoology, 259, 23–29. How to cite this article: Dur~ao, A. F., Muñoz- Thewissen, J. G. M., & Taylor, M. A. (2007). Aquatic adaptations in Muñoz, F., & Ventura, J. (2021). Postnatal the limbs of amniotes. In B. K. Hall (Ed.), Fins into limbs - Evo- ontogeny of the femur in fossorial and semiaquatic lution, development, and transformation (pp. 310–322). Chicago, water voles in the 3D-shape space. The Anatomical CA: The University of Chicago Press. Toledo, N., Bargo, M. S., & Vizcaíno, S. F. (2015). Muscular recon- Record, 1–14. https://doi.org/10.1002/ar.24765 struction and functional morphology of the hind limb of