Introduction

Sexual dimorphism (SD) subscribing marked differences between sexes such as in morphology is a common phenomenon in the animal kingdom (e.g., Fairbairn et al., 2007). Different theories try to explain the causal relationship between selection and morphology, namely i. fecundity (e.g., Kupfer et al., 2004), ii. sexual selection (Shine, 1979) and iii. ecological niche portioning (Hedrick & Temeles, 1989, Shetty & Shine, 2002). In context with body architecture, SD can be expressed as size dimorphism (SSD) or shape dimorphism (SShD) underlying different selection processes (e.g., Schwarzkopf, 2005, Kupfer, 2007, Pogoda & Kupfer, 2018). While SSD is better studied and known in many vertebrates (e.g., Cox et al., 2007, Lindenfors et al., 2007), SShD needs much more attention as it substantial influence species ecology (e.g., Shetty & Shine, 2002, Alcorn et al., 2013, Pogoda & Kupfer, 2018, Pogoda et al., 2020).
To understand the processes leading to different morphologies between males and females, comparative studies are needed allowing conclusions from species ecology to its morphology. Amphibians are a vertebrate group with heavily female-biased SSD (e.g., Shine, 1979, Kupfer, 2007). Nevertheless, only approximately 61% of known salamander species exhibit female-biased SSD, while about 19% exhibit a male-biased SSD (Kupfer, 2007, Amat, 2019). Although less diverse in terms of species numbers, urodeles evolved various reproductive modes and behaviours including diverse life history strategies (Sparreboom, 2014, Kieren et al., 2018). As Salamandridae or true salamanders includes most of the variability in reproductive biology known from urodeles (Sparreboom, 2014, Frost, 2018), making the group well-suited for the investigation of SSD and SShD.
Pleurodeline newts, named Pleurodelini, comprise a diverse monophyletic group of salamandrid salamanders (e.g., Veith et al., 2018, Wang et al., 2018) thus, being important in the reconstruction and understanding of the evolutionary processes forming SD among Salamandridae (Pogoda & Kupfer, 2018). Among Pleurodelini monophyletic crocodile newts evolved a variety of reproductive modes and strategies (Salvador & García-París, 1999, Hernandez, 2016, Kieren et al., 2018). For mating either a ventral amplexus or a circular mating dance is performed either taking place in aquatic or terrestrial habitats. Also, female crocodile newts deposit egg clutches either in water or on land (Igawa et al., 2013, Phimmachak et al., 2015b, Pasmans et al., 2017, Gong et al., 2018, Hernandez, 2016). These differences in terms of mating and reproductive ecology can even be observed within one genus Tylototriton comprising most crocodile newt species. Some of the reproductive traits correlate with phylogeny e.g., oviposition site, while others do not e.g., mating mode. Nevertheless, both phylogeny and phylogenetic independent ecological traits leading to coevolution of similar cranial shapes among crocodile newts (Pogoda et al., 2020). So far, SD was rarely investigated in crocodile newts (Seglie et al., 2010, Phimmachak et al., 2015b). Many studies are regularly descriptive and limited to a few body measurements such as snout-vent length, cloacal size and some others (e.g., Khatiwada et al., 2015, Fei & Ye, 2016, Hernandez & Hou, 2018) often with low sample sizes. A variety of SSD and SShD patterns such as longer and wider trunks in females and longer limbs and crania in males are known in many other salamandrids (e.g., Malmgren & Thollesson, 1999, Romano et al., 2009, Amat et al., 2015, Reinhard et al., 2015, Reinhard & Kupfer, 2015, Altunışık, 2017). Consequently, it is likely that also a variety of dimorphic traits are undetected so far in crocodile newts. There is even less known about SD in the osteology of urodeles in general (Ivanović & Kalezić, 2012, Pogoda & Kupfer, 2018). Researchers just started exploring this field of morphology research accessed mainly by modern non-invasive CT technology (Broeckhoven & du Plessis, 2018). In a terrestrial salamandrid salamander, it was shown that the same patterns of SSD can be detected in the osteology as in the external morphology (Pogoda & Kupfer, 2018). Further, excluding soft tissue leads to an enormous increase of morphological structures which can be used in studying especially SShD otherwise covered. In ventral amplecting Pleurodeles , the sister taxon to the crocodile newt genera Tylototriton and Echinotriton , differentially shaped humeri between sexes are known for long time (e.g., Herre, 1952). The common ancestry of ribbed and crocodile newts and the interspecifically different reproductive strategies may imply variable SD patterns tightly linked to ecology. Understanding SSD and SShD patterns in context with phylogeny and ecology will aid understanding the evolutionary biology of salamanders. Thus, rather then just providing intersexual comparisons of measurements or shape data, these must be linked to other traits of the investigates system in order to be able to reveal potential selection sources leading to observed morphological differences.
The aim of our study was to investigate SD in crocodile newts, linking patterns to the different reproductive ecologies of the species. Our focus was laid on the cranium and fore limb morphology. Male cranial morphology was different among amplectant and dancing species (Pogoda et al., 2020). Thus, we assumed different interspecific patterns of SShD of crocodile newt cranial morphology applying different mating modes. Fore limb morphology is of special interest as the fore limb likely experiences different mechanical needs when a ventral amplexus is applied or not. We hypothesize that species which apply an amplexus during mating have a more pronounced SSD of their fore limbs than species mating without physical contact and that the different mating patterns lead also to differences in SShD between these reproductive groups. We used µCT scans of crania and humeri of crocodile newts and employed 3D geometric morphometrics (GM) to test our hypotheses.