Ana Golubović

Sexual body size and body shape dimorphism of Testudo hermanni in central and eastern Serbia

Hermanns tortoise (Testudo hermanni) is widely distributed in western and southern Europe. Most populations in the western part of the distribution range (e.g. Spain, France, Italy) are severely reduced, and relatively well studied, whilst the species is still abundant in eastern areas (i.e. the Balkans). However, essential biological information (e.g. main morphological, ecological, and behavioural characteristics) for the Balkans are still extremely limited. As reptiles exhibit strong geographic variation in most morphological, as well as life history traits, gathering data from distant areas is important. We present data from two populations of T. hermanni in Serbia, focusing on sexual dimorphism in body size and body shape. We found that almost all of the 43 morphological traits analysed were significantly different between sexes and that sexual size and sexual shape dimorphisms were not expressed in similar ways. Notably, sexual size dimorphism (SSD) was more pronounced than sexual shape dimorphism (SShD). Our analyses suggested that SShD is more stable than SSD, and that the scale of the focus (i.e. whole body proportions versus morphological details) is a key factor to test this notion. When general measurements were considered, the expected consistency of SShD between populations was verified; nevertheless, when more specific morphological attributes were considered, substantial variations were observed. These results provide a baseline for comparisons between populations to further examine geographic variation of sexual dimorphism.

Ontogenetic shift of antipredator behaviour in Hermann’s tortoises

Armoured animals generally exhibit two main antipredator responses: They either flee or stay motionless, withdrawn in their protective armour. The transition between these two threat reactions can be affected by the degree of armour sturdiness. Tortoise shell stiffness gradually increases through ontogeny due to ossification. Additionally, neonates do not benefit from parental protection. Thus, juvenile survival could rely strongly on behavioural adaptations. This experimental approach addresses the effects of age (size), morphology, sex and population of origin on the transition between the two strategies. Predator attack was simulated by overturning individuals on their backs. Juveniles displayed bolder threat response comparing to adults. They also spent shorter periods of time withdrawn in shells and inspecting surroundings. Immature tortoises from all localities had high self-righting success, contrary to adults. The deterioration of righting success coincides with age of sexual maturation. Prompt switch from hiding to fleeing strategy in threatened juvenile tortoises implies that natural selection acts strongly on their swiftness and agility. Sexes did not diverge in antipredator displays. Self-righting speed correlated with shell shape in both juveniles and adults. Morphological measurements used in this study affected self-righting speed only in adults. These effects were accompanied with a general negative effect of increase in body size. Further studies should explore how frequency of predator encounters (i.e. experience) shape antipredator behaviour of tortoises. This could have conservation implications, especially for efficient releasing of animals from captive breeding programs.

Skills or strength—how tortoises cope with dense vegetation?

Dense vegetation cover undoubtedly offers certain advantages for small and slow-moving animals, but its disadvantages concerning some aspects of spatial ecology (e.g. movements) were neglected in previous studies. Tortoises could get stuck in vegetation by protuberant part of the shell and thus succumb to overheating, dehydration or predators. To examine how vegetation cover shapes behavioural responses of ‘trapped’ tortoises, we tested adults of six populations from habitats with contrasting vegetation cover. The tortoises were fitted with a non-stretchable rope, representing a piece of vegetation, stuck on the protruding front part of the plastron. Results suggested the existence of two distinct releasing techniques. First, and only successful in this study, is frequent changing of the movement direction, with a minimal pulling force, until the obstacle detached. The other involved the maximal pulling force aimed at ripping out the constraint. Tortoises from shrub habitats had more releasing success, used less pulling force and needed shorter time period to release, contrary to tortoises from herbaceous habitats. Although sexes showed similar releasing success, females obtained lower number of direction changes and higher yanking force compared to males, suggesting slightly different liberating strategies between the sexes. For immobilized tortoises without suitable shelter from overheating and dehydration, appropriate behavioural response could be vital, especially during drought years, due to increased physiological stresses. Variability of behavioural patterns among tortoise populations, described in this study, could have an adaptive significance.

Locomotor performances reflect habitat constraints in an armoured species

Physical limits of speed performances impose strong selective pressures on animals, influencing important functions such as predator avoidance and foraging success. Armoured herbivorous species such as tortoises represent a peculiar case: features that optimise encounter rate during reproduction, the ability to reach favourable sites for thermoregulation and resting, foraging or nesting sites may be more important than running speed. To explore this issue, we measured three characteristics related to travelling ability but that are independent of running speed: (1) muscular strength, (2) time needed for overturning and (3) jumping from a high step as proxies of the ability to overcome various obstacles. Additionally, reaction times in tortoises placed in a normal or overturned position were measured as a proxy of antipredator response. More than 400 adult Hermann’s tortoises from six populations were tested in the field during two seasons. Measures of travelling ability and antipredator response varied markedly among populations, and thus with environmental characteristics such as habitat type, terrain ruggedness and presence of predators. Tortoises from rugged and hot habitats (e.g. Mediterranean macchia) were the most successful and the fastest to accomplish the tests. Overturned tortoises were more reluctant to start moving compared to those in normal position, but this effect was absent in the two localities free from predators. Sex and season had limited effect on the measured performances. Overall, locomotor features essentially varied with environmental constraints. Future environmental studies should explore whether the observed differences among populations are linked to genetic adaptation or phenotypic plasticity.Significance statementLocomotor performances crucially influence habitat use, foraging and reproductive success, thus directly affecting individual fitness. Although running speed is often considered as a main indicator of agility, in armoured terrestrial vertebrates some other agility components might be more important. Heavy, rigid armour imposes trade-off between protection of soft body parts and locomotor performances and consequently habitat use. This study compiles three tests of locomotor performance which might be important for overcoming various obstacles present in mosaic habitats of tortoises. Testing adult Hermann’s tortoises from six populations with various habitat characteristics, we found significant inter-population differences. The ability of tortoises to complete the tests positively correlates with climatic conditions and topography of their habitats. Additionally, presence of predators in the habitats dramatically affects tortoise’s antipredator behaviour. Measured locomotor performances and antipredator behaviour showed surprisingly low level of sexual dimorphism.