On the strictly protected island of Golem Grad in North Macedonia, the tortoises are destroying their own population. During prolonged courtship, aggressive males are exhausting the females and frequently pushing them off the cliffs. Consequently, there are now one hundred males for every female capable of laying eggs. This is the only known example of demographic suicide in the wild to date.
Under favourable, stable and protected environments, large animal populations have no reason to die out. This should not happen unless a catastrophe, such as a devastating fire or the destruction of their habitat, or over-exploitation, wipes out all individuals or weakens the population, making it vulnerable to disease and other disturbances and hazards.
Well sheltered by the steep cliffs that line the island of Golem Grad on Lake Prespa in North Macedonia, Hermann’s tortoises (Testudo hermanni boettgeri) thrive on the wooded plateau.
After basking in the morning sun, they graze in the meadows, rest, and court, with the males emitting high-pitched sounds during mating. At first glance, nothing seems to threaten this population.
As is the case with other long-living species, maintaining populations requires high survival rates among adults. On Golem Grad, the adults have no predators, as wild boars, dogs, rats and humans are absent from this strictly protected island. The mild Mediterranean climate of this lake, situated at an altitude of 850 metres, is also favourable for reptiles.
All these factors explain the extraordinary population density, which stands at around 50 individuals per hectare – the highest ever recorded for tortoises. The ease with which these tortoises can be observed and studied led to the establishment of the field-monitoring programme in 2008. This was the result of a fruitful scientific collaboration between North Macedonia, Serbia and France, and this long-term monitoring programme was awarded the CNRS’s SEE-Life label in 2023.
But appearances can be deceiving: this population is in a critical state.
The extensive demographic, behavioural, physiological and experimental data collected over nearly 20 years show that, although highly active sexually and reproductively, this population is effectively committing suicide!
Demographic suicide
Demographic suicide is a strange and counter-intuitive theoretical process. The conditions under which it may arise are quite specific. For a given species, one must imagine a high-density population in which violent sexual behaviour is so prevalent that it threatens the survival of females. This would gradually lead to an imbalance in the sex ratio (the proportion of males and females in a population), in this case an excess of males. This would put increasing pressure on females, who would become fewer in number and more harassed as a result. This would eventually create a vicious circle that leads to the disappearance of females and, ultimately, the extinction of the population.
Fourni par l’auteur
Coercive and violent mating behaviours are fairly common in nature. Typically, males harass females until they mate, sometimes injuring them in the process. In some cases, such behaviour can result in the death of the female, as has been observed in elephant seals (where the males are considerably stronger than the females), as well as in wild sheep, grey squirrels, otters, deer, toads, fruit flies, humans… However, such fatal outcomes do not benefit the males, as they will have no offspring if the females die during mating. Therefore, such excessively violent behaviours are maladaptive and remain marginal.
Furthermore, in wild populations, various regulatory mechanisms prevent this type of vicious circle, or extinction vortex, from emerging. Females can employ a wide range of avoidance and defence strategies. For example, they can hide, seek the protection of a dominant male, or form alliances. Excessively violent males generally produce fewer offspring than those who spare the females, meaning their behavioural traits are less likely to persist over time. Furthermore, when males become overcrowded, they tend to emigrate in search of better mating opportunities, thereby reducing the pressure on females. Thus, conflicts between the sexes in coercive mating systems are resolved through effective equilibria, without a harmful escalation for either sex.
However, rare experiments conducted on animals studied in captivity have shown that males can have a strong negative impact on populations when the sex ratio and population density are artificially skewed in favour of males. For example, in a species of Japanese shrimp, an excess of males reduces female fertility and mating opportunities. In the common lizard, an excess of males leads to increased aggression, reducing both the fertility and survival of females. This theory has thus received partial confirmation through experimentation.
What is causing disruption to the population in Golem Grad?
Information on the sexual behaviour of terrestrial tortoises, alongside a comparison with a control population, would be useful for understanding the situation in Golem Grad. The mating system of tortoises is coercive: males chase females, bump into them (like bumper cars) and sometimes bite them until they bleed and, in the case of Eastern Hermann’s tortoises, press on the females’ cloaca with their sharp tail spurs until they yield.
Fourni par l’auteur
Hermann’s tortoises are still abundant in North Macedonia. We were therefore able to study another dense population located on the shores of the lake, just 4 kilometres from the island. Genetically very similar to the Golem Grad population, this population lives in a protected environment without cliffs. The females are large and heavy, with many weighing between 2.5 and 2.9 kg, and highly fertile, as shown by X-rays. They are slightly more numerous than the males and larger than them, and they effectively resist their intermittent sexual assaults. No demographic problems have been detected; population forecasts suggest an increase in numbers.
Fourni par l’auteur
However, the situation at Golem Grad is quite different. On the plateau, over 700 adult males roam around looking for the forty or so adult females.
Furthermore, if physiological and environmental conditions are unfavourable, a Hermann’s tortoise may fail to lay eggs after mating. For example, if they are too thin or stressed, they are unable to build up reserves in the ovarian follicles and the eggs do not develop. In reality, therefore, there are more than 100 males for every female capable of laying eggs. However, our analysis of neonate and juvenile cohorts shows that the sex ratio is balanced at birth and during the first years of life, becoming imbalanced later on.
The surplus males often act in groups of three to eight. They harass the females all day long and injure them. They then lay down beside the females in the evening, ready to start again the next day. The females have little respite and do not have enough time to feed. They are thin, very few exceed 1.6 kg, with a maximum of 1.75 kg and when they lay eggs, they produce half as many as those in the control population.
Unable to escape, the females are regularly driven to the cliff edges, where the obstinate and clumsy males sometimes push them over. On July 18 2023, a GPS device fitted with an accelerometer, attached to a female, recorded her fall of over 20 metres; she died, broken in two, along with her three eggs.
Fourni par l’auteur
Since the start of the study, we have identified almost all the turtles that have been found dead in the field, where their shells remain intact for a long time. Of the females that died, 22% suffered a fatal fall, compared to 7% of the males.
In collaboration with British colleagues, we have also developed an epigenetic clock to estimate the age of individuals from a blood sample.
The oldest males are over 60 years old and the oldest female is 35. These results are consistent with morphological, growth, and demographic analyses. The survival rate is abnormally low among females, due to male aggression.
The vicious circle of extinction
Over time, the decline in the number of adult females, coupled with a drop in their fertility, slows down the renewal of the population, both relatively (the proportion of females) and absolutely (the total number of females). In 2009, we captured 45 adult females in the field, compared to 37 in 2010, 20 in 2024, and just 15 in 2025.
However, it takes a female around fifteen years to reach adulthood. Frustrated by the lack of sexual partners, males mate with other males, carcasses, stones, and immature females. Through this latter behaviour, they prematurely compromise the survival of females and exacerbate their demographic problem.
Population dynamics can be modelled by incorporating the above parameters and others. It is also possible to make predictions. The last female could die in 2083. The males, now deprived of females, will survive for decades, as these tortoises can live for over eighty years and will eventually die out. This is a prediction; perhaps the population, which is currently on the brink of extinction, will recover, even if we cannot see how. While the tortoises’ very slow pace of life has given us the opportunity to observe an extinction vortex in the wild and test a strange theory, intensive field monitoring has provided us with the data and inspiration above all else.
A weekly e-mail in English featuring expertise from scholars and researchers. It provides an introduction to the diversity of research coming out of the continent and considers some of the key issues facing European countries. Get the newsletter!
