Eurasian lynx (Lynx lynx) have recently naturally recolonized southern Sweden. The first documented reproduction of lynx in recent times occurred in 2003, and the population increased from two to 48 family groups (the unit of measurement in Swedish monitoring) during its first 18 years (2003/04 – 2020/21). We did not detect any Allee effect, i.e., lower growth rate at low population density, during the recolonization of southern Sweden, although our population simulations revealed a non-negligible (30 %) chance that population observed development could include an Allee effect. The probable absence of an Allee effect was likely because colonizing females did not lack mating partners, as a larger number of wide-ranging males were established in the area before documented reproduction took place. Despite the absence of an Allee effect, the growth rate during recolonization was lower in southern Sweden (lambda = 1.20) than in central Sweden (lambda = 1.29). We have no evidence of higher mortality, including that from poaching, or lower reproduction in southern Sweden could explain the lower growth rate. Instead, we suggest that the lower growth rate during the recolonization of southern Sweden was explained by fewer immigrants arriving from central Sweden due to areas of less suitable habitat between central and southern Sweden, partially preventing immigration southward. From a conservation point of view, it is positive that this small population could recover without being negatively influenced by an Allee effect, as small populations with an Allee effect experience lower viability than those without.

We used data from the Swedish lynx monitoring system, available in the in the official carnivore database (Rovbase; rovbase30.miljodirektoratet.no). Lynx monitoring in Sweden is based on non-replicated counts of family groups. The monitoring is primarily conducted from December to the end of February and largely based on snow-tracking and identifying lynx tracks from two or more individuals, which are then assessed as a family group consisting of an adult female and young of the year. Simultaneous snow tracking or a distance criterion based on home-range sizes and movement patterns from radio-marked female lynx with kittens are used to separate observations of different family groups, to assure that counts of family groups are distinct. Additional observations that are used to confirm reproduction include camera-trap images of kittens, and any kittens shot in the early part of the hunting season (February) or killed in traffic accidents. Trained and authorized personnel from the Swedish County Administration Boards perform the lynx monitoring. Game wardens, hunters, and the public can report records of lynx tracks, but all observations need to be verified by the authorized personnel before being confirmed and entered into the carnivore database and thus used in the national count of family groups. The family group counts are multiplied by a conversion factor to encompass the entire lynx population, including males and non-reproducing females. The conversion factor is on average 5.48 (± 0.40 SD) in central and southern Sweden. There was a good fit between monitored number of lynx family groups and reconstructed population size. Thus, the lynx monitoring provides a proxy of all lynx in an area. In the population models we included dead lynx. Data on dead lynx was downloaded from the carnivore database (Rovbase; rovbase30.miljodirektoratet.no).