Introduction
Terrestrial carnivores are some of the most imperiled species today due to their large home range requirements, high metabolic demands, sensitivity to habitat fragmentation, and persecution by humans (Crooks, 2002; Palomares & Caro, 1999; W. J. Ripple et al., 2014; Woodroffe & Ginsberg, 1998). Carnivores are also important top-down regulators in ecological communities (Beschta & Ripple, 2009; W. J. Ripple & Beschta, 2006; W. Ripple J. & Beschta, 2012). The loss of key carnivore species can have devastating ecosystem effects (Effiom et al., 2013; W. J. Ripple et al., 2014) and changes in abundance or occurrence of carnivores can trigger trophic cascades (W. Ripple J. & Beschta, 2012). As such, the recovery of apex predators as a conservation tool to restore ecosystem functions (termed trophic rewilding) has become increasingly popular (Jørgensen, 2015; Seddon et al., 2014). Trophic rewilding is an ecological restoration strategy used to promote self-regulating ecosystems (Svenning et al., 2016).
Rewilding efforts in the context of apex predators requires not only an understanding of their ecological interactions within the carnivore guild, but also the broader context of these interactions including sources of anthropogenic impacts. Many apex predators readily reestablish in human-dominated landscapes and exhibit potential coexistence with humans (Chapron et al., 2014; Lamb et al., 2020). Although the effects of apex predator recovery in natural landscapes are relatively well understood, there are significant knowledge gaps regarding the effects of their recovery in shaping species interactions (both intraguild and across trophic levels) in human-dominated landscapes (Dorresteijn et al., 2015). Interactions between carnivores are complex in nature, but are integral to shaping the ecology and structure of wildlife communities. Therefore, examining such interactions in landscapes that harbor viable carnivore populations may provide important insights into the effects of carnivore recovery on the mesocarnivore communities that often dominate landscapes where apex predators have been eliminated.
Grey wolf (Canis lupus ) and Eurasian lynx (Lynx lynx ) are top predators in many temperate ecosystems in Europe and Asia, but their co-occurrence has been severely limited by extirpation of one species (most often wolf). This is particularly the case for most of Western and Central Europe due to a long history of human habitation and persecution of carnivore species. Both wolf and Eurasian lynx are recovering in Europe’s landscapes (Chapron et al., 2014; Kaczensky et al., 2013) either through natural range expansion (wolf) or reintroductions and population augmentation (lynx). The wildcat (Felis silvestris ) is a mesocarnivore that was once common in Europe, has also been extirpated, and currently at the core of reintroduction programs in some European Union states. In this context, the Romanian Carpathians represent one of the few natural areas in Europe that still harbor intact viable populations of all three species and serve as a stronghold for carnivore populations in Europe, despite anthropogenic influences common (hunting, forestry, farming, and livestock production) (Popescu et al., 2016; Salvatori et al., 2002).
While no work has been conducted on understanding the spatial relations and interactions between these three species simultaneously, research exists on pairwise interactions between species, particularly for lynx and wolf. Lynx and wolf are sympatric across most of their range and there is some diet overlap between them. Research addressing coexistence between these species differ in their findings, but recent studies looking at spatial interactions between these species in Europe found that these two apex predators coexist and competition between them is low (Schmidt et al., 2009; Wikenros et al., 2010). In Poland, lynx and wolf territories overlap and researchers concluded that the co-occurrence of these two species was facilitated by heterogeneous habitat and specialization on different prey (Schmidt et al., 2009). These predictors, habitat heterogeneity and diet, are also explaining competitive interactions between canids and felids in North America, with a lack of interference competition in heterogeneous habitat (Dyck et al., 2022). Therefore, we expect to observe similar co-existence (high co-occurrence) and little evidence of interference competition (neutral or positive conditional occupancy values) between lynx and wolf in our study area. Additionally, we expect to observe differences in co-occurrence based on seasonal changes in these species’ behaviors. For example, male lynx move further during the mating season (January-March) and female lynx move further during periods of extensive kitten care (May-August) (Jedrzejewski et al., 2002), which could cause increased interactions with wolves. Research on wildcats is scarce, but a study conducted in the Jura Mountains of central Europe found no evidence of avoidance between lynx and wildcat (Zimmermann & Raoul, 2011). No published research examines interactions between wildcats and wolf, however interspecific interactions have been assessed between wildcats and red foxes. A study in Spain found that the two carnivores exhibited active avoidance towards one another and exhibited aggressive behaviors during encounters suggesting interference competition (Ruiz-Villar et al., 2021). Given the size difference between wolf and wildcats and their different diets, it is likely that the relationship between wildcats and wolf will be similar to that of wildcats and lynx.
In this study, we aimed to address these knowledge gaps by studying the intraguild interactions of two apex carnivores, the Eurasian lynx and the grey wolf, and a mesocarnivore, the wildcat in the Romanian Carpathians using multispecies occupancy models (Rota et al., 2016). Unlike traditional occupancy modeling, multispecies occupancy models allow for the estimation of co-occupancy probabilities for more than two species and do not assume asymmetric interactions (i. e., dominant and subordinate species). This is useful for estimating co-occupancy probabilities between species for which there is not a priori knowledge about interspecific relationships or for which there is not an obvious dominant or subordinate species. Multispecies occupancy models also allow for the estimation of marginal occupancy (occupancy of a single species irrelative of other species) and conditional occupancy (occupancy of a single species based on the presence or absence of another species) probabilities in relation to variables of interest (e.g., altitude). This approach has been used effectively to assess habitat use, interspecific interactions of carnivores in a variety of landscapes (Dechner et al., 2018; Lombardi et al., 2020; Van der Weyde et al., 2018). Previous research on lynx-wolf and lynx-wildcat interactions suggests a high capacity for coexistence, low interspecific competition, and little to no intraguild killing. However, this research is limited and there has been no work on lynx-wolf dynamics or interactions of lynx, wildcat, and wolf in the same region. Additionally, none of the published literature has been conducted in an area with a fully intact carnivore guild. This information is crucial to understanding the effects of apex predators on mesocarnivores and the carnivore guild. By using a multi-species occupancy approach, we can analyze complex intraguild interactions and better understand competition and coexistence patterns. Results can elucidate variables and thresholds important for occurrence and coexistence of elusive species and help inform management or reintroduction efforts. Our specific objectives were: 1) evaluate seasonal predictors for occupancy of each species, 2) characterize the spatial relationships (co-occurrence) of each species in winter and autumn, and 3) identify predictors that facilitate co-occurrence. Specifically, we analyzed the effects of potentially dominant apex carnivores on the occupancy and detection of a mesocarnivore to understand potential impacts reintroductions of apex predators may have on smaller carnivores. We also evaluated seasonal changes is marginal and co-occupancy probabilities to better understand how species persist and interact under different environmental conditions.