Lynx d'Espagne

The major threat is the risk of collapse of European Rabbit populations across large areas within the current Iberian Lynx distribution. During the last six decades, several outbreaks of different viral diseases became enzootic and provoked strong rabbit declines either in the short-term only (Delibes-Mateos et al. 2014) or lasting for extended periods (Moreno et al. 2007, Delibes-Mateos et al. 2009). Lynx response to rabbit scarcity was also strong and persisting (López-Bao et al. 2010, Ferreras et al. 2011). Indeed, during five out of the last six decades, rabbit scarcity was the single most important driver limiting the distribution and abundance of the Iberian Lynx (Delibes et al. 2000, Ferreras et al. 2010). Now rabbit populations thrive across broad areas, even becoming a pest in some agroecosystems (Delibes-Mateos et al. 2018), whereas in montane areas rabbit recovery was lower, or some populations may even decline, perhaps due to the interaction between disease, predation and other factors (Calvete 2006). Population trends during the last decades illustrate how the European Rabbit may undergo large and rapid fluctuations in numbers. Therefore, a sudden decrease in rabbit abundance, associated with new disease outbreaks or other causes hard to predict, could also occur in the future entailing a serious threat for the Iberian Lynx.

The effects of other important threats are also indirectly mediated by marked shifts in rabbit abundance. Changes in land use constitute a good example. Rabbits are very scarce in pulpwood and timber plantations making intensive forestry an unsuitable land use for breeding lynx (Palomares et al. 1991, 2000). Conversely, open scrublands in fine-scale mosaics are quite favourable for rabbits and lynx (Palomares 2001, Fernández et al. 2003), and this relationship also operates at the landscape scale. Land use shifts from small to big game tend to induce the encroachment of small clearings by shrubs thus reducing the quality of mosaics for rabbits (Rodríguez and Delibes 2002). Reduced habitat suitability may also arise from the abandonment of pastures and cultivated fields within forest landscapes, as well as the structural simplification of mosaic landscapes following agricultural intensification.

A number of secondary threats can also be listed. The genetic diversity of the Iberian Lynx is remarkably low, being genetic erosion a consequence of drift after severe population bottlenecks during its evolutionary history (Casas-Marcé et al. 2013, 2017). Indication of inbreeding depression has been reported in the isolated Doñana subpopulation, where low genetic diversity, poor reproductive performance, signs of immunosuppression, and high susceptibility to disease were correlated (Palomares et al. 2012).

Although habitat suitable for the Iberian Lynx does not seem to be in short supply, substantial habitat loss may quickly result from large wildfires. The risk of large wildfires has increased due to a combination of human depopulation in rural areas, the parallel growth of secondary forest in abandoned agricultural lands, and the increasing forest dryness produced by anthropogenic climate change (Turco et al. 2018, Calheiros et al. 2021). The mean size of large wildfires recorded in Mediterranean ecosystems of the Iberian Peninsula until 2010 was 15 km² (Cardil and Molina 2013). Fires of this size could well wipe out the habitat of small lynx population nuclei.

Lynx mortality of anthropogenic origin was an important threat in the past (Ferreras et al. 1992, Rodríguez and Delibes 2004) and should still be taken into consideration (López et al. 2014). Between 2006 and 2011 the annual mortality rate attributed to direct human threats, mostly poaching and road kills, was 0.05 based on a sample of 78 radio-tagged individuals. Published updates are not available but preliminary data suggests that annual mortality rates might have increased following lynx occupancy of highly anthropized areas. Poaching is always present but hard to detect, which makes it difficult to quantify its actual contribution to lynx non-natural mortality. As a result, the impact of poaching on lynx demography may be systematically underestimated. There is no indication that the Iberian Lynx is generally and actively persecuted, as happens with other carnivore predators. This is inferred from high adult survival rates in most subpopulations, once the effect of known causes of mortality are accounted for. Poaching associated with conflict with lynx predation on livestock or poultry might be locally important when losses are not properly compensated (Garrote et al. 2013). Road kills are a threat with strong population effects on specific localities where middle-size roads accommodate high traffic volumes across dense lynx nuclei.

The importance of disease as a cause of lynx mortality is also underestimated because fresh carcasses are rarely found in the field. Between 2006 and 2011 the annual mortality rate attributed to disease was 0.06. Twenty-eight pathogens have been described in the Iberian Lynx (Ferreras et al. 2024). Most of them circulate in the lynx population and may cause low to moderate mortality except in lynx nuclei prone to immunodepression (Palomares et al. 2012, López et al. 2014). Frequent contact with domestic carnivores may increase pathogen transmission (Millán et al. 2009, Meli et al. 2009, López et al. 2014), and increased lynx exposure to disease agents in agricultural areas is expected to be larger than in the scarcely populated montane areas where the species was confined in the recent past. In the mountains, however, big game may enhance pathogen spillover from wild ungulates (e.g. tuberculosis, Gortázar et al. 2012). Another source of concern comes from density-dependent effects. Very high rabbit densities may increase not only local lynx carrying capacity but, sometimes, may also elicit some relaxation of territorial behaviour. High lynx density may favour the transmission of new pathogen variants, and may increase mortality rates if outbreaks emerge (Bradshaw and Brook 2005, Meli et al. 2009).

Altogether, mortality rates from poaching, road casualties and disease seem to be outweighed by current recruitment rates in most lynx subpopulations. Exceptions exist where high anthropogenic mortality may limit population growth, as in the Gualdalmellato cluster within the Sierra Morena subpopulation.

The colonization of agroecosystems, that are increasingly managed adopting intensive agricultural practices, may increase the risk of lynx drowning in irrigation ponds as well as lynx exposure to toxic chemical compounds either directly acquired (e.g. while drinking) or potentially bioaccumulated in prey tissues.

Actions currently in place

The following conservation actions are undertaken to preserve consolidated subpopulations, and to stabilise reintroduced or naturally colonised lynx nuclei.
 
Monitoring plays a key role. Every year, considerable effort is made to recapture in camera traps as many individuals as possible. Spot patterns in the lynx coat allow individual identification. Breeding territories are geolocated and an important goal is determining whether reproduction takes place in them. Immigrants are tried to be detected and identified. Some individuals are tagged with radio-collars to obtain valuable data on mortality rates and movements. The relative abundance of rabbits is also regularly assessed. All known mortality events are recorded and investigated. Intensive field work allows the detection of local threats associated with habitat alteration and hazards for specific individuals. The effect of implemented conservation actions is evaluated. A network of collaborators built over the years has proven useful to get information relevant for conservation. Intensive monitoring is a trace of the procedures implemented in the early 2000s (Simón 2012), when the fate of each individual was highly relevant for species survival. With a rapidly expanding lynx population, maintaining such an effort is becoming unsustainable, and a design adapted to the new circumstances is needed.

Habitat management protocols, including guidelines for the recreation of a patchy structure of vegetation and rabbit restocking, are still be implemented in some localities, but the mid-term sustainability of habitat quality relies on the commitment of landowners and gamekeepers to keep on applying lynx-friendly management procedures. This important action has been inherited from conservation agreements in private lands signed during the last 20 years. Such agreements often involved some sort of technical support or economic compensation, but favourable habitat management has often continued after the reduction or even the absence of incentives. Informative campaigns have also increased the awareness of local communities about the benefits of lynx occurrence for ecosystems and their own livelihoods. Lynx conservation has considerable social support which is maintained by regular communication of relevant news to the general public.

After reintroduction is completed, even in scenarios where newly created nuclei show signs of self-sustainable growth, additional lynxes are sometimes restocked to fill gaps in the local distribution, to help balance adult sex-ratios, or to replace losses in sites of interest.    

Where needed, habitat management is also applied to stepping stones, that is, relatively small areas selected in the intervening space between existing subpopulations, meant to facilitate transitory occupancy of dispersing lynx and eventually demographic exchange. Stepping stones are initially selected on the basis of habitat models. In some subpopulations, lynx productivity is high enough to produce many emigrants, some of which discover and colonise suitable areas. After detection of lynx presence, such areas may be defined as stepping stones. Management in these areas includes deeper monitoring, a survey of mortality risk, an assessment of the area containing enough food, contacts with landowners to create a receptive environment, and prospective lynx releases.

In stable lynx nuclei, recent reintroduction sites, and stepping stones, non-natural mortality is monitored and measures are actively taken to reduce mortality rates. The mere presence of conservation technicians and their informative talks may suffice to reduce the risk of lynx losses, especially from hunting and predator control. In most locations, any conflict from predation on domestic animals is promptly addressed and compensated. Economic support is supplied to farmers to adopt measures preventing lynx to access poultry pens. Individual lynx recurrently involved in livestock predation may be captured and translocated. Irrigation ponds and wells are inventoried in lynx areas, drowning hazards are evaluated while fencing may be reinforced and escape devices can be installed in the most dangerous spots. Hunting fences are also checked to avoid lynx entanglement. Road kill prevention includes the detection of risky situations (e.g. black spots) and the implementation of measures to increase the permeability of transport infrastructures (underpasses, overpasses, culverts, and devices channelling lynx movement), to increase mutual detection between lynxes and drivers (right-of-way clearings, visual and acoustic deterrents), to calm traffic (speed limitation, bumps), and to increase safety in high-speed roads or railways (efficient fencing combined with the construction or rehabilitation of safe passages). Mortality rates due to natural causes are also minimised especially by sanitary checks to determine the circulation of pathogens in lynx subpopulations and the fraction of individuals that actually develop a disease. Annual health checks help to prevent disease outbreaks that have proven highly detrimental for lynx conservation in the past.

The genetic composition of each subpopulation (and clusters within subpopulations) is also monitored through the accumulation of biological samples collected during sanitary checks and from lynx found dead. The genetic composition of the ex situ population is also known (Kleinman-Ruiz et al. 2019). Both sources of information are combined to maximise genetic diversity within lynx nuclei and clusters. Genetic criteria are established to (i) prioritise access to breeding in the captive population, (ii) to design mating schemes, (iii) to select founders for reintroduction, (iv) to select lynx for restocking, (v) to extract individuals from the wild that could improve genetic diversity of the ex-situ population, and (vi) to inform translocations between populations.   

To create new subpopulations through reintroduction, conservation actions include a thorough evaluation of the suitability of the reintroduction site using criteria as size, quality and landscape connectivity of habitat patches, mortality risks, legal protection, rabbit abundance, land use, health parameters of co-occurring wildlife, road density, or social attitudes, among others. The proper assessment of an area for reintroduction may take several years of data collection and analysis. Releases of 4–10 individuals per year are usually sustained for four to six years to cope with the risk of high emigration rates during the first attempts. Occasional restocking may be applied later on, if needed. After that, the regular management described above is applied.

The Iberian Lynx is listed in Appendix II of the Bern Convention, and Annexes II* and IV of the European Union Habitats and Species Directive. It is also listed in Appendix I of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES).

Actions needed to mitigate major threats

The most reasonable strategy to buffer the effects of a potentially sudden rabbit population collapse on the lynx population is spreading the risk by creating as many reintroduced populations as possible. Ideally, this should be done during the current phase of generalised abundance of wild rabbits. This strategy relies on the assumption that a new outbreak would take some time to spread at a large scale, but this depends on the speed of pathogen transmission and how rabbit abundance is geographically distributed when the outbreak takes place. Early detection is essential to try to isolate the outbreak. Current plans consider monitoring rabbit abundance as an indicator, using specific abundance thresholds as a warning. Monitoring rabbit pathogens could also be useful and it is being planned. The reaction to a sudden decrease in rabbit density mainly relies on mitigation measures such as rabbit restocking and supplementary feeding. This conservation response might work during an intense but transitory reduction of food availability. However, in most cases, predicting how long it will take for rabbit populations to recover from an emerging disease will be difficult.

More generally, increasing artificially the abundance of rabbit populations at a large scale is costly and has proved largely inefficient. It may pay moving lynx to more rabbit-rich areas if they exist. This action would mimic the ability of lynx to track prey across the landscape. Indeed, the selection of stepping stones aimed to facilitate connection between existing subpopulations has been, in most cases, actually made by lynxes themselves, reaching and pointing out sites meeting the ecological conditions suitable for establishment. Essentially, moving lynx where rabbits thrive instead of increasing rabbit abundance within occupied areas with low lynx productivity has been the successful approach followed during the selection of areas suitable for reintroduction and the subsequent releases.

Regarding the question of where to undertake new lynx introductions, currently suitable habitat seems to abound. A trade-off exists between the advantages and disadvantages of spacing lynx subpopulations as much as possible. Broad spacing protects subpopulations from threats originated in distant subpopulations to spread and affect the others, and also reduces the exposure of the population to wildfires and other large-scale disturbance. On the other hand, spacing hampers demographic and genetic exchange. Increasing separation between subpopulations also reduces the probability of subpopulations to be rescued (Rodríguez and Delibes 2003), but that could be done artificially with new reintroductions or restocking. Two new reintroduction attempts in southeastern Spain have started in 2023, one of them in the Murcia region, 500 km apart from the western-most Portuguese subpopulation of Vale do Guadiana. A number of additional reintroductions in distant localities are also planned.