With infectious diseases causing calamitous declines in a wide range of endangered populations, there can no longer be any doubt that diseases are a matter of serious concern for conservation biologists. The outbreak of feline leukemia virus (FeLV) in Iberian lynx Lynx pardinus in Spain, described by López et al. (2009), involves a depressingly familiar suite of culprits, but also raises some troubling new concerns about disease impacts in endangered species. The outbreak involved a recurring triad of factors: a small, endangered wild carnivore population (the Iberian lynx), a larger population of related domestic animals (domestic cats) and a virus (FeLV).
Small (endangered) populations are unable to maintain virulent viral pathogens independently, as the pathogen rapidly depletes susceptible hosts and burns itself out. Infectious disease threats for endangered wildlife therefore commonly originate from other species. Domestic animals, which have the potential to maintain a wide range of pathogens in relatively abundant populations, are among the most dangerous of these sources and feature prominently in a recent litany that includes canine distemper virus (CDV) in Serengeti lions (Roelke-Parker et al., 1996); rabies in Ethiopian wolves (Sillero-Zubiri, Gotelli & King, 1996; Randall et al., 2004); rabies and CDV in African wild dogs (Gascoyne et al., 1993; Kat et al., 1995; van de Bildt et al., 2002); CDV and heartworm in Channel Island foxes (Timm et al., 2000; Crooks, Scott & Van Vuren, 2001); infectious keratoconjunctivitis in Ibex and Chamois (Giacometti et al., 2002); CDV in Lake Baikal seals (Mamaev et al., 1996).
The close phylogenetic relationship with domestic carnivores and a high susceptibility to domestic carnivore pathogens has been suggested as one reason why, within mammals, carnivores feature most prominently among species threatened by infectious diseases (Pedersen et al., 2007). But, within the carnivores, felids appear to have been less troubled by disease than canids. There may be several reasons for this: domestic cats are probably less numerous in and around wildlife-protected areas than dogs (although there is surprisingly little data available on domestic cat populations in most of these areas); direct and indirect contact rates between domestic cats and wild felids may be lower than between domestic dogs and wild canids; and because felids are generally less social than canids, the spread of diseases within wild felid populations may be less rapid and population-level impacts less severe than in wild canids.
The FeLV outbreak in Iberian lynx in Doñana in 2007, which was the result of disease transmission from domestic cats, was therefore particularly troubling. Clearly, contact between domestic cats and Iberian lynx (presumed to have occurred through fighting) was sufficient to introduce the virus into the population. This, in itself, was not of undue concern to wildlife managers, as FeLV infection, although widespread among domestic cats, had only rarely been reported as a cause of disease in wild felids. In contrast, the Doñana outbreak was characterized by high levels of mortality associated with FeLV infection (often due to secondary infections) and rapid spread of the virus within the lynx population once introduced.
Although the reasons for increased pathogenicity of FeLV infection in this outbreak are not known, the appearance of diseases in a new guise should not be unexpected. Many pathogens, and viruses in particular, have a propensity for host-switching and also for infection in the same host population to result in widely differing outcomes. Co-infections, environmental factors and a myriad of host-related factors can all interact to profoundly affect susceptibility to disease. What is clear is that wildlife managers must be prepared for the appearance of new diseases. As contact between domestic animals and wildlife species increases and environmental conditions change, new patterns of wildlife disease and mortality will inevitably emerge. How, whether and when to intervene to tackle these disease problems are major issues in conservation management and the work by López and colleagues makes a major contribution to these discussions.
For over 200 years, vaccination has been one of the most effective and cost-effective preventive health measures for improving the health of both human and domestic animal populations. Vaccination also offers a potential solution for reducing disease threats for endangered species. This is particularly true for viral pathogens transmitted from domestic dogs and cats, for which safe and effective vaccines have already been developed. But, since the vaccination of chimpanzees against polio in Gombe National Park in 1966 (Van-Lawick-Goodall, 1971), vaccinating endangered wildlife has often led to controversy, and decisions to intervene influenced by a complex suite of scientific, practical, political and ethical issues.
In the wake of the deaths of four FeLV-positive lynxes in Doñana, the decision was made to intervene in order to prevent further spread of infection, remove the virus from the population and minimize the risk of future outbreaks. The interventions focused on removal of FeLV-viremic lynxes, vaccination of non-infected lynxes, changing management practices to reduce intra-specific contact (e.g. modification of feeding stations) and reducing contact between lynx and domestic cats (e.g. cat removal). These interventions appeared to have successfully contained the outbreak to within one sub-population.
Several widely debated issues are raised by these interventions. First, the decision to intervene at all is often criticized, with the argument that diseases are natural phenomena and that wildlife managers and veterinarians should not meddle in natural processes. Indeed, parasites and pathogens are important components of natural ecosystems and can play an important role in population regulation. But, the rapid expansion of domestic animal populations is entirely anthropogenic and most wildlife emerging disease threats are associated with human activity (Daszak, Cunningham & Hyatt, 2000).
Second, the question of how to design vaccination intervention strategies has been the subject of much discussion. The failure to evaluate sufficiently the impact of vaccination interventions has previously led to highly damaging controversies (e.g. African wild dog vaccination in East Africa, reviewed by Woodroffe, 2001) and highlighted the critical importance of effective population monitoring and evaluation following any intervention (Knobel et al., 2008). However, evaluating the impact of an intervention in endangered wildlife can be difficult. In terms of protecting individuals (and assessing the safety of the intervention), comparison of morbidity and mortality of vaccinated and non-vaccinated individuals can be carried out if some animals remain unvaccinated, and the intensity of population monitoring in Doñana should allow this evaluation to be carried out.
However, the availability of a non-intervention control population for evaluating the effectiveness of interventions and population-level outcomes (such as population persistence) is rarely an option. In the study by López et al. (2009), disease management measures were evaluated through monitoring of FeLV infection status in both the Doñana population and Sierra Morena population, and the interventions put in place were considered to have contained the outbreak when no positive lynxes had been detected for a period of four months. However, it can be difficult to determine precisely how an intervention has affected the outcome of an outbreak, particularly when several management strategies are implemented concurrently. In these situations, mathematical models have proved enormously valuable. For example, a spatially explicit individual-based model used to assess the impact of a rabies vaccination campaign on an endangered population of Ethiopian wolves in the Bale Mountains National Park indicated that reactive vaccination had limited the scale of the outbreak and should significantly enhance the persistence of the population (Haydon et al., 2006). Models can also guide managers as to whether, and when, in the trajectory of an outbreak implementation of reactive disease control strategies are likely to be effective.
The intervention in Doñana involved the capture of an impressively high proportion of the population, with more than 80% (n=34) of the individuals being trapped and their infection status determined before removal (of viremic individuals) or vaccination (of naïve individuals). An important consideration for undertaking an intervention of this kind must clearly be access to sufficient resources and field capacity, as well as the availability of appropriate and rapid diagnostic tests and technical expertise. Although the intervention resulted in a relatively high level of coverage, with 22 lynxes in Doñana vaccinated out of a total population of 33 (after removal of viremic animals), mathematical models have shown that, by protecting a demographically viable ‘core’ of individuals, even low-vaccination coverage can be effective in reducing the threat of extinction (Haydon et al., 2006; Vial et al., 2006), and can be considered where resources or logistic constraints limit access to a larger proportion of the population.
In summary, the case study presented by López et al. (2009) provides valuable information for conservation managers about the dangers of infectious diseases and also the potential strategies for managing these threats. There is no doubt that wildlife diseases will continue to appear in endangered populations and we need to be well prepared to respond to these threats. Well-designed interventions that can be monitored and evaluated will do much to help build confidence that interventions to control wildlife disease threats are feasible and can be effective.