**2. Methods**

274 Epidemiology Insights

2008), but this is often hampered by the lack of epidemiological data (Artois *et al*., 2001;

Bovine tuberculosis control programs in cattle are in place for several decades in Iberian Peninsula and consequently incidence has been decreasing (Allepuz *et al*., 2011; Cunha *et al*., 2011). However in the last few years incidence has stabilized, or even slightly increased in both Portugal and Spain (Allepuz *et al*., 2011; Cunha *et al*., 2011). The role of wildlife hosts in this scenario remains speculative; nevertheless the existence of wildlife reservoirs may compromise the goal of eradication in cattle. Besides livestock, attention should be given to spill-over from wildlife to other domestic animals (e.g. goats and free-ranging pigs) and even to humans, namely hunters and others that handle wild ungulate carcasses (Gortazar *et al*., 2011b, in press). Wildlife-to-human transmission of *M. bovis* is hard to prove and no single case has been documented in Iberian Peninsula, but it is known to occur elsewhere (e.g. USA – Wilkins *et al*., 2008). Bovine tuberculosis is also one of the main infectious diseases affecting the critically endangered Iberian lynx *Lynx pardinus*, with several freeranging and captive lynx killed by this infection (Millán *et al*., 2009). Iberian lynx is subject to an intensive multinational conservation program in Iberian Peninsula, which includes releasing captive-bred animals to former range. The persistence of *M. bovis* on the environment

and in prey species poses a threat to this conservation action (Millán *et al*., 2009).

changes have potential implications on bTB epidemiology (Gortázar *et al*., 2006).

Iberian Peninsula ecosystems display a high degree of human intervention and have experienced some profound changes in the last decades. The most important alterations were a shift from domestic ungulate to wild ungulate production for hunting purposes (Miguel *et al*. 1999) and an increasing intensification of the later (Vargas *et al*. 1995). This management of wild ungulate populations aims to increase profits by increasing harvest, translating into increased densities of hunted species. This has been accomplished through introduction/restocking, provision of food and water (mostly during the summer shortage), fencing and sometimes even medication (Miguel *et al*. 1999, Gortázar *et al*., 2006). All these

In the Iberian Peninsula, ungulates such as the wild boar *Sus scrofa* and the red deer *Cervus elaphus* have been recognized as the most important maintenance hosts for wildlife tuberculosis (Gortázar *et al*., 2011b). Nevertheless other species have also been identified as locally non-negligible hosts, such as the fallow deer *Dama dama* and the badger *Meles meles* (Gortázar *et al*., 2011b; Balseiro *et al*., 2011). Several other species of ungulates and carnivores were also found infected (Rodriguez *et al*., 2010). This situation fits the definition of a multihost pathogen within a multi-species ecosystem (Renwick *et al*., 2007; Gortázar *et al*., in press), in which pathogen persistence and spread is dependent on the density of each maintenance host species and also on the effective interspecies contact rate (dependent on

Research on host-pathogen interaction usually deals with single-host single-pathogen systems, where disease persistence depends solely on the intra-species transmission rate (Tompkins *et al*., 2001). If transmission is density-dependent, then population thresholds for disease invasion and persistence are expected and have been described (Swinton *et al*., 2001). By contrast, in multi-host pathogens systems, disease persistence is dependent on both intra and inter-species transmission rates and densities of several host species (Renwick *et al*., 2007). Moreover, these rates depend on pathological, epidemiological, ecological and

Corner, 2006).

the ecology of each species).

behavioural factors (Corner, 2006).

We conducted a systematic bibliographic review for epidemiological studies on tuberculosis in wildlife in Iberian Peninsula by searching MEDLINE/PubMed, up to the 31st of August 2011, using MeSH and keywords: "*Mycobacterium bovis*", "*Mycobacterium caprae*", "wild boar", "deer", "epidemiology", "Iberian Peninsula", "Portugal" and "Spain". Combinations used were: ("Portugal" OR "Spain") AND ("*Mycobacterium bovis*" OR "*Mycobacterium caprae*"), ("*Mycobacterium bovis*" OR "*Mycobacterium caprae*") AND "wild boar" AND "epidemiology" and ("*Mycobacterium bovis*" OR "*Mycobacterium caprae*") AND "deer" AND "epidemiology". Abstracts were selected according to their relevancy and excluded if dealing exclusively with laboratory or pathology investigations, domestic species or humans or other geographical regions. Articles were reviewed in full text.

For each article, information about the type of epidemiological study and study design, sample size and sampling methodology, screening and diagnostic tests used, prevalence rate, time frame of the study, study areas, characteristics of the populations studied, risk factors identified and host epidemiological status was summarized and presented in table format for easy comparison. Due to their idiosyncrasies, molecular epidemiology articles were characterized differently according to the number of isolates studied, genotyping technique, mycobacterial species reported, number of genotypes found, host and geographical clustering of genotypes and study areas. Due to differing methodologies and sometimes incomplete reporting of results, meta-analysis was not applicable except for a small number of studies.

For the purpose of this review, wildlife tuberculosis was defined according to the OIE definition of bovine tuberculosis, but *Mycobacterium caprae* was also considered etiological agent, besides *M. bovis*.
