LATENT INFECTIONS (APPARENTLY HEALTHY CARRIERS)

Latent infections (apparently healthy carriers) may be the most common type of reservoir among pathogenic bacteria, and examples are found among a wide variety of bacterial groups.

SALMONELLA SPP.

causative agent Salmonella spp. are gram-negative bacteria (Family Enterobacteria- ceae; App. 1: Table 7) shaped as short rods; they are well-adapted intestinal bacteria of many birds, mammals, reptiles, and amphibians. At least 2,435 serotypes (serovars) of Salmonella spp. have been identified, based on their sur­face and somatic antigens (Bopp et al. 1999, Morner 2001b); most of these serotypes show little specificity for their host species (Daoust and Prescott 2007). Until the 1970s, species within the genus Salmonella were defined by their epidemiology, host range, biochemical reactions, and antigenic structure. These for­mer species were reduced to serotypes of two species: S. enterica and S. bongori; S. enterica contains five subspecies and the vast majority of serovars (Murray 1991, Clark and Gyles 1993, Bopp et al. 1999, Farmer 1999, Brenner et al. 2000). Most of the older literature used the serotype name as a species name (e.g., Salmo­nella typhimurium), and for convenience some scholars still continue to do so (Gast 2003b).

host range and distribution Salmo­nellae have a worldwide distribution (Morner 2001b), and the occurrence of salmonellae in the intestinal tract of wild animals often is correlated with the proximity of their habi­tats to those of humans or livestock (Daoust and Prescott 2007). Serovars Enteritidis and Typhimurium are the two most widespread salmonellae (Acha and Szyfres 2001). Among mammals, salmonellae are found among a wide variety of taxonomic groups throughout the world, many of which are able to serve as recovered carriers; examples include wild pigs (Sus scrofa), many deer species, hares, rodents, marsupials, insectivores, primates, and carni­vores (Morner 2001b).

The potential avian host range of the genus Salmonella appears unlimited (Daoust and Prescott 2007). Among North American birds, salmonellae are reported regularly among gulls and terns, songbirds, waterfowl, herons and egrets, and pigeons and doves (Friend 1999c); salmonellosis has been a significant contributor to mortality, particularly among the Passeriformes and Piciformes, but also among the Ciconiiformes, Pelecaniformes, and Gaviiformes (Hall and Saito 2008). Sal­monellae also commonly are reported from reptiles, fish, and insects (Janssen and Meyers 1968, Murray 1991). While salmonellae infect humans (Bopp et al. 1999), humans rarely acquire salmonellae from wildlife (Murray 1991, Acha and Szyfres 2001).

reservoirs and transmission Although salmonellae can survive for extended periods in the environment, the animal host is believed to be the primary habitat of Salmonella spp. (Winfield and Groisman 2003); salmonellae carry a number of genes that aid in the invasion of and survival in host cells (Scherer and Miller 2001). Many specific serotypes of Salmonella spp. lack special host adaptations and are capa­ble of colonizing a wide variety of hosts (Foltz 1969). Other serotypes are highly host adapted and can cause high mortality in their respec­tive hosts, but also establish long-term carrier states in some of these hosts to ensure their continued transmission (Daoust and Prescott 2007).

Among wild birds, 10% of apparently healthy coots (Fulica americana) in the Impe­rial Valley of California had Salmonella spp. in their feces (Rosen et al. 1957). However, most surveys of Salmonella in wild birds has shown are not exposed to food or water contaminated with salmonellae (Brittingham et al. 1988, Daoust and Prescott 2007). Salmonella spp. were isolated from 37% of over 300 reptiles exam­ined live or postmortem at the National Zoo in Washington, D.C., with a range of 3% in turtles to 55% in snakes; no apparent disease generally was attributed to these bacteria (Cambre et al. 1980, Acha and Szyfres 2001). A number of tropical reptiles and amphibians also regularly carry salmonellae (Kourany et al. 1970, Kourany and Telford 1981).

The occurrence of salmonellae in the natu­ral environment (water, soil, feed) often is a function of the degree of contamination by fecal material from infected hosts (Daoust and Prescott 2007). Salmonella spp. have high sur­vival rates in aquatic environments (Chao et al. 1987), and water can serve as a bacterial reser­voir and aid transmission between hosts (Foltz 1969). Some strains pathogenic to humans can occur among wildlife in relatively pristine areas (Gaertner et al. 2008)

Salmonella spp. also are widely distributed in soil and sediment (Chao et al. 1987, Abdel- Monem and Dowidar 1990) and can survive at least 12 months in soil (Thomason et al. 1977, Davies and Wray 1996), and 16 months in poultry litter (Williams and Benson 1978). Salmonella enterica Typhimurium can survive up to 9 months in soil, which, in an avian col­ony or around bird feeders, would be sufficient to maintain its presence from one year to the next (Literak et al.

Salmonellae can be isolated from the envi­ronment for one or several years (Winfield and Groisman 2003), including in very harsh condi­tions (Thomason et al. 1975). This capacity for extended survival in the environment also may be enhanced by the ability of some salmonellae to use free-living amoebae as hosts in the environment (Tezcan-Merdol et al. 2004).

Transmission commonly occurs through ingestion of fecally contaminated food or water; however, transmission by direct contact, inha­lation, and ovarian transmission also occur (Friend 1999c). Contaminated water may be particularly important for some serovars of salmonellae (Murray 1991). Among songbirds, transmission also occurs readily at bird feeding stations (Friend 1999c). Humans can acquire infections from direct contact or fecally con­taminated materials from infected pets, includ­ing turtles and rodents (Anonymous 2005a). Flies (e.g., Musca domestica) can carry and transmit Salmonella spp. (Mian et al. 2002).

clinical effects Among susceptible ani­mals, salmonellae typically cause intestinal infections resulting in primary enteritis and colitis, which may further develop into septi­cemias or abortions. Endotoxins released by the bacteria upon their death are an important cause of the signs and symptoms observed in the host (Barker et al. 1993, Clark and Gyles 1993); birds also can experience cecal infec­tions. Salmonellae can spread from the intes­tine to the liver and esophagus of infected hosts. Most Salmonella spp. intestinal infec­tions among wild birds are not associated with clinical illness and, without reinfection, are likely to last no more than a few weeks (Daoust and Prescott 2007). Among birds experienc­ing disease, there are few distinctive signs, and death may be very sudden; clinical symptoms may include ruffled feathers, diarrhea, and lethargy (Friend 1999c).

population effects Salmonella enterica serovar Typhimurium, which itself includes many strains, is the serotype most often asso­ciated with disease in wild birds (Daoust and Prescott 2007). Prior to 1980, most Salmonella spp. isolations from wild birds were taken from apparently healthy free-ranging wild birds, or were from epizootics involving small numbers of birds. However, extensive mortalities among birds using feeding stations have become com­mon in the United States, and also are reported from Canada and Europe (Friend 1999c). Birds most commonly infected at feeding stations in the United States include pine siskins (Spinus pinus), American goldfinches (Spinus tristis), evening grosbeaks (Hesperiphona vespertina), house sparrows (Passer domesticus), brown­headed cowbirds (Molothrus ater), and cardi­nals (Richmondena cardinalis) (Friend 1999c). It is apparent that several serovar Typhimurium strains have become established among some songbird species frequenting bird feeders (Daoust and Prescott 2007).

special problems Mortality from salmo­nellosis has been an extended problem among passerine birds in many parts of the world and commonly is associated with birds congregat­ing at bird feeders (Friend 1999c, Refsum et al. 2003); many cases go unreported, or are reported less formally by news services such as ProME D-mail (http://www.promedmail.org/). A small proportion of infected birds remain lifelong carriers that intermittently excrete sal­monellae into the environment. Among mam­mals, salmonellae may have a delicate balance between overt presence and undetectable levels of infection; stress on the host can shift this balance (Williams and Newell 1970).

Wild-caught slider turtles (Pseudemys spp.) used in the house pet trade in the United States have been an important source of salmonellae for humans. This problem declined after turtle importation was prohibited and a requirement was enforced for infection-free status to be established prior to any interstate commerce (Acha and Szyfres 2001).

control There is little likelihood of controlling salmonellae in the wild (Morner 2001b). Although recovered carriers are the probable source, much control is focused on eliminating point sources of infection and reducing the number of organisms discharged into the environment (Murray 1991); for sal­monellae in songbirds, one important step is keeping bird feeders clean and reducing the densities of birds present (Daoust and Prescott 2007). While humans and captive animals can be treated with a variety of antibiotics (Gast 2003a), elimination of point sources of infec­tion such as refuse dumps, untreated sewage outlets, and runoff from livestock and poultry operations is effective in reducing outbreaks in wildlife. Because of the regular association of salmonellae in wildlife with environments contaminated by humans and livestock, elimi­nating point sources of infection reduces the likelihood of carrier animals amplifying bacte­rial contamination in the environment (Daoust and Prescott 2007).

BRUCELLA spp.

causative agent Brucella spp. are gram­negative, non-motile bacteria shaped as cocci or short rods (App. 1: Table 7). The bacteria are fastidious aerobes requiring complex media and a narrow range of environmental condi­tions for successful isolation and growth in the laboratory (Shapiro and Wong 1999). Eight species currently are identified, each with distinct host specificities (Alton et al. 1988, Maquart et al. 2009).

host range and distribution At least 91 species of mammals from nine orders are susceptible to Brucella spp. (Thorne 2001). Gen­erally brucellae affect only mammals, although they have been isolated on occasion from wild and domestic birds (Acha and Szyfres 2001).

At least eight major species are recognized, each commonly linked with principal hosts: B. abortus occurs in cattle as well as bison (Bison bison), elk (Cervus elaphus), Asian buffalo (Bubalus bubalis) (Metcalf et al. 1994, Shapiro and Wong 1999), and several species of African wildlife (Godfroid 2002). Brucella suis most commonly occurs in swine, but also occurs in cattle, caribou, and reindeer (Rangifer spp.), the European hare (Lepus capensis), and rodents (Metcalf et al. 1994, Shapiro and Wong 1999). Brucella melitensis occurs in domestic goats and sheep, as well as alpacas (Vicugna pacos) and camels (Family Camelidae) (Shapiro and Wong i999)∙ Brucella canis occurs in dogs but also has been isolated from coyotes (Canis latrans) and several species of foxes (Shapiro and Wong 1999, Thorne 2001). Brucella ovis occurs primarily in sheep, and B. neotomae is known only from desert wood rats (Neotoma lepida) in Utah (Metcalfet al. 1994, Thorne 2001). Brucella ceti occurs among cetaceans and Brucella pinni- pedialis among pinnipeds (Maquart et al. 2009). An argument also has been made for three dis­tinct species occurring among dolphins, por­poises, and cetaceans (Groussaud et al. 2007).

Brucellosis is a zoonotic disease, with human infections caused by B. abortus, B. melitensis, B. suis, and B. canis (Thorne 2001), as well as B. pinnipedialis (Sohn et al. 2003, Whatmore et al. 2008). Brucella melitensis is considered the most important cause of human disease (Nicoletti 1989).

Brucellae have a worldwide distribution, although the geographic ranges for particu­lar species may vary considerably (Acha and Szyfres 1987, Thorne 2001). Brucella abortus is most widespread; in contrast, B. neotomae appears to be limited only to wood rats in Utah. Distribution of the marine species has not yet been clarified, but appears to be very broad, if serology is an indicator (Nielsen et al. 2001, Van Bressem et al. 2001, Foster et al. 2002, Neimanis et al. 2008).

RESERVOIRS AND TRANSMISSION Both domestic animals (Shapiro and Wong 1999) and wild ungulates (Acha and Szyfres 1987) have been proposed as reservoirs for terrestrial brucellae. Thorne (2001) notes that, except for B. suis biovar 4, B. neotomae, and possibly brucellae from marine mammals, free-ranging wild mammals are not primary reservoir hosts of brucellae (Thorne 2001). Brucellae generally are transmitted by ingestion, inhalation, and direct contact via skin abrasions and mucous membranes, and the conjunctiva (Shapiro and Wong 1999). Transmission of B. abortus among wild ungulates most typically occurs by licking and ingestion of the bacteria found on fetal membranes, on an aborted fetus, or in uterine discharge (Thorne 2001).

Although brucellae can survive for an extended period in the environment (Wray 1975), there is no evidence that the bacteria can replicate outside of a host (Thorne 2001). There also is no direct evidence for arthropod transmission (Thorne 2001). Among marine mammals, Brucella spp. have been isolated from lungworms, and it is proposed that lung­worms may be a means of transmission among marine mammals (Garner et al. 1997, Dawson et al. 2008).

clinical effects Brucellosis in ungu­lates usually is characterized by abortion in the female and, to a lesser extent, orchitis and infection of accessory sex glands in the male; infertility can occur in both sexes (Fraser and Mays 1986). Abortion and retained placentas are the most common signs of B. abortus infec­tion in bison (Thorne 2001). Abortion and birth of nonviable calves were the most frequent and important signs of brucellosis in captive elk studies; these same signs have been observed in free-ranging elk (Thorne 2001). Among marine mammals, intracerebral infections and associated neurological problems can occur (Hernandez-Mora et al. 2008, Alba et al. 2013).

There are several surveillance programs to determine brucellae presence within a popula­tion or geographic area; these usually involve serologic tests. Although many serological tests have been tried for bison, none is reliable alone for diagnosing Brucella spp. infection, however; combinations of tests have been fairly success­ful (Davis et al. 1990). Similarly, for elk, no one test is satisfactory, but a combination of tests can be useful (Thorne 2001). Culture of Brucella spp. from tissue or fluid is the defini­tive basis for diagnosing infection (Thorne 2001), and a multiplex single-nucleotide poly­morphism detection assay is available for the six classic Brucella spp. as well as the marine mammal groups (Scott et al. 2007).

population effects Among wildlife, Brucella abortus has been studied extensively in bison and elk. Although B. abortus causes abor­tions in bison, the incidence of abortion among free-ranging bison is unknown (Thorne 2001). Combining several studies, loss of the first calf following infection was 57% among elk, and one of nine cows monitored lost a second calf (Thorne 2001). Brucellosis (along with tuber­culosis) is associated with a decline in the bison population of Wood Buffalo National Park (Joly and Messier 2005).

special problems Brucella abortus infection among bison and elk has been controversial in North America because of concerns that domes­tic cattle could become infected through contact with infected bison and elk; U.S. states or Cana­dian provinces with infected cattle likely would lose their brucellosis-free status, with the conse­quent potential for serious economic losses.

Brucellosis among free-ranging bison has occurred in several areas of North America, but is of special concern in the Greater Yellowstone area of the western U.S., and in the Wood Buffalo National Park of Alberta, Canada (Thorne 2001). Brucella abortus has occurred in the Greater Yellowstone ecosystem since at least 1917, infecting both bison and Rocky Mountain elk (Thorne 2001). While B. abortus can be transmitted to cattle under experimen­tal conditions from both bison (Davis et al. 1990) and elk (Thorne et al. 1979), the contin­ued absence of brucellosis in cattle that associ­ate with infected bison is evidence that the risk of transmission is low (Thorne 2001). However, management of infected bison in the vicinity of domestic cattle in Yellowstone National Park has continued to be very controversial (Petersen and Graham 1996, Bransom 1998).

Among free-living elk, B. abortus also is present only in the Greater Yellowstone Area of North America. Prevalence among elk on feeding grounds ranged from 13 to 58% but was only 2.2% where elk wintered on native range adjacent to the feeding ground (Toman et al. 1997). There also is evidence that infected free-ranging elk would not be a likely source of brucellosis for cattle and that elk populations probably are not capable of maintaining brucel­losis in the absence of concentrations caused by winter feeding grounds (Thorne et al. 1997, Thorne 2001). However, infected elk were identified as the likely source of infected cattle in Montana, resulting in Montana's loss of bru­cellosis-free status (Anonymous 2008). It also is argued that elk are the greater disease threat to cattle, compared to bison, but that elk have less disease control intervention because hunt­ers and outfitters bring considerable revenue to the region; a gradual program to phase out elk feeding grounds has been recommended as the best means to reduce or extirpate brucellosis in elk (Bienen and Tabor 2006).

Brucella abortus also has been of special con­cern among the bison of Wood Buffalo National Park (Alberta, Canada). Two factors contribute to the controversy. The current bison are an introduced strain, different from the original bison that had been extirpated earlier. And this new strain of bison also introduced B. abortus and Mycobacterium bovis to the site (Tessaro 1987, 1989a, 1989b; Gates et al. 1997). Canada's national cattle herd achieved brucellosis-free status in 1985, and bovine tuberculosis also was nearly eliminated (Thorne 2001). To complicate matters, a remnant population of the original bison strain later was located in the nearby region of Canada. The Federal Environmen­tal Assessment Review Panel recommended a lethal removal of the infected bison, but a strong public opposition to this plan developed. Further studies are being conducted (Joly and Messier 2005).

control Control of brucellosis among bison and elk has been controversial. For many years, bison that left Yellowstone National Park were either shot or first captured and then tested, with Brucella-positive animals then slaughtered to reduce the likelihood of trans­mission of B. abortus to cattle on surround­ing areas (Cheville et al. 1998). While test and slaughter, in combination with vaccination, has been used to eradicate brucellosis from con­fined ranch and park bison herds, logistic dif­ficulties and general lack of public acceptance make test and slaughter and depopulation strat­egies unlikely to be successful with large free- ranging wild mammal populations (Thorne 2001, Bienen and Tabor 2006).

An Interagency Bison Management Plan for the Greater Yellowstone Area was developed with the National Park Service and the State of Montana (Anonymous 2000), involving three federal agencies (National Park Service, For­est Service, Animal and Plant Health Inspec­tion Service) and two Montana state agencies (Department of Fish, Wildlife and Parks; Department of Livestock). The plan calls for an ongoing removal of infected bison in concert with a vaccination program, with the goal of eradicating B. abortus from the ecosystem.

The Bison Management Plan includes three steps (Anonymous 2000). In Step 1, the National Park Service prevents bison, by haz­ing procedures, from leaving the park during periods when they could come in contact with susceptible cattle. When hazing fails, bison are captured, seronegative animals can be vac­cinated, and seropositive animals are slaugh­tered. Step 2 of the plan is designed to allow some seronegative bison to move off the park onto areas used by cattle during winter, after cattle are removed from those pastures. In Step 3, some untested bison will be allowed in areas overlapping with cattle, but must be back in the park before cattle are released onto these pastures in the spring. Vaccination also is to be used, with the expectation that this process eventually will lead to the elimination of brucellosis in bison.

Although there appears to be little risk of transmission of Brucella abortus between bison and cattle in Wood Buffalo National Park, the only practical solution proposed thus far has been a call for the complete elimination of infected bison, with creation of a plan for dis­ease elimination and reestablishment of the original bison strain (Anonymous 1990). This has not yet been done.

Ultimately, successful brucellosis control programs may require extensive vaccination programs, as well as a more complete under­standing the disease in both wildlife and domestic animals (Davis et al. 1990). Currently, an attenuated live calf vaccine strain for cattle, RB51, is being evaluated for use among both bison (Olsen et al. 2002b, 2003) and elk (Olsen et al. 2002a). This vaccine is viewed as the most promising vaccine in bison (Olsen et al. 2003). Brucellosis-infected bison and elk continue to provide a significant challenge to wildlife and livestock managers.

PASTEURELLA SPP. AND MANNHEIMIA SPP.

classification Pasteurella spp. are gram­negative, facultatively anaerobic non-motile, rod-shaped bacteria (App. 1: Table 7) (Holmes et al. 1999). About 20 species are included in the genus, although the taxonomic classifica­tions of several are being revised (Holmes et al. 1999). Recently, P. haemolytica was reclassified as Mannheimia haemolytica. However, because M. haemolytica has many similarities with other pasteurellae, we share the view of some (Miller 2001) that it still be included with this group. Among pasteurellae, the most impor­tant species affecting wild birds are P. multo­cida, with P. multocida and M. haemolytica most important among mammals.

host range and geographic distribution Most species of birds and mammals probably can be infected with P. multocida (Miller 2001, Samuel et al. 2007); however, avian strains typically infect only birds, and mammalian strains only mammals, with rare exceptions (Gregg et al. 1974, Smit et al. 1980, Korbel 1990). Pasteurella multocida and M. haemolytica both have a worldwide distri­bution (Miller 2001, Samuel et al. 2007). Among mammals, P. multocida has been isolated from most mammalian families (Mutters et al. 1989). For birds, close to 200 species, among 44 fami­lies, have been reported as naturally infected by P. multocida (Samuel et al. 2007).

reservoirs and transmission Many wild mammals carry one or more strains of pas- teurellae as commensal flora (Brogden and Rhoades 1983, Quan et al. 1986). Pasteurella multocida occurs in the mouths, respiratory tracts, and intestinal tracts of a variety of mammals, including the respiratory tracts of domestic sheep (Foreyt and Jessup 1982), as well as the mouths of Norway rats (Rattus norvegicus) (Schipper 1947), domestic dogs and cats (Owen et al. 1968, Hubbert and Rosen 1970, Arnbjerg 1978), and a variety of wild mammals (Owen et al. 1968, Bond et al. 1972, Quan et al. 1986). In an unusual case, a number of Newfoundland (Canada) caribou (Rangifer tarandus) calves died from cervical and neck abscesses of Pasteurella multocida type A infections following failed predatory attempts by lynx (Lynx lynx) (Bergerud 1971).

Recovered birds are carriers for avian strains of Pasteurella multocida in domestic fowl (Wobeser 1992), and there is increasing support that this also is the case among wild­fowl (Samuel et al. 2005, 2007). While an alter­native reservoir proposed for avian strains of P. multocida has been soil or water of wetlands (Botzler 1991), recent evidence does not sup­port that hypothesis (Blanchong et al. 2006a, Samuel et al. 2007). However, avian cholera strains of P. multocida can grow and multiply in free-living amoebae (Acanthamoeba polyphaga, Hartmanella vermiformis) (Hundt and Ruffolo 2005); such interactions also could contribute to their extended survival within natural environments under some circumstances.

Transmission between mammals can be through inhalation following nasal contact (Foreyt and Jessup 1982, Foreyt 1989), as well as by bites of infected animals (Hubbert and Rosen 1970). However, the young of many host species likely are colonized with commensal strains at birth during postpartum interactions between the mother and young (Miller 2001). Transmission to susceptible birds can occur through inhalation of contaminated aerosols, ingestion of contaminated water, cat bites, and arthropod bites (Botzler 1991). However, nei­ther cat bites nor arthropod bites are associated with avian cholera epizootics among wildlife.

CLINiCAL effects Among mammals, clini­cal signs of pasteurellosis resulting from bite or scratch wounds often involves painful local­ized swelling, with an accompanying fever and depression (Miller 2001). Pneumonic pasteurellosis in ruminants is characterized by depression, anorexia, mucopurulent nasal discharge, coughing, and respiratory distress (Miller 2001). Pasteurellae may be involved with lungworm-pneumonia complex among bighorn sheep; this syndrome is an interac­tion between the lungworm Protostrongylus stilesi and concurrent infections involving Parainfluenza-3 viruses, M. haemolytica, P. multocida, and Corynebacterium spp. Stress often appears to initiate or increase the sever­ity of this disease. Among rabbits, a purulent nasal discharge (snuffles) is characteristic of enzootic pasteurellosis.

Most wild birds die from avian cholera with few preliminary warning signs (Samuel et al. 2007). It is common to find dead birds interspersed with apparently healthy birds, and with few or no sick birds evident. In the acute disease, fever, systemic hypotension, shock, and rapid death may occur (Collins 1977). In more prolonged cases, petechial hemorrhages of the heart and necrotic foci of the liver also develop.

Diagnosis of pasteurellosis and avian chol­era is determined most definitively by isola­tion and identification of the causative agents (Miller 2001, Samuel et al. 2007). Serotyping, biotyping, biogrouping, genomic fingerprint­ing, and polymerase chain reaction-based assays are used for further characterizations of suspected epizootics (Jaworski et al. 1997, Green et al. 1999, Miller 2001, Blehert et al. 2008). Use of amplified fragment length poly­morphism analysis has allowed distinguishing of many isolates among serotype ι P. multocida, and holds considerable promise in further elu­cidating the epizootiology of avian cholera in wildfowl (Blehert et al. 2008).

population effects Among mammals, pasteurellae most commonly function as endemic, opportunistic pathogens readily influ­enced by predisposing factors such as trauma, stress, or concomitant infections (Miller 2001). Thus, many cases of pasteurellosis are indi­vidual occurrences only. However, epizootic disease can occur among free-ranging animals when a novel pathogen is introduced to a sus­ceptible population (Miller 2001). In particular, wild sheep (Ovis canadensis, O. dalli) are quite susceptible to pasteurellae from domestic sheep, and population losses have occurred among populations of wild sheep (Foreyt and Jessup 1982, Foreyt 1989, Foreyt et al. 1996, George et al. 2008); some bighorn sheep popu­lations may be limited by recurrent pasteurel­losis epidemics (Hobbs and Miller 1992).

Among birds, population impacts are con­sidered at three levels of assessment: local wintering or breeding populations, regional populations, and continental species (Samuel et al. 2007). Impacts can vary among species depending on their abundance, number of individuals at risk of infection, species and individual susceptibility, and behavioral char­acteristics that increase the risk of exposure to P. multocida (Samuel et al. 2007). There are no consistent relationships of susceptibility to spe­cies, age, sex or nutritional status of the host (Samuel et al. 2007).

Avian cholera epizootics are most studied on wintering grounds, where they usually are explosive in wildfowl, appearing with little warning (Botzler 1991). In North America, waterfowl and coots (Fulica americana) expe­rience the greatest known mortality, followed by crows (Corvus spp.) and gulls (Larus spp.) (Friend 1999a, Samuel et al. 2007). Wildfowl epizootics commonly are associated with dense concentrations of susceptible birds involving many species (Blanchong et al. 2006b). Observed mortalities can be consider­able and probably are underestimates because of the difficulties in finding and collecting all carcasses (Humburg et al. 1983, Stutzenbaker et al. 1986). There also are regular occurrences of low-level, chronic mortalities among some populations (Wobeser 1992, Botzler 2002). These chronic low-level mortalities may occur throughout the year and may be a significant proportion of the total annual losses for some populations (Wobeser 1992).

Although there have been years with unusu­ally severe and widespread epizootics, there is little evidence that avian cholera has substan­tially impacted continental populations of birds (Samuel et al. 2007). At local levels, avian chol­era impacts vary from minor (90% to songbirds appear similar to each other, and differ from known poultry strains from which they probably are derived (Ley et al. 1997, Mikaelian et al. 2001).

Among reptiles, Mycoplasma agassizii has been associated with upper respiratory tract dis­ease of desert tortoises (Gopherus agassizii), either alone or in combination with Pasteurella testudi- nis; however, the mycoplasma appear to play a more significant role than the pasteurellae in the expression of this disease (Brown et al. 1994).

reservoirs and transmission Among both mammals and birds, healthy and recov­ered carriers typically are the basis of repeated occurrences of mycoplasmosis in populations; inapparent carriers may well persist once infections are established in wild populations (Friend 1999b, Whithear 2001, Luttrell and Fischer 2007). Most mycoplasma of mammals are transmitted between hosts by direct contact or aerosol transmission (Whithear 2001).

Infective keratoconjunctivitis, caused by M. conjunctivae, is a highly contagious disease among wild chamois and ibex, and is transmit­ted predominantly by direct contact; however, aerosol transmission as well as spread by eye­frequenting insects also can occur (Loison et al. 1996, Giacometti et al. 2002a). It appears that M. conjunctivae is not self-maintained in Alpine chamois, however, and requires another source as a reservoir, possibly domestic sheep (Giacometti et al. 2002b).

Transmission of M. gallisepticum commonly occurs by aerosols of droplets and dust, as well as by direct contact of susceptible birds with infected carriers; carrier birds are essential to the epizootiology the disease (Ley 2003, Luttrell and Fischer 2007). Mycoplasma gallisepticum is believed to spread due to contact with infected birds at bird feeders (Dhondt et al. 1998); bird feeders also can serve as fomites. Recovered birds likely become carriers of M. gallisepticum.

For upper respiratory disease of tortoises, transmission of M. agassizii may follow direct contact or aerosol transmission of the myco­plasma; mucous droplets in the burrows are a possible source of infection (Brown et al. 1994).

clinical effects Among mammals, most pathogenic mycoplasmas invade moist mucosal surfaces of the respiratory and genital tracts, where they establish persistent and often clini­cally inapparent infections (Whithear 2001). Infectious keratoconjunctivitis in mammals resulting from M. conjunctivae is specifically a disease of the eyes, with no other tissues affected (Giacometti et al. 2002a). The first vis­ible signs are ocular effusions containing high concentrations of mycoplasma; characteristic effects are a mild keratitis that can develop into a severe edema, and corneal ulceration that eventually leads to permanent blindness (Giacometti et al. 2002a).

Mycoplasma gallisepticum causes conjunc­tivitis in songbirds, with lesions limited to eyes, nasal turbinates, and trachea (Luttrell et al. 1998). In wild turkeys, infections involve periocular swelling and sinusitis (Davidson et al. 1982). With experimen­tal infections, the organism also can cause reproductive suppression among galliform birds such as pheasants, chukar partridges (Alectoris graeca), and peafowl (Cookson and Shivaprasad 1994).

Upper respiratory disease of tortoises is characterized as a chronic disease eventually leading to severe occlusion of the nares, with viscous exudates and destruction of the respira­tory epithelium (Brown et al. 1994, Christopher et al. 2003). A variety of shell lesions and plague-like oral lesions also are characteristic, and tortoises with oral lesions were most likely to be culture-positive for M. agassizii (Christopher et al. 2003).

Diagnosis of mycoplasmosis is based on clinical signs, history, and detection of the organism by culture or molecular techniques (Whithear 2001, Luttrell and Fischer 2007). Isolation and identification is the most certain way to identify Mycoplasma spp. (Ley 2003).

population effects There is only lim­ited information available on the population impacts of mycoplasmas on wild populations (Friend 1999b, Whithear 2001). Following a population decline of about 37% among Alpine chamois in the northern French Alps, the population required approximately 5 years to recover (Loison et al. 1996).

Population impacts directly related to mycoplasmosis in wild turkeys are difficult to assess (Rocke and Yuill 1987, Hoffman et al. 1997). Mycoplasmal conjunctivitis spread relatively quickly in house finches (Carpodacus mexicanus) (Dhondt et al. 1998). Mycoplasma gallisepticum appeared to suppress free-flying house finch populations (Luttrell et al. 1998). In one study, house finch numbers decreased throughout winter in areas of the United States with cold winters and high conjunctivitis prevalence, giving further evidence for signifi­cant mortality from this disease (Dhondt et al. 1998). The occurrence of mycoplasmosis in other songbirds may result from a spill-over of an epizootic in house finches, rather than a sustained interspecific transmission (Hartup et al. 2001b). There is evidence that some host populations are becoming resistant or that the organism is evolving to these new hosts (Luttrell et al. 1998, Hartup et al. 2001a, Roberts et al. 2001, Luttrell and Fischer 2007)

special problems Mycoplasma gallisepti- cum was first observed in house finch popu­lations of the eastern United States in early 1994, when it caused a severe conjunctivitis (Fischer et al. 1997). Between November 1994 and March 1997, it spread rapidly through the eastern house finch population, partly following migratory routes, with evidence of significant mortality associated with the con­junctivitis among the house finch population (Dhondt et al. 1998). The mycoplasma spread to other songbirds as well, including the American goldfinch (Friend 1999b). Increased risk of conjunctivitis among house finches was associated with cooler nonbreeding seasons (September to March), as well as bird feeders containing platforms, hoppers, or tube-type feeders (Hartup et al. 1998). The disease now appears well established in eastern popula­tions of house finches. Western populations of house finches, while not yet infected, are very susceptible (Farmer et al. 2002).

control For M. conjunctiva in chamois, post-epizootic management is recommended as a response to infections, rather than elimi­nating infected animals (Loison et al. 1996). Immunoprophylaxis in domestic sheep or limiting interspecific transmission of the mycoplasma from sheep to chamois have been recommended (Giacometti et al. 2002b). Avoid­ing unnecessary human disturbance to wild animals also is seen as an important tool of prevention (Giacometti et al. 2002a).

Measures for prevention of the spread of mycoplasmal conjunctivitis in songbirds include modifying bird-feeding activities based on sea­son and type of feeder. For feeders, this would include reducing competition and crowding potential by having more spacious feeding strate­gies using raised platforms (Hartup et al. 1998).