Salmonellosis was diagnosed in garden birds from 198 incidents in Scotland between September 1995 and August 2008. Salmonellosis was essentially a disease of finches in the north of Scotland, but in the south of Scotland it was also a problem in house sparrows. Almost all of the incidents were caused by Salmonella Typhimurium phage types 40 or 56/variant, but regional variation in phage types was observed. In the north of Scotland, one phage type (DT 40) predominated, but in the south of Scotland two phage types were commonly isolated (DTs 40 and 56/variant, with the latter the more common of the two phage types). This regional difference was statistically significant for salmonellosis in greenfinches, chaffinches and ‘other garden birds’, but not for house sparrows. Different temporal patterns for different species of bird and different phage types were also observed within regions. These findings suggest that the epidemiology of salmonellosis in garden birds varies depending on the phage type of Salmonella and the species of garden bird, with additional regional differences depending on the wild bird populations and the phage types of Salmonella in circulation. An awareness of these differences will help when formulating guidelines aimed at reducing the impact of salmonellosis in garden birds.
Statistics from Altmetric.com
Salmonellosis has been the cause of outbreaks of mortality in garden birds in several areas of the world, including North America (D’Aoust and others 2000, Hall and Saito 2008), Sweden (Tauni and Osterlund 2000), New Zealand (Alley and others 2002), Norway (Refsum and others 2002) and Great Britain (Pennycott and others 2006). The causal organism, originally designated Salmonella typhimurium, has now been reclassified as Salmonella enterica subspecies enterica serotype Typhimurium (S Typhimurium) (Le Minor and Popoff 1997). Much of the early work investigating deaths in garden birds was carried out by Jim MacDonald at the Veterinary Laboratory in Lasswade, Scotland; for example, Wilson and MacDonald (1967), MacDonald and Cornelius (1969), MacDonald and Brown (1974). Around the same time, bacteriophage-typing schemes were being developed to differentiate isolates of S Typhimurium, and Anderson and others (1977) presented an updated list of definitive phage types (DTs). Between 1968 and 1985, MacDonald and colleagues at Lasswade recorded 194 isolates of S Typhimurium from garden birds, mostly from greenfinches (Carduelis chloris) and house sparrows (Passer domesticus) (G. Bennett, unpublished data). Based on the accepted nomenclature of Anderson and others (1977), 70 isolates were DT 40, 64 isolates were DT 160, 39 isolates were DT 129, 14 isolates were DT 37, six isolates were DT 161 and one isolate was DT 110. Most of the incidents occurred in eastern and south-east England, but carcases were also received from other parts of England and southern Scotland.
Pennycott and others (2006) diagnosed salmonellosis due to DT 40 and 56/variant in garden birds from 63 sites (mostly in Scotland) between 1995 and 2003. These two phage types have also been recovered from cats with enteric disease in the UK (Philbey and others 2008), and the Scottish Salmonella Reference Laboratory has identified these two phage types in over 75 people since 2001 (Philbey and others 2008). Possible geographic regional variations and interactions between phage type and bird species were not considered in the paper by Pennycott and others (2006), and the presentation of the data by carcase numbers rather than incidents could have obscured trends if large numbers of carcases were submitted from some of the incidents. The data were therefore re-examined to look for possible regional variations within Scotland, looking at incidents rather than carcase numbers. The database was also substantially expanded to include results from a further five years and to include results from other diagnostic laboratories in Scotland that had examined the carcases of garden birds, thus reducing possible sources of bias.
Materials and methods
Dead or moribund wild birds were submitted directly or indirectly by members of the general public to SAC Veterinary Services – Ayr, to other laboratories of SAC Veterinary Services, and to Veterinary Laboratories Agency (VLA) – Lasswade. Some of the carcases examined after April 2005 were submitted as part of the Garden Bird Health Initiative (Cunningham and others 2005). Bacteriology was carried out as described by Pennycott and others (2002). The results were expressed as incidents of salmonellosis rather than isolations: an incident of salmonellosis included the first and subsequent isolation(s) of a particular phage type of S Typhimurium from garden bird(s) on a single site within a Salmonella seasonal year. The seasonal year was taken to run from September to August rather than January to December because of the seasonal nature of salmonellosis (Pennycott and others 2006). The month of onset of the incident was taken to be the month in which the first carcase with confirmed salmonellosis was found.
Sites submitting garden birds were divided into those from the north of Scotland (Scottish Unitary Authority areas of Highland, Moray, Aberdeenshire, Aberdeen City, Western Isles, Orkney Islands, Shetland Islands), referred to as region 1, and the remainder of Scotland (region 2). These areas correspond to the British Trust for Ornithology regions of North Scotland and South Scotland.
The data were analysed using a computerised statistical software package (GenStat release 11.1; VSN International). Statistical significance was set at P<0.05.
Geographic region, species of bird and phage type
Salmonellosis was diagnosed in garden birds from 198 incidents in Scotland, most commonly affecting greenfinches, chaffinches (Fringilla coelebs) and house sparrows. Other garden birds with salmonellosis were goldfinches (Carduelis carduelis), siskins (Carduelis spinus), great tits (Parus major), tree sparrows (Passer montanus), dunnocks (Prunella modularis) and a brambling (Fringilla montifringilla). A total of 111 incidents were caused by S Typhimurium DT 40, and 85 incidents by S Typhimurium DT 56/variant (Table 1). A further two incidents were caused by S Typhimurium DT 41. A total of 88 incidents occurred in region 1 and 110 incidents were recorded in region 2. In region 1, most of the incidents (88 per cent) involved S Typhimurium DT 40 (Fig 1). In region 2 there was more fluctuation between years, but overall 31 per cent of incidents involved DT 40, and 68 per cent of incidents involved DT 56/variant (Fig 2). This pattern of a greater prevalence of DT 40 in the north of Scotland and DT 56/variant in the south of Scotland was analysed using a twotailed Fisher’s exact test and was statistically significant for greenfinches (P<0.001), chaffinches (P=0.012) and ‘other garden birds’ (P<0.001) (Table 1). A similar though less pronounced trend was seen for house sparrows, but this was not statistically significant (P=0.119).
Months of onset of incidents
In both regions, most (87 and 88 per cent) new incidents in greenfinches caused by DT 40 were first reported in the months January to March, peaking in January/February (Fig 3). However, the pattern for DT 56/ variant in greenfinches in region 2 was different, with 90 per cent of new incidents occurring in November to March, again peaking in January/February (Fig 3). New incidents of salmonellosis in chaffinches mostly occurred in January or February (Fig 4), regardless of phage type or region. The month of onset of incidents of salmonellosis in house sparrows is shown in Fig 5; unlike the situation in greenfinches and chaffinches, no marked peaks in January/February were noted.
Overall there were 111 incidents of DT 40, 85 incidents of DT 56/ variant and two of DT 41. Two main phage types (DT 40 and DT 56/ variant) appear to be circulating in wild birds in the south of Scotland, but in the north DT 56/variant has not yet become widespread. This small range of phage types from Scotland contrasts with the findings of MacDonald and colleagues (unpublished data), who, in addition to DT 40, also commonly recovered DTs 160, 129, 37 and 161 from garden bird carcases from eastern and south-east England between 1968 and 1985. It is also interesting to note that MacDonald did not demonstrate DT 56/variant in any of the garden birds examined. A comparison of these results suggests that the phage types of S Typhimurium involved in garden bird mortality may vary with the region of the country, with time, or with a combination of both factors. Monitoring of garden bird mortality should be continued, to detect any future changes in the pattern of phage types in circulation.
Salmonellosis was essentially a disease of finches (greenfinches, chaffinches, goldfinches and siskins) in the north of Scotland, but in the south it was additionally a problem in house sparrows. Further work is required to determine if these differences reflect the population size and behaviour of house sparrows in the north of Scotland compared with the south, or whether they are the consequence of predominantly one phage type circulating in the north compared with two phage types in the south.
Although most new incidents of salmonellosis were reported in the winter months, different seasonal patterns were observed depending on the species of bird and phage type of Salmonella. These findings suggest that the epidemiology of salmonellosis in garden birds varies depending on the phage type of Salmonella and the species of garden bird, with additional regional differences depending on the wild bird populations and the phage types of Salmonella in circulation. An aware- ness of these differences will help when formulating guidelines aimed at reducing the impact of salmonellosis in garden birds. Increased hygiene or a reduction in the amount of food offered to wild birds may be required at times of greatest risk, based on knowledge of the phage types in circulation and the commonest species of birds visiting the feeding stations. For example, in areas where DT 40 circulates, greenfinches are at greatest risk during January to March and disease control measures should be enhanced during these months. However if DT 56/variant is known to be present in the region, the risk period should be extended to include the months November to March. Similarly, awareness of the differences in the regional and temporal patterns of salmonellosis in garden birds will be of value when investigating outbreaks of salmonellosis caused by DTs 40 and 56/variant in human beings or livestock, and when issuing guidelines aimed at reducing salmonellosis in human beings and livestock.
Attempts should be made to minimise possible sources of bias in any surveillance programme. In the present study, the inclusion of data from several diagnostic laboratories in Scotland permitted the addition of results from carcases over a greater geographic distribution. Similarly, participation in the Garden Bird Health Initiative in the final three years of the study increased the numbers of members of the public submitting carcases from both regions of Scotland. Analysis of the results as the number of incidents rather than the number of carcases removed the potential for the data to be skewed if large numbers of carcases were submitted from a small number of incidents. With the accumulation of increased numbers of data points from both regions over a long period of time, it is becoming apparent that salmonellosis of garden birds is not one single disease but varies depending on the region of the country, the phage types of S Typhimurium present in the wild bird population and the species of garden bird involved. These patterns will not remain static but are likely to change as existing factors change or as new factors arise, for example, a new seasonal disease of garden birds such as trichomonosis that emerged in Great Britain in 2005 (Pennycott and others 2005, Duff and others 2007). To increase the understanding of garden bird salmonellosis and its implications for individual bird welfare, wild bird population dynamics and possibly human, domestic cat and livestock health, it is essential that monitoring of garden bird mortality continues.
Much of this work was funded by the Dulverton Trust, the Game Conservancy Trust (now the Game Conservation and Wildlife Trust) and the Garden Bird Health Initiative. The work was also part-funded by the Scottish Government as part of its Public Good Veterinary and Advisory Services. VLA – Lasswade provides Defra with wild bird surveillance as part of the VLA Diseases of Wildlife Scheme.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.