A costly case of the flu

An adult ostrich in the Oudtshoorn area. Note the yellow stock tag on the back of the neck. A typical enclosure on a chick-rearing farm. The ostrich production system typically involves multiple stages beginning with hatching birds in a hatchery, then moving them to a chick-rearing facility for 2-3 months, followed by an adult-rearing farm for 9-12 months, before they spend 14 days in quarantine prior to going to the abattoir for slaughter. These different specialised production stages are often on separate farms, resulting in the movement of large numbers of birds. SAEON student Christine Moore photographed with the legendary broadcaster and naturalist Sir David Attenborough, while on a visit to the University of Cape Town during the filming of the BBC series Africa.

By Dr Jasper Slingsby, SAEON Fynbos Node and Christine Moore, SAEON student

Disease outbreaks are increasing in frequency and are of growing concern internationally.

This trend is likely to be due to a combination of globalisation increasing the opportunities for human-mediated spread of disease and increased disease susceptibility in social and ecological systems due to degradation or poor management of the system.

Beyond direct threats to human health, there are many diseases that threaten domestic and wild biological resources and are of considerable concern for the conservation of rare and endangered species, for the economic viability of various farming industries, and for food security.

Major questions with respect to these diseases are: a) How do diseases move within either domestic or wild systems and how does human management of these systems promote or prevent outbreaks? And b) When and where are we likely to see spillover of generalist pathogens between domestic and wild species?

Bird flu in the ostrich industry

Christine Moore, a SAEON student and graduate of the University of Cape Town’s (UCT’s) Conservation Biology Master’s programme, has recently published her dissertation work exploring the evolution of vulnerability to Avian Influenza within South Africa’s ostrich farming industry.

This work, co-authored with supervisor Graeme Cumming (UCT’s Percy FitzPatrick Institute of African Ornithology) and co-supervisors Jasper Slingsby (SAEON Fynbos Node) and John Grewar (Department of Agriculture), was published in open-access journal PLoS One. The study uses social network analysis techniques to explore the movement of ostriches between farms in the Western Cape in the five years leading up to the 2011 outbreak of Highly Pathogenic Avian Influenza (H5N2). This outbreak caused severe economic losses and near-collapse of the ostrich industry, resulting in the government paying out over R65 000 000 in compensation.

The study found that as time progressed, and the production system expanded and became better geared towards increasing economic efficiency, the network became increasingly vulnerable to pathogen outbreaks. Farms that became infected during the outbreak showed significantly higher connectivity (i.e. had more links to other farms) and centrality (birds moved between other farms often passed through these farms), which predisposed them to be more vulnerable to contracting and spreading the disease.

Recommended policy and management interventions

The study developed several recommendations for policy and management interventions that could reduce the epidemic potential in the ostrich industry, reduce the probability of spillover to wild populations, and improve monitoring to facilitate earlier detection of infected birds.

Vulnerability could be reduced by:

• Reducing the rate at which the disease could spread among farms by constraining the direction of exchanges such that farms could not transfer birds to farms from which they have recently received transfers.
• Reducing the potential size of outbreaks by fragmenting and compartmentalising the industry. This can be done by reducing or limiting the numbers of farms that are permitted to transfer animals between one another, creating localised "neighbourhoods" of interacting farms.

Early detection of H5N2 outbreaks is constrained by financial and logistical limitations on testing birds for the disease. These efforts could be made more efficient if:

• monitoring is focused on farms with high vulnerability scores (e.g. connectivity and centrality), because these are likely to become infected early in the outbreak;
• monitoring/testing is more frequent and intensive for transfers of birds between farms with high node-level vulnerability scores; and
• when an outbreak is first detected, the known network of bird movements can be used to implement a network quarantine zone around infected farms. This would include intensively testing farms that have received or sent birds to the infected farm within a specific time period.
  

A visualisation of the ostrich movement network using social network analysis techniques. Nodes represent farms and links represent bird transfers. Node size is proportional to the degree to which each farm is connected. Farms that tested positive for H5N2 are shaded red.

While enforcing regulations of this nature might incur additional costs, reduce individual farmer profits and reduce the short-term economic efficiency of the production system, it would greatly reduce the vulnerability of the system to future outbreaks of H5N2 or any other pathogen and avoid severe economic losses and the collapse of the industry.

Broader ecological relevance

Studying disease dynamics in commercial animal production systems is of great ecological and conservation significance, for a number of reasons. Firstly, disease outbreaks in animal production systems typically occur on a scale rarely seen in wild populations and present a significant threat for spillover into wild bird populations. This can threaten rare and endangered species, but can also decimate common wild bird populations and threaten important ecosystem functions such as pollination, seed dispersal, pest control and other trophic interactions.

Secondly, large comprehensive datasets on disease dynamics in wild populations are rare due to the time, effort and financial resources required to collect them, but are relatively easy to collect or infer from other data commonly recorded in production systems. Studies of “artificial” ecosystems such as these, allow the development of methods and the identification of generalities that may be relevant to wild populations.

For example, one could draw an analogy between farms for the domesticated birds and habitat fragments for wild bird species. One could then examine bird movements between these fragments to test whether individuals behave in such a way as to minimise their potential to contract disease, and thus limit the potential for catastrophic outbreaks. The beauty of this kind of ecological research is that it is of great relevance for commercial production systems. If wild bird populations do exhibit behaviour that reduces disease spread, then production systems would do well to adjust their protocols to mimic this behaviour.

Further research

Further information on the research programme exploring Avian influenza and other pathogens in ostriches and wild bird populations can be found here.