Bird Flu’s Industrial Revolution

Public talks can be private affairs. In sharing our work we bare the triumphs and limits of a thinking born in long hours alone or with a few fellow conspirators sworn to secrecy. But in breaking our oaths we are able to take the next step.

In talking bird flu twice this past month I learned something new.

Much of the talk was dedicated to describing the role industrial poultry played in the evolution and diversification of highly pathogenic influenza, including H5N1, bird flu’s marquee star. The evidence is presently circumstantial, but for me offers a convincing working hypothesis:

• A phylogenetic burst of influenzas capable of infecting humans coincided with the globalization of the industrial model of poultry: H5N1, H7N1, H7N3, H7N7, H9N2, in all likelihood H5N2 and perhaps some of the H6 serotypes. 
• Duan et al. (2007) identified low pathogenic relatives of more deadly H5N1 in migratory birds dating back to the 1970s. None of the recent low pathogenic H5 relatives became established in aquatic or terrestrial poultry. The origins of recent H5 virulence appear characteristic of domestic poultry alone.
• Vijakrishna et al. (2008) meanwhile showed the source 1996 H5N1 strain entered southern Chinese poultry with all eight genomic segments intact. The subsequent diversification into multiple genotypes, including the deadly Z genotype that characterizes outbreaks since 2003, began in domestic ducks in China mid-1999 to 2000.
• Industrial poultry populations are characterized by little genetic diversity and offer few immune firebreaks against outbreaks. Populations packed together facilitate viral transmission. Bird immune systems are depressed under such densities. And the high turnover rate of poultry processing—the duration from birth to sacrifice has been reduced to 40 days—likely selects for virulent influenza strains that must reach their transmission threshold before their host is killed for meat.

But if that’s so, I was asked at both talks, how does that explain the 1918 pandemic, which killed 50-100 million worldwide? The implication is that because no such industry existed in 1918, poultry intensification need not explain the first H5N1 outbreak among humans in 1997 (and the sequelae thereafter). Something else must explain the recurrence.

I hemmed and hawed through an answer, talking through Paul Ewald’s hypothesis about the evolution of virulence in WWI’s trenches. It turns out I was onto something, but nothing I could articulate even to myself until days later:

Pathogens have histories. They have their origins, their diasporic migration, their classical eras, Dark Ages, and Industrial Revolutions. And as human pathogens evolve and spread in a world of our making, these eras often coincide with our own.

Examples abound. As Steven Johnson describes it nicely, for most of its history the cholera bacterium made its living eating plankton in the Ganges delta. Only once humanity became urbanized and, later, interwoven together by 19th century transport and trade was cholera able, like protagonists of the Bildungsroman of the era, to make its way to the world’s cities and find its fortune. A Dickensian trope, the bacterium was able to transform its marginality into roaring success when cities Metropole and provincial, their inhabitants piled atop each other, began drawing drinking water from the same place they dumped their shit.

Other pathogens engaged in similar shifts in epidemiological practice. As the standing hypothesis puts it–Edward Hooper notwithstanding–SIVs of various stripes emerged from the deep tropical forest when the colonial, neo-colonial and neo-liberal enterprises turned subsistence hunting into voracious trade. It appears HIV emerged via a burgeoning market in butchered bush meat before hitting the global circuit.

Influenza, for eons feasting on an archipelago of waterfowl cities across Siberia, and on human populations further south since at least the 1500s, has its own history. My explanation for 1997’s flu, then, need not be turned on itself because of its inability to explain 1918. The latter was in fundamental ways a different virus, one that evolved out of the historical circumstances it found itself, whatever we may discover those to be in the future. But for now, that’s neither my responsibility, nor my refutation: industrial poultry had an important role in the emergence of the latest influenza.

The 1918/1997 comparison is an attempt to impose an algorithm upon contingent history, which resists such easy abstraction. The failure reflects a physics-envy that has crippled evolutionary biology since the Modern Synthesis mathematically reconciled Darwinian selection and Mendelian genetics. Natural selection is on its face an algorithm–propose, dispose, repeat–but the concrete–or rather fleshy–results are dependent on an era-specific mash of previous adaptations, quirky genomic architecture, and passing ecological circumstances.

Evolutionary epidemiologist Karl Marx put it best, viruses “make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given and transmitted from the past.”