If we are to believe the news stories making the rounds of the media, the export banana is a metaphorical heartbeat away from extinction. It shouldn’t come as a surprise then that a blogger saw a parallel between the Heartbleed security bug and Cavendish bananas.
The blogger, Mark Gibbs, compared the ubiquitous open source security protocol OpenSSL to monocultures of Cavendish bananas. “When a biological monoculture becomes dominant it's guaranteed that a pest or disease that exceeds some level of virulence can threaten the entire biome, which is exactly what's happening with bananas because the bananas you buy at the supermarket are overwhelmingly of one variety: Cavendish.” The two examples he uses are black Sigatoka (black leaf streak in our book) and Panama disease (Fusarium wilt), which he says used to be less virulent and have re-emerged in more virulent forms. “Currently there's an outbreak of a new strain of a disease called Black Sigatoka, which destroys Cavendish bananas and which originally appeared in a less virulent form in 1970's. This has re-emerged along with another contagion, Panama Disease. These two contagions are successful simply because there are some many identical hosts so close together.”
In the first example, the less virulent form of black leaf streak Gibbs is referring to is actually a different disease, Sigatoka leaf spot caused by Mycosphaerella musicola. That species is indeed less virulent than Mycosphaerella fijiensis, the fungus responsible for black leaf streak, but it is still a problem in places where it has not been displaced by M. fijiensis. My admittedly nitpicking point is that the problem commercial growers of Cavendish cultivars are wrestling with is not an increase in the virulence of the black leaf streak fungus, but an increase in its resistance to the fungicides used to control it – a classic example of a ‘pesticide treadmill’.
Fusarium wilt is different. Cavendish cultivars are resistant to race 1 strains and their cultivation in fields infested with these strains could theoretically select for more aggressive strains. But it doesn’t explain the evolution of tropical race 4 (TR4), the causal agent of the more virulent form of Fusarium wilt Gibbs is alluding to. TR4 came to the attention of scientists when it turned up in Indonesia and Malaysia in fields that had no history of growing Cavendish bananas. If monocultures had been responsible, TR4 would have evolved in Central America, where Cavendish cultivars have been grown in infested soils for more than 60 years.
Gibbs’ takeaway message to website developers is that they expose themselves to greater risk by using the most popular product on the market rather than use a diversity of less well-known products. I asked a colleague who uses open source tools in building websites what lesson the community was drawing from the Heartbleed episode. His answer was increased vigilance (he pointed out that one of the problems with diversifying coding solutions is that it would make it more difficult for the community of programmers to detect flaws). Coincidently, the FAO said something similar in a recent press release when it called on banana-producing countries to “step up monitoring, reporting and prevention” to contain TR4.
So there is a parallel between the two situations after all. It's just not the one favoured by Gibbs and many observers of the banana's predicament. To be fair to Gibbs, scientists didn't expect the resistance of Cavendish cultivars to the race 1 strains that took down Gros Michel to be so long-lasting. Many, like Gibbs, thought the fungus would overcome it. The fact that the fungus has no known sexual form probably explains why it hasn't cracked the Cavendish's defences.
But what about growing a diversity of banana cultivars? Surely it can do no harm, you might say. Well, given that the majority of banana types tested so far are more or less susceptible to TR4 (including the dessert types that have commercial potential), it's not clear why Cavendish growers would switch to a less commercial type or add resistant types to protect the susceptible Cavendish. In the absence of the fungus, there is no incentive to grow cultivars that have a lower market value, even if it is resistant. And once the fungus is present in the soil, there is no evidence that the resistant types would protect the susceptible ones from getting infected, let alone be economically viable.
Solutions to the unique problem posed by TR4 are more likely to work if they are informed by the biology of both the host and the pathogen than by sweeping generalisations.