Disorderly conduct
Can we predict which species will be affected by bird flu? (12 minute read)
I’m standing on a tiny beach, looking north-east across Auckland’s Waitematā harbour. It’s a warm summer evening, under a blue sky, with a breeze barely ruffling the harbour’s surface. While the water should be inviting, I’m no more inclined to jump into the water than I was in Antarctica, where I saw the penguins I wrote about last week. I know people do swim here, but I also know that whenever there’s significant rainfall, sewage pipes overflow and raw sewage flows into the harbour. These inner harbour beaches are frequently marked as unsafe for swimming, shellfish collecting and fishing1.
Nobody’s told the terns, though. Just off the beach there’s a tiny flock, four birds soaring and circling above the water. Sometimes, one will suddenly plummet head first into the water, emerging a few seconds later with a fish, or without one, if it’s unlucky. The water might be polluted and the surrounding land altered beyond recognition, but I’m delighted that there are still enough fish left in the harbour to attract these elegant birds.
Terns have been on my mind lately, because they’ve been particularly hard hit by bird flu. The terns I was watching last Christmas were white-fronted terns, or tara, which are mainly found around New Zealand but are also seen around Tasmania and Australia’s east coast, so until very recently they wouldn’t have come into contact with bird flu. But they are a close relative of one of the most affected species in the Northern Hemisphere, the common tern. This suggested to me that the white-fronted terns may also be at risk, but this was just my assumption. I decided to see whether there had been any work which would clarify what was likely to happen in New Zealand, whether my assumption was likely to be correct and whether I could apply any of my own risk assessment experience to the issue.
Predicting how a species will behave when it reaches a new area has been part of my work, in one way or the other, for nearly 30 years now. Actually, it might be more accurate to say I’ve been attempting to predict how species will behave, because such predictions are frustratingly inexact. I have learned that it’s usually possible to determine that some scenarios are unlikely. But among the realistic possibilities, it can be difficult to narrow down which are likely.
Many diseases of both animals and plants are shared among related species but rarely, if ever, infect unrelated species. For example, foot and mouth disease, the livestock disease which New Zealand hopes will never reach here, almost exclusively infects cloven-hooved animals like cattle, sheep, goats, deer and pigs2. Another example is myrtle rust, a fungus which infects only plants belonging to the myrtle family. However, within the myrtle family, different strains of the fungus affect different plants to different degrees. Prior to its arrival in New Zealand, we couldn’t say exactly which of our native plants in the myrtle family would be infected and which would suffer the most damage. However, we could be pretty confident that it wasn’t going to infect any plants outside the myrtle family.
When I started looking at bird flu, I wondered whether there were any patterns of this kind in the birds which were being infected. I particularly wanted to see whether the birds which were being killed in large numbers were related to one another. This is relatively simple, because the classifications which scientists use for living things reflect these relationships. Species are grouped together if they share common ancestors.
I started with the highest level grouping of birds, known as the order. There are around 40 orders of birds, and around half of all bird species, well over 6000, belong to a single order commonly known as perching birds. This order includes most of the familiar small birds such as sparrows, blackbirds, robins and swallows, as well as a few larger birds such as crows and magpies. However, in New Zealand we have only 22 native perching birds which still survive, including our robins and tomtits, tūī, riroriro (grey warbler), tīeke (saddleback) and kōkako.
Before I continue: I encountered something of a problem in trying to make this article clear. I realise that I need to pause and explain my approach, so please bear with me.
Most orders of bird, apart from the perching birds, don’t have a clear common name, because these classifications aren’t widely used by non-scientists. The scientific terms, however, can be a real mouthful. Phoenicopteriformes doesn’t roll easily off the tongue, for example. The usual way to refer to the orders is by the common names of some of the birds in that order. This works perfectly for some orders, such as the flamingos, which are the only bird in the order Phoenicopteriformes. But other orders contain a number of well-known bird groups.
For clarity, when I first mention an order, I will name some of the birds in that order, but I will write one of the names in italics. When I next mention the order, I will use only the italicised name. So, when I later refer to the order which contains pigeons and doves, I will refer to it as ‘pigeons’.
Aside from the perching birds, no other order of birds contains more than a few hundred species. Among the orders with a few hundred birds are pigeons; parrots and parakeets; and eagles, hawks and vultures (but not falcons, which belong to another order). Chickens belong to another moderately large order, along with quail, pheasants, turkeys and peacocks. The birds we usually consider seabirds belong to several different orders. Penguins are in an order of their own, another order contains albatrosses and petrels, another gannets and cormorants, and then there’s a single order which includes the birds we commonly call shorebirds, such as plovers, stilts, godwits, knots and snipe, as well as gulls, terns and skuas. Ducks, teals, geese and swans are another order, but these are separate from grebes such as the pūteketeke (the bird made famous during the Bird of the Year competition, thanks to some election interference from John Oliver). Pelicans, herons and ibises are another order, which is separate from flamingos and also separate from cranes and rails (such as the weka).
These orders become important when we start to look at which birds are infected with bird flu. As I’ve mentioned in previous articles, mild forms of the virus are found in birds around the world, including New Zealand. Some are restricted in the species of bird they infect, such as two which are known only from shorebirds. But the pandemic bird flu we are worrying about at the moment is a descendant of a severe form which arose in the early 2000s and has been spreading and evolving ever since. It also belongs to a subgroup of bird flu viruses, first found in domestic geese in 1996, with a known tendency to jump from one species to another. Bird flu viruses in this subgroup didn’t move just between related species, such as geese and ducks. They also moved between species in different orders, such as ducks and chickens.
Because there are different forms of bird flu, estimates of how many bird species can be infected vary. One recent list of birds known to be infected with bird flu names under 400 species, from 26 different orders. This list uses records from 1973-2023, so some records pre-date the bird flu pandemic. It also uses only one source of data and is missing some records, such as the African penguin. However, it was still useful as it came with a spreadsheet of all the data, so I sorted through it until I found only the type associated with the pandemic bird flu. There were around 250 different bird species listed where there was at least one record of the pandemic bird flu type. The two orders which are most important for bird flu, the ducks and the shorebirds, comprised around 40% of the species listed, and the perching birds and eagles each comprised about 10% of the species. There were 20 other orders where at least one species was recorded with bird flu.
The perching birds and eagles on this list are interesting for two different reasons. Although 32 species of perching bird have been recorded with bird flu, this is a tiny proportion of the whole order, less than 1%. The order includes species such as the starling and sparrow, which are found in large numbers in cities, suburbs and agricultural landscapes. One of the first serious bird flu outbreaks in wild birds, in Hong Kong in 2002, included one sparrow. But it’s still not clear how important perching birds are in spreading bird flu, and there don’t appear to be records of any species suffering severe impacts.
In comparison to the perching birds, eagles are over-represented as bird flu hosts, for one obvious reason. Eagles and their relatives are mainly birds of prey, with a few species which are mostly scavengers. If they eat an infected bird, they can be infected and killed by the virus. As the pandemic has spread, more and more eagle species have been affected, and there have been significant numbers killed. There have also been a number of deaths among other predatory birds such as falcons and owls.
The birds of prey are an important indicator of a pattern with bird flu: relatedness doesn’t tell us which species will be infected or killed. This pattern becomes even more apparent when we look at the outbreaks where large numbers of wild birds have been killed. In one of the first outbreaks, in Hong Kong in 2002, the birds which died in the greatest numbers were ducks and geese, which belong to the same order, and flamingos, which have their own order. Another early outbreak, at Qinghai Lake in China in 2005, mainly affected four species: one goose, one cormorant and two gulls. These birds belong to three different orders. The 2017 and 2018 outbreaks in southern Africa mainly killed terns, but also killed more than 1000 gannets and hundreds of African penguins; again three different orders were affected.
The new strain of bird flu which emerged in 2021 caused larger and more frequent outbreaks in wild birds. Sometimes, it affected the same kinds of birds as previous outbreaks, particularly ducks, shorebirds (especially terns) and gannets. Sometimes, though, it killed large numbers of other birds, such as cranes, pelicans and sooty shearwaters. These belong to three different orders, and while there were occasional cases of bird flu reported from birds in these orders in the past, they had not previously suffered severe outbreaks.
The pattern to these outbreaks, if there is one, doesn’t have much to do with relatedness between different bird species. What matters is the opportunity for transmission. Birds of prey become infected when they feed on infected birds. Some outbreaks are associated with areas where large numbers of migratory birds are found together. Other outbreaks are associated with birds which breed in dense colonies, creating ideal conditions for the spread of bird flu.
What’s less clear is whether all these different bird species are able to spread bird flu, as opposed to catching it but not passing it on. Nor is it clear why some bird species seem to escape the outbreaks. In the South American outbreaks of 2022 and 2023, some abundant seabirds had only a few deaths recorded due to bird flu. Were these species not particularly susceptible, and the few dead birds already weakened by something else? Did few birds of those species come into contact with the virus? I’m still wading through the numerous papers on the subject, trying to see whether anyone’s worked that out.
So what does this mean for our ability to predict what might happen in Australia and New Zealand? It’s not good news. A 2024 risk assessment concluded that many species have overlapping habitats, which will allow migratory birds carrying bird flu to pass on the virus to birds which don’t migrate, including those which live mainly on land, rather than on the coast. But there’s no clear indication of which species will be affected. This is not an orderly virus, behaving in a predictable way.
But there’s something the risk assessment doesn’t mention and which I haven’t seen discussed elsewhere. One factor may make some of New Zealand’s rarest birds even more vulnerable. Many are concentrated on small islands which also have many seabirds. This may increase the chances of bird flu being spread to our rare birds.
One potentially vulnerable bird is the kākāpō. Almost all live on three predator-free islands, and on one of these, Whenua Hou, they come into contact with seabirds which may carry bird flu. If you watched the live video of a kākāpō on her nest last summer, you may remember when a petrel entered the burrow and was chased off by the kākāpō. While there are very few records of parrots infected with bird flu, it’s difficult to predict what will happen when parrots come into close contact with seabirds. Many of the kākāpō on Whenua Hou have been vaccinated against bird flu for this reason.
Some of our other rare birds have been vaccinated as well, but this is really only worthwhile where birds can be given two doses of the vaccine. In reality, this means that it’s mostly birds which are being bred in captivity or are under intensive management, such as kakī (black stilt) and tuturuatu (shore plover). Other birds will have to fend for themselves.
Some of our rare birds are extremely vulnerable. There are fewer than 100 tara iti (fairy terns), fewer than 200 tchaik (Chatham Island tāiko) and around 200 Whenua Hou diving petrels. Two of our three surviving snipe species have fewer than 2000 birds and there are no more than 3000 hoiho (yellow eyed penguins). All of these birds belong to orders where some species have experienced large outbreaks and live in habitats where they may encounter infected birds from other species. Maybe they will be lucky and maybe they won’t. I’m not sure that we can do more than hope.
While I’ve been looking at birds, I’m also aware of an elephant in the room, or rather an elephant seal. Bird flu is not a disease which is confined to birds. As I mentioned in last week’s article, it has spread to marine mammals, including elephant seals. These are far from the only mammals affected. Next week I will be away in Taranaki, so I’ll write an article related to my trip, but I will return to bird flu and its spread to mammals, including humans, in a couple of weeks.
Although I’m hopeful that the newly-opened Central Interceptor sewage pipe will stop most of the overflows and make the beaches much cleaner.
There are a few exceptions, which I’ve previously written about in the following article: Desperate measures
You are wonderful Melanie. I enjoy your scientific articles in normal language on topics we didn’t even know we needed to know about. Thank you Thankyou Thankyou for your enquiring brain and research interests . We are so lucky to have you in our midst.
Thank you ... very informative ... I would have thought starlings were susceptible ...