J.Konstapel,Leiden, 1-11-2026.

Aanleiding FD : Ook in ‘jubileumjaar’ voor vogelgriep worstelen pluimveehouders nog steeds

Het artikel onderzoekt waarom de meeste influenza-infecties vanzelf eindigen, terwijl sommige (zoals ernstige vogelgriep) dat niet doen, en betoogt dat het antwoord ligt in de eigenschappen van de gastheer-viruscombinatie, niet in het virus of de gastheer alleen.

De centrale hypothese is dat herstel afhangt van een race tussen twee snelheden: hoe snel de infectie een onomkeerbare toestand bereikt versus hoe lang het duurt voordat een adequate immuunrespons zich heeft ontwikkeld.

Volgens deze theorie ontstaat de niet-stoppende, ernstige vorm van vogelgriep vooral door specifieke houderijomstandigheden (zoals hoge dichtheid en lange transmissieketens in de pluimveehouderij), niet doordat het virus uit de natuur komt.

Huidige maatregelen zoals ruimen en ophokken pakken deze onderliggende oorzaak niet aan, waardoor uitbraken terug blijven keren; vaccinatie kan helpen, maar heeft als risico dat het de verspreiding onopgemerkt laat doorgaan.

1. The puzzle: most flu ends on its own — and nobody knows why

An ordinary bout of flu lasts about a week and then goes away — even though the person or animal that caught it had no immunity to that particular virus beforehand. Nothing was “stored” in advance. Whatever ends the episode is already present at the very first encounter, in a wild duck just as much as in a human being.

This simple observation pulls the rug out from under the usual way of framing the bird flu problem, which asks: how much immunity is present, and how can we raise it? But vaccination does not happen in nature, and the natural case is precisely the one that resolves most reliably. An account that explains recovery by stored immunity explains the uncommon case and leaves the common one unexplained.

So the essay asks a narrow, basic question: what ends an infection when nothing was stored — and under what conditions does that fail?

The failure is not a small matter. In the Netherlands, three outbreaks in three days in late October 2026 triggered a nationwide order to keep birds indoors; the first, at a laying farm in Lutjegast on 26 October, involved about 45,000 birds. Worldwide, the H5N1 clade 2.3.4.4b now circulates in wild birds, poultry, and a growing list of mammals, including 1,177 infected American dairy herds across 20 states by late September 2026. The outbreaks that do not end are the whole of the problem.

2. The lens: “closure” belongs to the pair, not to the virus or the host alone

The essay reads the problem through the Vacuum.Net theory, used here only as a way of structuring thought — not as a physical claim. Its key idea is that “closure” (infection ending and the system returning to normal) is a property of a whole path, not of a single point on it. You can measure a point as precisely as you like; that tells you nothing about whether the circuit completes.

Applied to infection, this has a sharp consequence: whatever ends an episode is a property of the host–virus pair. No property of the virus alone, and no property of the host alone, can carry the explanation.

This is also why the essay refuses the usual vocabulary of viruses that “evade,” “hijack,” or “outwit” and immune systems that “fight” or “defend.” A path has no intentions, and an account written in intentions cannot be checked.

Closure is defined operationally: disturb the system at time zero and watch two compartments — a transient one (things that rise on contact and are expected to subside) and a registering one (things that stay changed afterwards). The path has closed when the transient returns, within a characteristic time, to roughly its prior values, while the registering compartment holds a stable trace. It has failed when the transient does not return. Two quantities follow: how much of the disturbance was recovered, and how fast. Both require at least two measurements over time — neither can be read from a single snapshot.

3. The evidence

Five bodies of evidence carry the argument:

• Mallards (wild ducks). In experimentally infected mallards, mild (low-pathogenicity) flu virus replicates only in the gut’s surface lining, concentrated at the tips of the intestinal villi. Virus was shed in droppings from day one. The ducks showed no symptoms and no tissue damage at all (Vet Res 2018).
• Ducks versus chickens. A gene-activity comparison of the two species infected with H5N1 (severe) and H5N2 (mild) found their early responses share almost nothing. Chickens genetically lack the virus-sensor RIG-I (and its partner RNF135) and mounted almost no IFITM antiviral response; ducks raised IFITM1/2/3 strongly. Crucially, within chickens: against H5N1 their response rose on day one and faded — and they died; against H5N2 their response built over three days and held — and they survived (BMC Genomics 2015).
• The official definition. The WOAH Terrestrial Animal Health Code states that mild H5/H7 viruses “have the potential to mutate into” highly pathogenic ones, and treats a sudden virulence jump in poultry as the notifiable event. In other words, the severe form is officially defined as something that appears in poultry.
• Wild bird deaths. In the 2020–2023 panzootic, colonial breeders were the main mass-mortality group. Northern gannets were among the worst hit: about 75% of North Atlantic colonies affected in summer 2022, a 70% colony-size reduction in Scotland, over 58% adult mortality in France, 11.5% population loss in eastern Canada (npj Biodiversity 2025).
• The current situation. Clade 2.3.4.4b dominates globally: 624 severe-flu events in 25 countries in one reporting month, plus infections in goats, swine, sheep, seals, dolphins, foxes and mustelids (FAO). Human cases since 2003 stand at 1,003 with 479 deaths; WHO judges sustained human-to-human transmission currently unlikely (WHO, 2 Oct 2026).

4. A first explanation — and its honest failure

The first explanation built from this material was elegant: an infection confined to a surface that faces outward and discharges outward (like the duck’s gut lining, shedding virus into droppings; or the human airway) ends by construction — the material leaves, the surface renews itself, the path has nowhere to run. Severe bird flu instead runs through blood-vessel lining and internal organs, which face inward with no exit — so the path can only continue, and it ends with the host. It even covered tuberculosis: a granuloma is an opening with no outlet, and it stays unresolved for decades. One rule: a path closes when its discharge has somewhere to go.

But this account is false. The duck–chicken data shows that ducks respond strongly to H5N1 in both lung and gut — meaning the infection in the duck is not confined to an outward-discharging surface, yet the duck survives. A host that resolves an unconfined infection refutes the claim that confinement is what resolves it.

The essay reports this failure openly rather than quietly replacing it, for two reasons: the failed account was built from individually correct facts, and the error was in which property was picked as decisive — an instructive lesson. And a theory whose failures are never reported cannot be judged for how often it fails.

5. What the evidence supports instead: a race between two clocks

The same dataset shows why a purely component-based explanation also fails: the chicken lacks RIG-I in every infection, yet it survives H5N2 and dies of H5N1. Constant gene content cannot explain a variable outcome. The difference must sit in the pair.

What differed between the two chicken cases was not whether a response occurred but whether it carried through: against H5N1 it flared briefly and collapsed; against H5N2 it developed and held. In the essay’s terms, one transient was displaced and did not return; the other ran its course.

This suggests the decisive property is a ratio of two rates:

1. how fast the infection reaches a state from which no response can retrieve it, versus
2. how long a response takes to develop.

Host components (like RIG-I) matter because they shorten the second clock; the severe virus matters because it shortens the first. Neither is decisive alone — their ratio is.

All four cases in the data line up:

Pair	Infection speed	Response speed	Outcome
Duck, mild virus	slow	barely needed	closes
Duck, severe virus	fast	fast (RIG-I, IFITM present)	closes
Chicken, mild virus	slow	~3 days	closes
Chicken, severe virus	fast	~3 days	fails

The reading also absorbs what would otherwise be anomalies: ducks are not universally spared by the current clade — waterfowl deaths have been recorded since 2021. On a ratio account, that is expected wherever dose, route, age or co-infection shifts either clock enough to flip the inequality. No special pleading needed.

Status: formulated and consistent with every case in the material — but not yet verified. Confirming it requires the experimental literature on replication speed versus early innate responses in chickens, which could not be retrieved for this work.

6. Where the non-closing form comes from

A second question hides behind the first: granted some pairs fail, where does the virus that cannot be closed come from? The governing instrument answers plainly: the severe form is defined as something that emerges in poultry. It is not what arrives from the wild — it is what a particular configuration makes of what arrives.

Three measurable features of that configuration:

1. Passage length — how many successive transmissions occur within a population before the chain ends. Long passage gives the virus many chances to change.
2. Route diversity — a commercial flock is one genome in tens of thousands of copies: one route, repeated.
3. Carry — whether a trace persists in the population. A broiler slaughtered at six weeks carries nothing forward; a laying flock can.

Age synchrony (a whole flock at the same point in its course) compounds all three.

This reading is testable in the wild — and effectively already tested: if configuration, not taxonomy, is decisive, wild populations should show the “stall pattern” precisely where their structure resembles a stall. Colonial breeders — nature’s version of a densely stocked barn — are exactly the main mass-mortality group, and coloniality is linked to persistent circulation. Dispersed waterfowl do not show that pattern. The same logic covers mammals without any extension: dairy herds combine density with a shared milking process; mink farms combine density, uniformity and naivety — and are repeatedly where mammalian-adaptation signals appear.

7. What current measures do — and don’t do

The Dutch package of 28 October 2026 (housing order, screening, exhibition ban, visitor restrictions, culling, with compulsory laying-hen vaccination due before Parliament by year’s end) is read through this lens:

• Culling touches none of the three quantities. It removes the birds but keeps the configuration — the “stall” remains, with its carry wiped to zero. Recurrence is then not a risk but a consequence; the annual Dutch outbreak cycle is what that consequence looks like.
• Housing orders act on the rate of introduction, and only while in force. They delay the next episode rather than reduce its likelihood.
• Vaccination is more interesting than a “memory-based” rationale suggests. Since closure at a first encounter does not depend on a stored offset, vaccination’s real effect is to shorten the time to a developed response — a genuine effect on the rate that matters. But it is double-edged: fewer clinical signs in a flock where transmission continues means the configuration keeps operating while becoming harder to observe.
• Passage length is the only quantity that acts on the cause rather than the episode — and nothing currently in force addresses it. Everything that shortens passage (smaller flocks, faster detection, separating waterfowl from poultry, fewer birds kept at passage-supporting density) would attack where the non-closing form is produced.

8. What would prove this wrong

Three tests, each capable of ending the reading:

1. The timing measurement. Apply a standardised perturbation and record the transient at two or more time points, in both species against both virus forms. If the failing pair recovers on the same timescale as the surviving ones, the ratio account is wrong. No new instruments needed.
2. The sequence record. If high-pathogenicity forms appear first where passage is sustained, the earliest detection of each such episode should be in poultry, not in dispersed wild birds. An episode that begins cleanly in a dispersed wild population would refute the configuration account. Testable against existing archives; not performed here.
3. The population-structure prediction. Within wild birds, mortality and persistent circulation should track colony structure, not taxonomy. Survived so far; every future season re-runs it.

An earlier, more specific claim (an inequality between re-exposure interval and a return time in dense flocks) is withdrawn: the return time has never been measured, so the inequality has no content, and it is superseded by the ratio formulation.

9. Honest limits

The essay gives a direction, not values: none of the quantities named (passage length, route diversity, carry, the two rates) is actually measured here. The two experimental papers that would confirm or break the ratio account could not be retrieved and are named as not read. The panzootic review was available only in part. The mink-farm detection in Utah is reported via trade press, and circulating claims about a specific mutation there could not be confirmed and are excluded entirely. Finally, the economic question is untouched — but density, uniformity and fast turnover are also the economic basis of the sector, and an account that names them without a plan is a diagnosis, not a plan.

---

Annotated reference list

Each source is listed with what it contributes and how much weight it can bear.

1. Alternate routes of influenza A virus infection in mallard (Anas platyrhynchos), Veterinary Research, 2018.
   Experimental infection of mallards; replication localised to gut surface epithelium at villus tips; cloacal shedding at days one and two; no clinical signs, no lesions. Carries the mallard evidence and the premise of the surface-confinement hypothesis. Directly observed primary experimental work — but its tissue analysis did not cover the respiratory tract, so it establishes where replication was found, not where it was absent.
2. A comparative analysis of host responses to avian influenza infection in ducks and chickens highlights a role for the interferon-induced transmembrane proteins in viral resistance, BMC Genomics, 2015.
   Transcriptome comparison of two species against H5N1 and H5N2; RIG-I/RNF135 absent from the chicken genome; strong IFITM1/2/3 response in ducks; the chicken’s H5N1 response flared and faded (death), its H5N2 response developed and held (survival). The single most load-bearing source: it both falsifies the surface-geometry account and supplies the within-species contrast that drives the ratio account. The authors themselves note the IFITM role is inferred from transcript data, not tested in vivo — a caveat carried through into the essay.
3. Infection with avian influenza viruses, WOAH Terrestrial Animal Health Code, chapter 10.4.
   The definitional position: low-pathogenicity H5/H7 can become high-pathogenicity; a sudden virulence increase in poultry is notifiable as an emerging disease. Carries the “where the severe form is produced” argument. A regulatory instrument, not a research finding: strong evidence of consensus, not independent evidence of mechanism.
4. The role of wild birds in the global highly pathogenic avian influenza H5 panzootic, 2020–2023, npj Biodiversity, 2025.
   Colonial breeders as the main mass-mortality group; regional gannet figures; coloniality linked to both mortality and persistent circulation. Carries the wild-population test of the configuration account. Read only as a truncated excerpt — figures reported as given; the full argument is not represented here.
5. Global AIV with Zoonotic Potential, FAO, update of 24 September 2026.
   Circulating clades; 624 high-pathogenicity events in 25 countries in one month; 1,177 US dairy herds in 20 states; mammalian host list. Current-status reporting. Read only in part (about two-thirds).
6. Avian Influenza Weekly Update, WHO, 2 October 2026.
   1,003 human cases and 479 deaths across 25 countries since 2003; sustained human-to-human transmission currently judged unlikely. Human figures. The implied fatality rate reflects surveillance weighted toward severe cases and should not be read as a per-infection risk.
7. Landelijke ophok- en afschermplicht ingesteld na vogelgriepuitbraken and Vogelgriep vastgesteld in Lutjegast, Rijksoverheid (Dutch government), 26 and 28 October 2026.
   The three October outbreaks, the 45,000-bird laying farm, the composition of the national measures, and the compulsory-vaccination plan due at year’s end. Carries the policy analysis. Primary announcements from the responsible ministry.
8. Association of RIG-I with innate immunity of ducks to influenza, PNAS, 2010.
   The original identification of RIG-I in ducks in relation to influenza. Cited for provenance only; read at abstract level.
9. Differential innate responses of chickens and ducks to low-pathogenic avian influenza, Avian Pathology, 2012.
   Companion comparative work on the two species against the mild form. Corroborating context; not read in full.

Named but not read — the two gaps that matter. Two papers would decide the ratio account of section 5 and could not be retrieved; they are listed so the gap is visible rather than hidden: Association of Increased Pathogenicity of Asian H5N1 HPAI Viruses in Chickens with Highly Efficient Viral Replication Accompanied by Early Destruction of Innate Immune Responses (Journal of Virology, 2009) — by its title, exactly the relation between replication speed and response collapse; and The pathogenesis of a 2022 North American highly pathogenic clade 2.3.4.4b H5N1 avian influenza virus in mallards (Avian Pathology, 2023) — bearing on how far ducks remain a “closing pair” under the current clade.

Earlier work in the author’s programme, referenced for the closure definition and the vocabulary: Protection Without a Component (operational closure definition, October 2026); Closure, Carry and the Ladder; Where Illness Comes From (8th edition); Curves in the Vacuum.Net; and the Vacuum.Net Theory Foundational Paper (9th edition).

DOI: 10.13140/RG.2.2.13874.77766