It seems unlikely that two separate and lethal infections devastated chickens and humans almost simultaneously; therefore, a single responsible agent is more probable. Newcastle disease can kill chickens rapidly but does not cause disease in humans.5 Psittacosis can infect domestic poultry6 and then humans; however, the rapidity and lethality of the disease in 1539 in both chickens and humans is not typical of psittacosis. Anthrax is a devastating zoonosis most typically associated with herbivores. Infection of birds with anthrax is usually asymptomatic and is detected by spores shed in faeces.7 Chickens, especially, seem resistant to the disease.8

Could avian influenza have caused the 1539 outbreak? In modern times, the H5N1 strain of avian influenza affects birds and humans, with human-to-human transmission considered to be very rare. In the current climate, one would have to consider avian influenza as a differential diagnosis for an infection targeting both domestic poultry and humans. Few zoonoses other than H5N1 influenza kill both.

Features of the 1539 outbreak have some similarity to avian influenza of the 21st century. The infection appeared to be an emerging infectious disease, not seen before by the indigenous people or the Portuguese. Also, throughout the islands, the illness seemed to affect poultry before affecting humans, supporting the idea of poultry-to-human transmission.

However, there are strong arguments against avian influenza being responsible for the 1539 outbreak. First, the large number of human casualties is in direct contrast to the situation in the 21st century. In today’s H5N1 outbreaks, human death rates are low, despite the exposure of large populations to chickens.9 This may be attributable to the H5N1 virus binding only to galactose receptors in the lower respiratory tract, unlike human influenza viruses, which can bind to galactose receptors in the upper respiratory tract.10 Therefore, if avian influenza was the cause of the 16th century epidemic, the degree of human devastation described is surprising, even if Galvão exaggerated the situation. An avian influenza virus capable of human-to-human transmission might explain so many human deaths, but a mutation rapid enough to allow this is questionable. Also, H5N1 influenza in humans has a distinct clinical syndrome of focal respiratory symptoms often accompanied by diarrhoea,9 yet Galvão did not describe focal features of the 1539 outbreak. One could attribute this to a failure to record the symptoms for one reason or another, but this seems not to have been the case. The treatise elsewhere is filled with details and elaborate descriptions by Galvão, including careful depictions of the clinical features of other epidemics and illnesses.1 Therefore, it is more likely that the 1539 illness truly lacked obvious focal clinical features, possibly because of its rapid and severe course, but this would certainly make avian influenza far less likely as a diagnosis.

Although the specific organism cannot be pinned down, the way in which the illness appeared to spread through this group of islands but not outside it also supports an infectious aetiology if the trading practices of the region are studied.

Strong monsoonal patterns strictly regulated the periods of external contact. At the time of year during which the outbreak occurred, there would have been trading only among the islands but not outside them. Generally speaking, sailing vessels from outside the region could arrive in the archipelago only twice each year during rather narrow windows of time. Most ships approached via the Java and Banda Seas, arriving in the southern part of the archipelago about mid February. Those approaching from the north would arrive around the end of October.

Regardless of when or by which route they had arrived, all sailing vessels had to depart from the Maluku archipelago via the southern route with the arrival of the south-east monsoon, about the middle of May.11 This monsoon is the “south winds” identified by Galvão as having marked the advent of the illness in Bacan. Consequently, the outbreak would have begun just as external trading vessels were departing from the island of Ambon about 350 km to the south, and it would be more than 5 months before the first outside ship could return to Maluku, presumably after the infection had run its course.

At the same time, traffic among the islands in the Maluku region would have continued without interruption. This was possible because korakoras and other native craft could be rowed in shallow, reef-infested seas regardless of the prevailing winds. Constant inter-island trade was necessary because so many of the spice-producing islands depended on imported foodstuffs. Bacan, where this illness was first observed, produced no food of its own and therefore relied entirely on supplies brought in from other islands, and Ternate was in almost the same situation.4

These inter-island food shipments would have included chickens.12 Chickens were imported from a handful of fertile areas within the Maluku region rather than raised locally;13-15 therefore, local inter-island trade in chickens would have been sufficient for an infection carried by fowl to “spread through all the islands”.1