Findings from a report on equine influenza by Dr Perkins were included in a recent communique from Queensland chief veterinary officer Dr Ron Glanville.
"There is some evidence," said Dr Perkins, "that horses in their own environment and on pasture may show quite mild signs, whereas horses at Morgan Park - stabled, out of their environment and in a high horse density environment - showed signs and progression of signs that were very consistent with published literature.
"There is a definite risk of under-reporting of disease due to lack of severe signs in horses that may be on pasture and not closely inspected at regular intervals."
Dr Perkin's report also pointed to evidence raising the possibility of airborne spread of the disease of up to 4km.
Analysis of trends by Dr Perkins and the state's Department of Primary Industries indicate the outbreak should largely be under control by Christmas, provided people continue to be vigilant about biosecurity.
Dr Perkins said observations from Morgan Park at Warwick, provide a well-documented record of the rate-of-spread of the disease through a confined population.
Presumed infectious horses arrived at Morgan Park on a single truck from New South Wales on the afternoon of August 24, and 100% of around 250 horses were showing clinical signs by October 1 - a period of less than eight days.
"These observations clearly indicate that in a completely susceptible population with close contact the rate of spread is very rapid."
Dr Perkins said initial spread was associated with movement of horses and subsequent direct contact between infected and susceptible horses.
"Major alternative spread mechanisms that are believed to have contributed to subsequent spread include indirect contact via fomites (vehicles, equipment and people) and airborne transmission."
The majority of the spread has been over short distances.
"There appears to be little evidence to confirm or rule out potential mechanisms for spread and the following list is provided for illustrative purposes only and is likely to include mechanisms that may not be playing any major role in the spread of EI:
"In some cases these reports are supported by evidence of direction of spread of infection being non-random and consistent with prevailing wind direction and by investigation ruling out other obvious explanations for spread.
"At the practical level, airborne spread is one of a large number of potential mechanisms of spread that can contribute to local spread.
"There is understandable interest in distinguishing the role of particular individual spread mechanisms and the factors that may influence these ... [but] these questions are difficult to resolve and would require experimental studies.
In the short term, he said, there was considerable value in attempting to understand the distances and broad factors that might influence local spread at a more general level.
Pointing to the outbreak in the Park Ridge region of Queensland, Dr Perkins said the disease appeared in the week ending September 13 and over the next five weeks a total of 113 infected properties were identified.
Almost all new cases occurred within 4km of an existing infected properties, with the sole exception of a single property 9.22km from the nearest infected property.
"There was a consistent spatial trend of movement of infection from east to west over weeks five to eight of the epidemic, which was consistent with recorded wind direction during the same period.
"While it is accepted that some local spread may in fact be the result of poor biosecurity, it is difficult to separate these particular mechanisms from any other mechanism.
"This very useful study has implicated airborne spread because of the association between direction of spread and wind direction. The results may also be used to indicate at a more general level that all forms of local spread may contribute to spread of infection at distances up to about 4km from existing infected properties.
"In contrast, spread of infection involving distances greater than 4km would seem almost certain to involve some form of breach of biosecurity and may be described as satellite or outlier spread.
"One preliminary analysis indicates that the relative proportion of new infected properties that were more than 10km from an existing infected property has been less than 2%.
"There are a small number of infected properties that have clearly involved some form of biosecurity breach because of the distance between these infected properties and other known infected properties.
"A number of these have been subjected to additional investigation in an attempt to identify mechanisms of spread. In most cases there has been no simple explanation and findings tend to indicate a number of possible mechanisms of spread.
"There are a number of infected properties where investigations have identified relatively simple and rational explanations for spread. Many of these involved human-assisted fomite transmission - for example, individuals with links to horses at multiple locations."
A number, he said, were difficult to explain, with investigations unable to identify plausible pathways that may explain how the property became infected.
Others hypothetical means of spread include indirect contact involving people contacting other people at an off-farm location, the role of produce stores either through people mixing while visiting a store and inadvertently acquiring virus or through movement of produce store personnel (along with vehicles, equipment and feed) while delivering feed to farms, and the role of flies and other indirect fomite mechanisms.