Seasonal rainfall and forest dieback epidemics in Gippsland.  K. C. Tregonning and P. C. Fagg.  December 1982.  29 pp. (unpubl.)

SUMMARY

Dieback epidemics, in which large numbers of trees die with drought-like symptoms, have been recorded (up to 1982) in native eucalypt forests of Gippsland, south- east Victoria, on four occasions - in 1953, 1956, 1967 and 1971. The root-rot fungus, Phytophthora cinnamomi Rands, has been implicated in these occurrences, though the conditions that trigger large-scale dieback remain to be elucidated. Unusual climatic conditions appeared to be at least partly responsible, because P. cinnamomi thrives under a restricted range of soil conditions (wet, warm soil being optimum for its propagation) and because of the drought-stress-like symptoms; both these factors are related to rainfall and prevailing temperature.

This study found a mechanism that successfully explains all observed dieback epidemics in Gippsland. The mechanism was derived using a new and novel methodology based on numerical and graphical analysis of observational data, the analysis being carried out without any assumptions in order to yield a robust model for the phenomenon. This methodology should be useful for investigating other complex and intermittently recurrent natural phenomena for which definitive experimentation is not feasible.

To investigate the relationship between climatic conditions and epidemic dieback, an exhaustive analysis of rainfall patterns was carried out, utilising data for that part of the year (November to May inclusive) during which soil temperatures are usually high enough to allow significant P. cinnamomi activity or in which drought-stressing could take place. The study investigated sites representative of East and South Gippsland (Orbost and Woodside respectively) and was based on data extending from the 1880s to 1980, although the post-1945 era was concentrated upon.

Graphical examination of post-1945 rainfall data showed that all dieback epidemics had occurred during years in which a three or four-month period of high rainfall in the warmest months, November to February, had been immediately followed by a three or four-month period of low rainfall (in the months February to May). It was calculated that the likelihood of such a coincidence was less than one in a million (P < 10^-6).

Accurate limits applying to both sites were placed on those rainfall levels conducive to epidemic dieback: greater than 92 mm month^-1 for the wet period and less than 58 mm month^-1 for the dry. For severe epidemic dieback, these levels were correspondingly higher: greater than 108 mm month¹ for the wet and less than 44 mm month^-1 for the dry.

The rainfall data was also tested for autocorrelation (serial correlation) and for any long-term trends; this allowed expected recurrence periods for dieback to be calculated, based on expected probabilities of particular wet-dry occurrences. Thus, for East Gippsland, it was calculated that epidemic dieback could be expected, on average, once every 15 ± 4 years (i.e. possibly as often as every 11 years), and for South Gippsland, once every 46 ± 18 years. Severe epidemic dieback alone was predicted to occur less frequently: 42 ± 16 years for East Gippsland and 82 ± 14 years for South Gippsland.

Based on the conditions found for dieback in the post-1945 years, examination of the pre-1945 years showed that epidemic dieback should have occurred in 1943. This was not observed, and as it is unlikely that such an event would have remained unrecorded, it appears that certain environmental conditions necessary for expression of dieback on a large scale were not met prior to the late 1940s or early 1950s.

Also published:

Tregonning, K.C. and Fagg, P.C. (1984)  Seasonal rainfall and Eucalyptus dieback epidemics associated with Phytophthora cinnamomi in Gippsland, Victoria.  Aust. For. Res. 14 (3): 219-34.