LAUSR

May I offer a comment on the recent article in your Journal by Dr E Davison. I am a forester, not a plant pathologist, but for 30 years, from 1964, I was associated with various aspects of jarrah dieback, including: mapping from aerial photos, screening for resistance, ecology, the introduction of quarantine, development of hygiene protocols, monitoring of logging trials, policy and the implementation of dieback protection plans for various National Parks. I have no doubt that site conditions and waterlogging play a substantial role in the disease syndrome observed in the (Eucalyptus marginata) jarrah forest, but there are still many questions that remain. Jarrah is not favoured by wetter sites and is replaced there by bullich (E megacarpa), blackbutt (E patens) and marri (Corymbia calophylla). I have no doubts that waterlogging alone can rapidly kill mature jarrah trees and have observed this myself , especially in forest located below sumps in cleared bauxite pits where the soil was sodden underfoot and water was pouring out from the road cutting. From the earliest descriptions by foresters, jarrah dieback was found predominantly in watergaining sites mostly on slopes and the “head’ of gullies. These sites were considered of lower quality (shorter in height and more open) and often with indicators of some impeded drainage. However the individual stumps of jarrah on these sites were mostly large, 1 to 1.5 m in diameter, indicating that these trees were very old, perhaps 250 years or more. I find it hard to envisage that during this long period of time there were no unusually high rainfall events, including heavy rainfall in summer, such as were observed between the 1940’s and 1960’s and in 1982. Had the jarrahs on these susceptible sites been killed by periodic high-rainfall events, one would expect the newer cohorts to be younger and smaller in size. Also, why did not marri and bullich invade opportunistically 150+ years ago, as has occurred since the 1950’s on dieback-affected sites? Rainfall data from Jarrahdale, one of the larger sawmill towns, are available from 1882 to the present. Mean rainfall is 1177 mm with an upper 90th percentile value of 1486 mm and a 95th percentile of 1604 mm. From 1904 to 1932, these values were exceeded on eight and four occasions respectively and from 1939 to 1964, they were exceeded six and three times. However between 1882 and 1904 and from 1965 to the present, the 90th percentile rainfall was never reached. The data show a cyclical drier to wetter to drier pattern, with the 60 year period from 1904 being wetter than normal. Exceptionally wet months were also recorded: 701 mm in June 1945, 515 mm in July 1946 and 510 mm in July 1917. The earliest reports of small patches of dead trees were filed in the 1920’s. Why then did dieback only become a major concern in the mid to late 1940’s and not during the earlier, wetter period from 1904 to 1932? I am not convinced that logging contributed significantly to raising water tables in the 1940’s to 60’s though it probably did so between 1880 and 1920. Departmental records show that the high rainfall jarrah forest, where most of the dieback areas are located, was very heavily logged before 1920, with minimal control, as the Forests Department was only formed in 1919. The logging was followed by intense bushfires in the accumulated logging debris on the forest floor. The forest is flammable and the dry conditions each summer are conducive to bushfire. The crown deterioration observed after logging may be due to exposure and higher water tables, but bushfire This comment refers to the article available at https://doi.org/10.1007/ s13313-018-0558-8, and an author’s reply to this comment is available at https://doi.org/10.1007/s13313-018-0580-x