442 
obvious areas of infection. However the early stages of ui 
the disease, where some of the understorey only are dying, s< 
cannot be detected on these photographs. tl 
Pi 
The same method of interpretation was first c< 
attempted in the southern forest in 1966 but with little 
success. The main reason for this appeared to be that 
fewer of the understorey species were susceptible, coupled 
with a higher proportion of marri in the overstorey. The 
death of the susceptible component of the stand therefore Ac 
had much less effect on the surface reflectance. Further, 
although the disease was spreading rapidly at that time, 
very little of it was at an advanced stage. It is worth fc 
noting here that while decadent jarrah crowns could be me 
detected on these photos, this factor alone was of little (2 
value as an indicator of .dieback in many areas in the south. he 
Most crowns in the area were secondary crowns with a number tl 
of dead leaders as a result of past wildfires. Unless the cc 
disease was well advanced, symptoms in the understorey were me 
the only real guide even for ground reconnaissance, 01 
le 
Attempts to detect diseased areas with colour and a] 
colour infra-red photos at conventional scales were also ui 
unsuccessful• 
It was concluded that a system of dieback mapping gi 
based on the detection of disease symptoms in individuals wj 
of the understorey offered greatest promise of success. te 
Limited trials by the author in 1968 and 1969 with large- Bi 
scale 70 mm photography were encouraging and established the f] 
basis for the project described in this paper. The early ac 
trials indicated that: tl 
1S 
(i) Colour was superior to colour infra-red for A 
detecting dying understorey. 
(ii) A scale of about 1: 3,000 was necessary to detect o\ 
and identify dying banksia, zamia and blackboy. To differ- Ai 
entiate between healthy and dying blackboy it is necessary ue 
to differentiate between the colour of the central 15-20cm ce 
of blackboy crown and its outer fringe. cl 
(iii) Transparencies were superior to prints because of 
the improved resolution which permitted greater effective ar 
magnification of the image. Live zamias and sometimes ra 
blackboys are difficult to see on prints of this scale but 
are more readily seen on transparencies. (The presence of 
live susceptible species is an important diagnostic property) 1: 
1: 
(iv) Photographs taken under cloud were superior to re 
those taken under a clear sky because the flat illumination na 
eliminated shadows which obscured a high proportion of the op 
understorey and also eliminated bright reflections from. 
shiny flat leaves which hampers colour differentiation. Ir 
The cost of complete coverage with large-scale 
photographs was prohibitive so that a two-stage program 2x