204 
SYMPOSIUM PHOTO INTERPRETATION, DELFT 1962 
in this situation, since it has been applied in so many parts of the world, 
is to stratify the forest first on the basis of aerial photographs into the 
different forest types, and then carry out field sampling of the commercially 
valuable forest types to determine their composition by species and size classes, 
producing as an end product a series of forest maps showing the location 
and extent of the valuable forests, with a report on their timber content. 
Yet this cannot be undertaken in many parts of the world with modern 
aerial survey techniques, either because weather conditions prevent photo 
graphy, or because the forest types are heterogeneous and cannot readily be 
distinguished on aerial photographs which are available. The tragedy of this 
situation is that it is the really rich tropical rain forests which have so far 
defied inventory on the lines developed so successfully in temperate climates 
and in the drier parts of the tropics. 
I have said enough about problems in the dense tropical forest to indicate 
that we are far from a solution. But there are two rays of hope. One is that it 
is possible to delineate on aerial photographs, site types based on ecological 
and geomorphological characteristics. As you will see from this the bulk of 
the commercially exploitable tree species are located in one particular zone. 
This zone can often be determined on aerial photographs. It is therefore 
possible to at least obtain a net forest area which will contain the bulk of the 
commercially valuable tree species. Using this as the basis for stratification 
then field sampling can be concentrated in this zone and the results obtained 
applied to it alone. Where the other two zones, i.e. those of impeded drainage 
and of too shallow soils form a significant portion of the whole, then an appre 
ciable saving in both the cost and time of forest inventory work may be obtained. 
The standard procedure for this type of work is to get down to net forest 
area of commercially valuable forest, apply to it a percentage of field sampling 
on a randomized basis, so that it may be subjected to a statistical analysis and 
the degree of precision involved in the estimate calculated, and follow this up 
by preparing a tabular summary giving the areas and volumes of the timber 
involved. This must be sufficiently comprehensive to give a breakdown by 
tree species and by size classes. Size classes - because this is of critical impor 
tance from the point of view of determining the most suitable forest industry 
to be established. For example large sizes of logs are more suitable for veneer 
production or sawmilling, while greater quantities of smaller size classes would 
indicate suitability for pulp and paper manufacture or board products. Size 
determines what you do; amount determines the scale of operation. 
This assessment of raw material supplies should always include not merely 
a statement of the existing situation but of estimated production five years 
ahead, ten years ahead, or even longer. Incremental studies carried out at 
the time of the forest survey will give an accurate indication of the rate of 
growth of the forest stands, and therefore an indication of the sustained annual 
yield to be expected from the forest resources under investigation in the area. 
This also is difficult in the tropics where most tree species do not have annual 
rings and the method applied in temperate forests of ring counts from stem