Permaculture Database

[Paul Cereghino <paul.cereghino@comcast.net>]

> or something of that sort, and then fill the 'attribute' table with
> 'nitrogen' or 'full sun' or what have you. (I can see huge problems with
> my example, like the fact that there's no way to specify the exact
> amount of water something needs, but I'm just using it as an example of
> simplicity and abstraction being able to create very complex and
> flexible interactions.)
>  
How about PRECISE but FLEXIBLE

Because we cannot define exactly how much water something needs we 
divide all things into groups based on how they act in relation to 
water.  Taxonomies again

water relations - (inundation tolerance, drought tolerance, yield impact)

saturation tolerance (intolerant of saturation, tolerates dormant 
saturation, tolerates short growing season saturation, tolerates 
permanent saturation) (might include a note on adaptation type -- 
aerenchyma etc..)
drought tolerance (intolerant of drought, tolerates some drought, 
tolerates severe drought)
yield impact (yield minimally affected by drought, yield strongly 
affected by drought, yield prevented by drought).

Drought tolerance and inundation tolerance are two different sets of 
adaptations, and so should not be considered in the same taxonomy.

If these attributes become an endless set of duplicative and overlapping 
classes THEN we are in real trouble.

The trick is normalizing the data so that you don't have a muddy 
taxonomy, and every object fits into (or at minimum straddles) a class 
boundary.  Different people from different places are going to have 
different labels, but trends will occur, and the data can be analyzed 
(80% of people from temperate conifer forest that were not in bottom 
land said their apple tree yield was strongly affected by drought, while 
bottom land dwellers said their trees were minimall affected by 
drought... but not so for those poor drought stricken aussies :)