permaculture

[Toby Hemenway <hemenway@jeffnet.org>]

As I've followed this thread, my leaning has been that the PFAF DB does most
of what is done by the plant DB we have talked about here, except for local
or specific DB's like the one Miekal originally described. But Georg's note,
that

>pc designers might know of different propagation
> strategies, uses (designers are aware of uses other people pay no attention
> to, e.g. fire-retardancy  . . .

raises an excellent point. A few of these uses show up as search fields in
PFAF (biomass, insectary, etc.) but permies, being oriented so strongly
toward function and relationship rather than size, color, or other physical
quality,  do make observations that horticulturists or botanists don't. The
question for me then is whether it's worth constructing a new DB based on
that, or just to add those qualities into an existing DB. It seems simpler,
perhaps, to organize these qualities as prose, as observations specific to
each plant, rather than adding dozens more fields to a DB, or we risk having
vast numbers of blank fields for each species.

But the idea of a guild DB . . . how many times have I been asked, "is there
a list of guilds for me to work from?" And the answer, outside of some
academic literature about the tropics, or listings of native plant
communities, is no. So perhaps it's time.

How would a guild DB be organized? I'm not a DB person, so I can only speak
from a user/permie viewpoint. Guilds can be organized several ways:

-Around a central element, such as a tree.
-Functionally, as a set of connections among insectaries, N-fixers, mulch
plants, etc.
-Structurally: plants that physically fit together--edge plants, understory,
vines, emergent trees, rooting depths, etc. that combine into a unit.
-Analog guilds that mimic natural plant communities.
-By environmental gradients: plants that like the same climate, soil, or
other conditions and thus tend to cluster.
-By natural selection: the end result of winnowing of high-diversity
plantings, survivors of various selection processes (neglect, drought,
disease) that stabilize into a cohesive group (I've generated a few of these
accidentally, and then have duplicated them elsewhere successfully; I like
'em).
And then there are groupings of plants that are just random collections and
not guilds (and how do we know the difference?).

Maybe I'm starting at the wrong point, because these varied definitions of
"guild" make it difficult to determine what are the useful relationships to
list in a DB. Each of the above guild types is ordered by different types of
relationships, thus constructing a Db based on relationships is a challenge.

So maybe it's useful to ask, How would someone searching a DB for a usable
guild find what they are looking for? What would they be looking for--a
guild for a particular location, function, or what?   Or are there so few
guilds--maybe only a couple of dozen?-- that a simple listing would suffice?

Since I'm generating far more questions than answers, I'll leave off for now
and think about this some more. But a key question, it seems to me, is what
would someone use a guild DB for? That would help structure the DB.

Toby
_____________________________________________
For a look at my new book on ecological gardening,
Gaia's Garden: A Guide to Home-Scale Permaculture, visit
http://www.chelseagreen.com/Garden/GaiasGarden.htm