Well, (no pun intended), Chicora Foundation,  (South Carolina) send
me a test of plant phytolith analysis on well fill -- ie. bottom,
middle and top samples.  The middle and top samples were relatively
impoverished, while the bottom sample was relatively rich with these
plant residue silicified cells.  My conclusion from this data was:
the bottom represents the time when the well was in use and receiving
significant amounts of leaf litter over time from which the
phytoliths are derived.  The impoverished nature of the middle and
upper samples indicated to me that the fill was not "botanically
active" -- probably sterile subsoil for the most part dumped in as
fill.  This also suggests that the fill might have been a
case of reverse stratigraphy; that is, biologically active surface
material gets dumped in first followed by deeper, sterile soil from
the spoil pit.  I doubt the well was filled slowly with cultural
trash. Cultural deposits contain substantial plant residues which
should occur throughout a slowly built-up deposit.  Not so in this
case.  In addition, a slow buildup of fill would continue to
receive coincidental leaf litter.  I can provide a citation for
this study upon request.
 
Sampling size aside for the moment, it appears that the well, once
abandoned was filled rapidly.   Sampling strategy NOT aside, three
samples from a well may not be enough to demonstrate this.  Also, one
well does not a pattern make.  Well fill may be a case-by-case
circumstance.  Obviously such a determination requires further
analysis.  Limiting recovery to sherds and bottles is....limited.
There's a host of other analytical approaches available.
 
Irv Rovner
Binary Analytical