The knowledge-base approach used to identify garter snakes, as discussed in the textbook Wildlife-Habitat Relationships: Concepts and Applications, Second Edition by Morrison, Marcot, and Mannan, can also be used to develop other expert identification models.  Here's another, for identifying conditions and classes of slump-formed ponds as habitat for wildlife.

Three Classes of Ponds

This model, called simply PONDS, is based on the pond classification system presented in Marcot (1990).  In that publication I identified three main classes of slump-formed ponds in the inland mixed-conifer forests of northwestern California:

ephemeral - which dry annually

astatic - which fluctuate seasonally >10% of their maximum depth

stable - which fluctuate seasonally <=10% of their maximum depth

What the Model PONDS Does

The model PONDS helps identify the pond classes based on simple, observable characteristics.  It also provides some simple management tips for maintaining the wildlife habitat values of each pond class. It runs by asking for specific information on vegetation and physical attributes, and provides the best interpretation of the pond class(es) based on your response.

I built the model, as with the GARTER model, in 1st Class using a rule-induction algorithm ID3.  This creates an optimal rule based on a set of examples.  In this case, the examples were the characteristics (physical limnology and vegetation) of a set of 13 slump-formed ponds I studied in northwestern California.

How To Get the Model

Get it here!  This link downloads the file PONDMODL.ZIP.  Unzip this file to any directory and run the model by typing PONDS <Enter>.

Run tip:  You can view all the information on all 3 pond classes in the model by entering "?" (without the quotes) to each question.

Citation

Marcot, B.G.  1990.  Limnology, vegetation, and classification of Coast Range slump-formed ponds.  Northwest Science 64(1):55-63.

Back to the WHR Book Page ...
Back to The Plexus Home Page ...