Forest and wildlife habitat analysis using remote sensing and geographic information systems

Forest and wildlife habitat analyses were conductedat the H.J. Andreigs Experimental Forest in the CentralCascade Mountains of Oregon using remotely sensed data anda geographic information system (GIS). Landsat ThematicMapper(TM) data were used to determine forest successionalstages, and to analyze the structure of both old and youngconifer forests. Two successional stage maps were
•
developed. One was developed from six TM spectral bandsalone, and the second was developed from six TM spectralbands and a relative sun incidence band. Including thesun incidence band in the classification improved themapping accuracy in the two youngest successional stages,but did not improve overall accuracy or accuracy of thetwo oldest successional stages. Mean spectral values forold-growth and mature stands were compared in seven TMbands and seven band transformations. Differences betweenmature and old-growth successional stages were greatestfor the band ratio of TM 4/5 (P = 0.00005) and the
multiband transformation of wetness (P = 0.00003). Theage of young conifer stands had the highest correlation toTM 4/5 values (r = 0.9559) of any of the TM band or bandtransformations used. TM 4/5 ratio values of poorlyregenerated conifer stands were significantly differentfrom well regenerated conifer stands after age 15(P = 0.0000). TM 4/5 was named a "Successional StageIndex" (SSI) because of its ability to distinguish forestsuccessional stages.
The forest successional stage map was used as inputinto a vertebrate richness model using GIS. The threevariables of 1) successional stage, 2) elevation, and 3)site moisture were used in the GIS to predict the spatialoccurrence of small mammal, amphibian, and reptile speciesbased on primary and secondary habitat requirements.These occurrence or habitat maps were overlayed to tallythe predicted number of vertebrate at any given point inthe study area. Overall, sixty-three and sixty-sevenpercent of the model predictions for vertebrate occurrencematched the vertebrates that were trapped in the field ineight forested stands. Of the three model variables, sitemoisture appeared to have the greatest influence on thepattern of high vertebrate richness in all vertebrateclasses.