|
Description:
|
This study investigated the use of high resolution satellite imagery in research involving habitat selection , and movement patterns of white -tailed deer and cattle in a semi -arid landscape . Vegetation classification was developed based on Ikonos satellite imagery that was then used to define habitat selection and characterize movement paths of deer and cattle to allow a better understanding of these 2 species . Pasture attributes were also measured to determine animal distribution throughout the study area in relation to roads , fences , water location , and supplemental feeders . Three cattle and 3 free ranging white -tailed deer were used during 3 trials to test seasonal differences in habitat selection and spatial distribution across the pasture .
Ikonos satellite imagery was classified to a final classification accuracy of 83 .6 % . Seven vegetation classes were defined in the classification with 1 class of bare ground / herbaceous that represents interspaces between shrub vegetation . Classification accuracy was obtained using a ? ? meter buffer to all ground control points increasing the accuracy from 71 .29 % .
All physical pasture attributes were significant to animal distributions in the study area when compared to the random distribution . Roads and water location were most important to cattle during the spring and summer . White tailed deer use of the pasture was more dependent on vegetation characteristics than physical attributes . Both cattle and deer selected habitat patches with a proportionately large percentage of bare -ground /interspaces ( >40 % ) . Deer were predominately found in areas containing higher percentages of shrub species , while cattle were found in areas containing a mixture of larger tree species in addition to shrub complexes .
Travel velocity and path tortuosity were measured to determine effects of vegetation attributes on animal movements . Both cattle and deer followed fairly linear paths (Fractal Dimension <1 .2 ) . Factors contributing to path tortuosity included mean patch size , number patches , and patch fractal dimension . Travel velocity was also measured and compared to vegetation association attributes . Travel velocity was significantly different between seasons for white -tailed deer however cattle and deer comparisons were not significant across the 3 trials . Number of patches , patch fractal dimension , mean patch size , and patch area were significant in the travel velocity model . |