Biostatistical Analysis (5th Edition)
Biostatistical Analysis (5th Edition)
| Hi-index | 0.00 |
In the deciduous forests of the eastern US, timber harvest programmes are often designed to increase the availability of woody browse for terrestrial wildlife. However, assessing the efficacy of timber harvest at increasingly available browse has traditionally required labour-intensive field-based measurements of woody plant growth and abundance. The objective of this study was to use readily available digital aerial imagery to estimate the amount of woody browse in regenerating clearcuts in central West Virginia. Aerial imagery from the National Agriculture Imagery Program and woody browse data collected from 11 regenerating clearcuts in the summer of 2007 were used in this analysis. Red, green and blue visible bands, as well as a simple texture metric, were used to create a multiple linear regression model to predict the amount of woody browse. The final model exhibited large correlation (R 2 = 0.94) and was statistically significant (F = 22.48, p = 0.0009), indicating that simple measures of image digital numbers and texture have potential utility in assisting forest and wildlife managers to assess habitat quality in forest regeneration areas.