Dynamic response of the scenic beauty value of different forests to various thinning intensities in central eastern China.

Forest management has a significant influence on the preferences of people for forest landscapes. This study sought to evaluate the dynamic effects of thinning intensities on the landscape value of forests over time. Five typical stands in Wuxiangsi National Forest Park in Nanjing, China, were subjected to a thinning experiment designed with four intensities: unthinned, light thinning, moderate thinning, and heavy thinning. People's preferences for landscape photographs taken in plots under various thinning intensities were assessed through scenic beauty estimation (SBE) at 2 and 5 years after thinning. The differences in scenic beauty value between different thinning intensities were then analyzed with a paired samples t test for the two periods. The results indicated that the landscape value of all of the thinned plots significantly exceeded that of the unthinned plots 2 years after thinning (p < 0.01) and that the heavily thinned plots were most appreciated, showing an average improvement of 9.71 % compared with the control plots. Additionally, the heavily thinned plots were judged to be more beautiful than the lightly thinned and moderately thinned plots, whereas there was no significant difference between moderate thinning and light thinning. At 5 years after thinning, however, the moderately thinned plots received the highest preference scores among the four intensities, displaying an average improvement of 11.32 % compared with the unthinned plots. A multiple linear regression (MLR) model indicated that landscape value improved with increases in the average diameter at breast height (DBH) and with the improvement of environmental cleanliness in the stand, whereas the value decreased with an increasing stem density, species diversity, litter coverage, and canopy density. In addition, we found that the performance of a neural network model based on a multilayer perception (MLP) algorithm for predicting scenic beauty was slightly better than that of the MLR model. The findings of our study suggest that moderate to heavy thinning should be recommended to manage forests for the improvement of forest landscape value.