Growth and yield models are critically important for forest management planning. Biophysical factors such as light, temperature, soil water, and nutrient conditions are known to have major impacts on tree growth. However, it is difficult to incorporate these biophysical variables into growth and yield models due to large variation and complex nonlinear relationships between variables. In this study, artificial intelligence technology was used to develop individual-tree-based basal area (BA) and volume increment models. The models successfully account for the effects of incident solar radiation, growing degree days, and indices of soil water and nutrient availability on BA and volume increments of over 40 species at 5-year intervals. The models were developed using data from over 3000 permanent sample plots across the province of Nova Scotia, Canada. Model validation with independent field data produced model efficiencies of 0.38 and 0.60 for the predictions of BA and volume increments, respectively. The models are applicable to predict tree growth in mixed species, even- or uneven-aged forests in Nova Scotia but can easily be calibrated for other climatic and geographic regions. Artificial neural network models demonstrated better prediction accuracy than conventional regression-based approaches. Artificial intelligence techniques have considerable potential in forest growth and yield modelling.

No abstract

Zhao, Z., Ashraf, M. I., Keys, K. S. and Meng, F-R. 2013. Prediction of soil nutrient regime based on a model of DEM-generated clay content for the province of Nova Scotia, Canada. Can. J. Soil Sci. 93: 193–203. Soil nutrient regime (SNR) maps are widely required by ecological studies as well as forest growth and yield assessment. Traditionally, SNR is assessed in the field using vegetation indicators, topography and soil properties. However, field assessments are expensive, time consuming and not suitable for producing high-resolution SNR maps over a large area. The objective of this research was to develop a new model for producing high-resolution SNR maps over a large area (in this case, the province of Nova Scotia). The model used 10-m resolution clay content maps generated from digital elevation model data to capture local SNR variability (associated with topography) along with coarse-resolution soil maps to capture regional SNR variability (associated with differences in landform/parent material types). Field data from 1385 forest plots were used to calibrate the model and another 125 independent plots were used for model validation. Results showed field-identified SNRs were positively correlated with predicted clay content, with some variability associated with different landform/parent material types. Accuracy assessment showed that 63.7% of model-predicted SNRs were the same as field assessment, with 96.5% within ±1 class compared with field-identified SNRs. The predicted high-resolution SNR map was also able to capture the influence of topography on SNR which was not possible when predicting SNR from coarse-resolution soil maps alone.

Zhao, Z., MacLean, D. A., Bourque, C. P.-A., Swift, D. E. and Meng, F.-R. 2013. Generation of soil drainage equations from an artificial neural network-analysis approach. Can. J. Soil Sci. 93: 329–342. Soil properties, especially soil drainage, are known to be related to topo-hydrologic variables derived from digital elevation models (DEM), such as vertical slope position, slope steepness, sediment delivery ratio, and topographic wetness index. Such relationships typically are strongly non-linear and thus difficult to define with conventional statistical methods. In this study, we used artificial neural network (ANN) models to establish relationships between soil drainage classes and DEM-generated topo-hydrologic variables and subsequently formulated the relationships to generate soil drainage equations for soil mapping. A high-resolution field soil map of the Black Brook Watershed in northwest New Brunswick, Canada, was used to calibrate/validate the ANN models, and the obtained equations. Independent data from an experimental farm, about 180 km away, were also used for validation. Results indicated that vertical slope position was the best predictor of soil drainage classes (r=0.55), followed by slope steepness (r=0.44), sediment delivery ratio (r=0.39), and topographic wetness index (r=0.38). The obtained soil drainage equations fitted well to the ANN model predictions (r 2=0.78–0.99; root mean squared error=0.39–4.55). Analyses indicated that soil drainage equations clearly reflected the actual relationships between soil drainage classes and DEM-generated topo-hydrologic variables, and have the potential to minimize bias originated from over-training the ANN models when applied outside the area of calibration, especially when the ranges of input variables were outside of the range of calibration data.