No abstract

No abstract

No abstract

Accurate soil information is critically important for forest management planning and operations but is challenging to map. Digital soil mapping (DSM) improves upon the limitations of conventional soil mapping by explicitly linking a variety of environmental data layers to spatial soil point datasets to continuously predict soil variability across a landscape. Thus far, much DSM research has focussed on the development of ultra-fine resolution soil maps within agricultural systems; however, increasing availability of LiDAR data presents new opportunities to apply DSM to support forest resource applications at multiple scales. This project describes a DSM workflow using LiDAR-derived elevation data and machine learning models (MLMs) to predict key forest soil attributes. A case study in the Hearst Forest in northeastern Ontario is used to illustrate the workflow. We applied multiple MLMs to the Hearst Forest to predict soil moisture regime and textural class. Both qualitative and quantitative assessment pointed to the Random Forest MLM producing the best maps (63% accuracy for moisture regime and 66% for textural class). Where error occurred, soils were typically misclassified to neighbouring classes. This standardized, flexible workflow is a valuable tool for practitioners that want to undertake DSM as part of forest resource management and planning.

Soil Nutrient Regimes (SNRs) are often incorporated in ecosystem classifications. Evaluation of actual nutrient levels associated with these SNRs and development of complimentary Soil Chemistry Regimes (SCRs) could broaden their utility. Using data from 618 forest stands in northwestern Ontario, we developed five-category SCRs using K-means clustering, and examined relationships among individual nutrients, SCRs, and the SNRs of the Canadian National Vegetation Classification Associations and the Ontario Ecological Land Classification Ecosites. F, A and B horizon samples were analyzed for organic C (OrgC), total N (TotN), C:N ratio (C:N), cation exchange capacity (CEC), exchangeable bases, base saturation (BaSat) and pH. CEC, pH and BaSat showed good correspondence across horizons, and together with C:N accounted for much of the variation in chemical properties. There was broad agreement between Association and Ecosite SNRs and B horizon (BHorz) and All horizon (AllHorz) SCRs. C:N decreased while pH and cation metrics increased with increasing SNR and SCR richness. User’s accuracies (SNRs vs. SCRs) for the classifications ranged from 31-39% but increased to 80-86% for SNR values within +/- one SCR class. Classification trees identified pH class, soil texture and overstory composition as the principal field-measured factors related to BHorzSCRs.