Enhanced Forest Inventories for Habitat Mapping: A Case Study in the Sierra Nevada Mountains of California

Traditional forest inventory systems, originally designed to quantify merchantable timber volume, often lack the spatial resolution and structural detail required for modern multi-resource ecosystem management. In this manuscript, we present an Enhanced Forest Inventory (EFI) and demonstrate its utility for high-resolution wildlife habitat mapping. The project area covers 270,000 acres of the Eldorado National Forest in California’s Sierra Nevada. By integrating 118 ground-truth Forest Inventory and Analysis (FIA) plots with multi-modal remote sensing data (LiDAR, aerial photography, and Sentinel-2 satellite imagery), we developed predictive models for key forest attributes. Our methodology employed a two-tier segmentation approach, partitioning the landscape into approximately 575,000 reporting units with an average size of 0.5 acre to capture forest heterogeneity. We utilized an Elastic-Net Regression framework and automated feature selection to relate remote sensing metrics to ground-measured variables such as basal area, stems per acre, and canopy cover. These physical metrics were translated into functional habitat attributes to evaluate suitability for two focal species: the California Spotted Owl (Strix occidentalis occidentalis) and the Pacific Fisher (Pekania pennanti). Our analysis identified 25,630 acres of nesting and 26,622 acres of foraging habitat for the owl, and 25,636 acres of likely habitat for the fisher based on structural requirements like large-diameter trees and high canopy closure. The results demonstrate that EFIs provide a critical bridge between forestry and conservation ecology, offering forest managers a spatially explicit tool to monitor ecosystem health and manage vulnerable species in complex environments.

💡 Research Summary

This paper presents a comprehensive workflow for creating an Enhanced Forest Inventory (EFI) and applying it to high‑resolution wildlife habitat mapping in the Eldorado National Forest (ENF), California, covering approximately 270,000 acres. The authors integrated 118 ground‑based Forest Inventory and Analysis (FIA) plots with three remote‑sensing modalities: high‑density LiDAR (USGS 3DEP, 2019, ≤0.35 m pulse spacing, ≥8 pts m⁻²), 0.3 m NAIP aerial photography (resampled to 1 m), and 10 m Sentinel‑2 multispectral imagery (13 bands, Level‑2A).

To address the Modifiable Areal Unit Problem (MAUP) and to align statistical units with ecological reality, a two‑tier segmentation was employed. First, “analysis units” mirroring the size and shape of FIA plots were generated for model training. Second, a finer “reporting unit” layer of roughly 0.5 acre polygons (≈575,000 units) was derived from a combination of the LiDAR‑derived digital terrain model, canopy surface, and NAIP vegetation indices. This hierarchical approach ensures that predictions are made at ecologically meaningful scales while preserving statistical comparability with field data.

For each analysis unit, the authors extracted ~1,600 raw features—including height‑stratified LiDAR point densities, terrain metrics (slope, aspect, curvature), and spectral indices (NDVI, EVI, NDWI, band ratios). Feature selection reduced this pool to the 700 most informative variables, mitigating multicollinearity and over‑fitting risks. Elastic‑Net regression—combining L1 and L2 penalties—was then used to model seven key forest attributes: total basal area, soft‑wood basal area, snag basal area, mean tree height, mean DBH, trees per acre (TPA), and canopy cover percentage. Hyper‑parameter tuning employed a grid search with ten‑fold cross‑validation; final models achieved R² > 0.75 and RMSE < 15 % across all targets.

The predicted forest structure metrics were translated into habitat suitability criteria for two focal species. For the California Spotted Owl, nesting habitat required canopy cover ≥ 60 %, soft‑wood basal area ≥ 50 % of total basal area, TPA > 9 stems acre⁻¹, and mean DBH > 25 inches; foraging habitat relaxed the canopy threshold to ≥ 40 %. For the Pacific Fisher, likely habitat required canopy cover ≥ 60 %, soft‑wood basal area ≥ 50 % of total, and mean DBH > 25 inches. Applying these rule‑based filters to the reporting units yielded 25,630 acres of owl nesting habitat, 26,622 acres of owl foraging habitat, and 25,636 acres of likely fisher habitat.

The discussion emphasizes that EFI bridges the gap between timber‑focused inventories and multi‑resource ecosystem management. By delivering forest structure information at a 0.5‑acre resolution, EFI captures fine‑scale heterogeneity essential for species that respond to individual tree size, canopy continuity, and dead‑wood availability. This spatial granularity supports practical management actions such as delineating conservation zones, prioritizing restoration treatments, and monitoring habitat change over time. The authors suggest future extensions including temporal EFI updates, incorporation of additional taxa, and field validation of predicted habitats to further refine model accuracy and ecological relevance.