Salvage logging resets stand structure toward lodgepole pine dominance

Prior to treatment in 2008, all four treatment plots exhibited comparable live basal area (Fig. 18), tree density (Fig. 19), and similar canopy cover (Fig. 22), confirming equivalent initial stand conditions. By 2025, disturbance type and intensity produced distinct overstory recovery trajectories. The Salvage treatment plots exhibited a dramatic increase in live basal area and tree density, reflecting the dense lodgepole pine regeneration pulse characteristic of stand-replacing disturbance. In contrast, the 100-killed treatment plots showed a dramatic decline in live pine basal area and tree density, with surviving stems unable to compensate for the loss of the original overstory cohort. The Control treatment plots showed a moderate decline in live basal area over the 17-year period, consistent with the natural senescence and self-thinning expected in an old lodgepole pine stand exceeding 120 years of age. 

Dead tree basal area and density showed a pronounced pulse in the 100-killed treatment plots, peaking in 2014 (Fig. 20, 21) before declining substantially by 2025 as standing snags collapsed. The Salvage treatment plots showed consistently low and declining dead tree basal area throughout the study period, confirming that clearcutting removed most standing dead material. This structural collapse was accompanied by a pronounced decline in canopy cover across all beetle-affected treatment plots, most severely in the 100-killed treatment (Fig. 22). The Salvage treatment plots showed similarly low canopy cover by 2025, reflecting the early successional stage of the regenerating stand.

At the species level, while shade-tolerant and early-successional species remained rare across all treatment plots (Table 6), live lodgepole pine density differed significantly among treatment plots over time (LMM: Treatment × Year: F₃,₂₀₁ = 168.1, p < 0.001, Table 7). By 2025, the Salvage treatment plots showed the highest lodgepole pine density (3440.7 ± 269.5 trees/ha), while the 100-killed treatment showed the lowest (152.3 ± 23.4 trees/ha), representing an 88% decline from 2008. All treatment plots pairs differed significantly in 2025 (all p ≤ 0.001, Table 8).

These results indicate that salvage logging promotes lodgepole pine dominance through a dense post-disturbance regeneration pulse, while high-intensity natural beetle mortality leads to a progressive decline in overstory density and basal area that does not recover to pre-disturbance levels within 17 years, leaving stands with an open canopy structure and reduced pine dominance whose long-term successional trajectory remains unresolved within the observation period.

Mountain pine beetle mortality accelerates surface fuel accumulation

The long-term monitoring of forest structure reveals a significant redistribution of biomass from the canopy to the forest floor over the 17-year study period. Following disturbance, small downed woody material (DWM) biomass declined progressively across all treatment plots by 2025 (Fig. 23), with Salvage treatment showing an initial pulse in 2010, likely reflecting debris from clearcutting, before declining to levels comparable with those in stands affected by simulated mountain pine beetle by 2025.

Large DWM biomass exhibited significantly different patterns of change in response to different types of disturbance, particularly in the 100-killed stands (Fig. 24). The 100-killed treatment plots exhibited a dramatic and sustained increase from ~30 Mg/ha in 2008 to ~120 Mg/ha by 2025, reflecting the progressive collapse of standing snags onto the forest floor as dead trees lost structural integrity over the 17-year period. This is consistent with the sharp decline in dead tree density observed between 2016 and 2025 in the 100-killed treatment (Fig. 20), indicating that most standing snags had fallen by 2025, transitioning from standing dead biomass to surface fuel loads. The Control and 50-killed treatment plots also showed moderate increases in large DWM by 2025, reaching ~63 Mg/ha, consistent with natural snag fall in aging lodgepole pine stands exceeding 120 years. In contrast, the Salvage treatment plots showed a sustained decline from ~28 Mg/ha in 2008 to ~8 Mg/ha by 2025, as clearcutting removed most merchantable timber and the regenerating stand has not yet produced large woody debris.

Large DWM accumulation differed significantly among treatment plots over time (LMM: Treatment × Year: F₉,₄₁₄ = 13.03, p < 0.001, Table 9). By 2025, the 100-killed treatment plots had accumulated significantly more large DWM than all other treatment plots, exceeding the Control and the 50-killed by 56 Mg/ha, and the Salvage by 111 Mg/ha (all p < 0.0001, Table 10). Control and 50-killed treatment plots did not differ significantly in large DWM accumulation by 2025. This indicates that moderate beetle mortality intensity does not accelerate large fuel accumulation beyond what would be expected from natural senescence in old-growth lodgepole pine stands.

The accumulation of large DWM in naturally recovering beetle-affected stands represents a significant increase in surface fuel loads. This shift in wildfire hazard is from aerial crown-fire potential to high-intensity surface fire risk, particularly in the 100-killed treatment plots. These results suggest that leaving high-intensity beetle-killed stands to recover naturally results in substantially greater coarse fuel accumulation compared to salvage logging, which may require active fuel management interventions to reduce wildfire risk in affected stands.

These structural changes were mirrored in the understory plant community. Prior to clearcutting in 2009, all treatment plots shared similar community composition (Fig. 25). Following disturbance, the Salvage treatment diverged most dramatically, reflecting the rapid establishment of early-successional species in the newly opened stand. However, by 2025 the Salvage community showed signs of shifting back toward a composition more similar to the Control stands, suggesting that as the regenerating lodgepole pine canopy closes, understory conditions may be recovering toward a more mature forest community.

In contrast, the 100-killed treatment plots showed a pulsed successional response, with rapid community shifts between 2010 and 2014 followed by a trajectory toward 2025, likely synchronized with the collapse of standing dead snags and the associated increase in light availability at the forest floor. The 50-killed and Control treatment plots showed the least community turnover, with the biggest shift occurring from 2016 to 2025. PERMANOVA confirmed that understory community composition was significantly influenced by treatment, year, and their interaction, with the full model explaining 16.8% of total community variation (Table 11). 

Altered nutrient cycling distinguishes salvage logged stands

Soil nutrient availability was significantly associated with understory community composition throughout the 17-year study period, with Al, S, Mg, and Fe emerging as the strongest drivers of community differentiation (Table 12, Fig. 26). The elevated base cation availability (Mg, Ca) in Salvage stands likely reflects post-harvest nutrient release following organic matter removal during clearcutting. Elevated ammonium nitrogen in the early Salvage trajectory is consistent with increased nitrogen mineralization following disturbance. In contrast, the association of beetle-affected and Control stands with Al, Fe, and S is consistent with the acidic soil conditions typical of undisturbed lodgepole pine forests. These results suggest that salvage logging produces a distinct soil nutrient environment compared to natural beetle recovery, which may contribute to the divergent understory trajectories observed across disturbance types.