Is initial post-disturbance regeneration indicative of longer-term trajectories?

The ability to estimate and model future vegetation dynamics is a central focus of contemporary ecology and is essential for understanding future ecological trajectories. It is therefore critical to understand when the influence of initial post-disturbance regeneration versus stochastic processes dominates long-term post-disturbance ecological processes. Often, conclusions about post-disturbance dynamics are based upon initial regeneration in the years immediately after disturbances. However, the degree to which initial post-disturbance regeneration indicates longer-term trends is likely to be contingent on the types, intensities, and combinations of disturbances, as well as pre-disturbance ecosystem structure and composition. Our relatively limited understanding of why initial post-disturbance regeneration is sometimes a poor predictor of future ecosystem trajectories represents a critical gap in post-disturbance ecological forecasting. We studied the composition and density of regeneration of tree species following wind blowdown in 1997, wildfire in 2002, and compounded disturbances by blowdown and wildfire in subalpine forests of Colorado. We examined regeneration of Picea engelmannii, Abies lasiocarpa, Pinus contorta, and Populus tremuloides in 180 permanent plots across 12 sites (classified by pre-disturbance age and composition) in 2003, 2010, and 2015. At sites that were blown down but not burned, regeneration was dense and dominated by Picea and Abies. At these sites, regeneration observed from 2003 to 2005 (hereafter initial regeneration) was also highly predictive of regeneration 5-10 yr later. In contrast, at sites that were burned and sites that were blown down and burned, regeneration was less dense and dominated by a mix of species. At these sites, initial regeneration was a poor predictor of longer-term trends as species dominance and overall density fluctuated over the 13-yr period. These findings call into question our ability to confidently predict ecosystem trajectories based upon observations made in the years immediately after large, severe disturbances such as wildfires and compounded disturbances. As compounded disturbances become more common under climatically driven changes in disturbance regimes, post-disturbance ecosystem trajectories may become increasingly stochastic and unpredictable.