Wood links terrestrial and freshwater ecosystems, and forest disturbances are a key factor driving its accumulation. Increasingly, wildfires are burning through forests of trees that are already dead, via insect outbreak, drought stress, and the combination of these factors. In this study we ask 1) how recruitment of wood varies according to the pre-fire status of trees: whether they were dead or alive prior to fire; and 2) whether an abundance of fallen ‘bridging’ wood spanning small (<3 m wide) streams affects stream microclimate. Understanding recruitment and potential ecological effects of wood is an important tool in forecasting stream ecosystem structure and function in an era of increasing disturbance loads in western North American forests. We measured tree characteristics (fire severity, size, mortality, species) wood loads, shade, light levels, and water temperatures two years after wildfire and roughly 10 years following an epidemic of native bark beetles in riparian pine-spruce forests (Pinus contorta/Picea glauca ) of central interior British Columbia, Canada. Generalized linear mixed models were used to estimate the probability of tree fall (toppling) into streams based on measured tree characteristics and for the influence of bridging wood on light and temperature of streams.
We found that disturbance history, being killed prior to fire, makes a tree 3.2 times more likely to topple post-fire and a tree’s height is an important attribute influencing the probability of recruitment to streams. The number of dead trees standing pre-fire did not affect wood recruitment overall, since not all trees that fall are oriented toward streams. However, pre-fire mortality effects on wood were most pronounced within lodgepole pine stands severely impacted by the latest bark beetle epidemic. Fire severity, species composition, and tree size were variable across the sites sampled, with pre-fire tree mortality ranging from 8% to 89%. We hypothesized shade and protection from recruited bridgewood might reduce light/temperature levels in the stream and therefore counteract loss of shading from the forest canopy due to wildfire. Toppled wood recruited to bridging positions did not provide a detectable buffer to incoming light/temperature, likely due to variability in additional factors such as shading from riparian understory. Our results indicate the tempo of wood recruitment to streams is affected by overlapping, short-interval disturbances of insect outbreak followed by wildfire. These results can assist forest managers seeking to understand the influence of changing disturbance processes on freshwater habitat in the fish-bearing streams of western North America.