Interactions of insects, fire and climate on fuel loads and fire behavior in mixed conifer forest

Mixed-conifer forests in the interior Pacific Northwest are subject to sporadic outbreaks of the western spruce budworm, the most destructive defoliator in western North America. Such outbreaks usually occur synchronously over broad regions and lead to widespread decreases in growth rates and low to moderate levels of mortality. In the last century, changing land use and fire suppression have led to an increase in the amount and density of host tree species, and changed fire regimes. This has altered the severity and frequency of both fire and western spruce budworm. In spite of the ecological and economic significance of these disturbances, their interactions with each other and with climate are not fully understood. We used two approaches to examine these interactions across a range of temporal and spatial scales. First, we used dendrochronological methods to examine the climatic drivers of budworm outbreaks and fires and to assess the association of fire and budworm over three centuries in 13 stands across Oregon, Idaho, and Montana Second, we used a mechanistic fire behavior model, the Wildland-urban interface Fire Dynamics Simulator (WFDS) to examine the sensitivity of crown fire to multiple aspects of defoliated crown fuels, including changing crown bulk density and branchwood moisture. The dendrochronological reconstructions revealed repeated western spruce budworm outbreaks and fires over the past several centuries, with different climate events associated with each disturbance. Outbreaks sometimes persisted more than a decade and were often synchronous among sites. An average of 12 outbreaks occurred at each site, each lasting an average of 12 years in length, with an average of 15 years between outbreaks. Outbreak initiation was often regionally synchronous. Synchrony was higher in the second half of the record (since 1900), possibly due to increased abundance and continuity of host trees during the fire suppression era. Outbreak duration and frequency were also somewhat higher after approximately 1890. We found that warm-dry conditions occurred one to three years preceding outbreak initiation, suggesting that drought-stressed trees permit population growth to a level at which predators no longer strongly limit the budworm population. The mean fire return interval in these mixed-conifer stands was 34 years (range: 16 – 53 years). Fires tended to occur during warm-dry years. We found no evidence of a consistent relationship between the timing of fires and western spruce budworm outbreaks. Western spruce budworm is associated with the ends of droughts and fire is simply associated with single drought years. The simulation study found that defoliation reduces both torching and crowning potential, requiring greater surface fire intensity for crown ignition than undefoliated tree crowns with the same crown base height. Single, highly defoliated trees (80%) experienced little or no torching, and moderately defoliated trees (50%) required about twice the surface fire intensity of undefoliated trees to produce the same heat output. For example, at a surface fire intensity of 700 kW/m 2, 99% of the canopy fuel from the undefoliated tree was consumed, leaving 2 kg of foliage on the tree, compared to 81% consumption of a moderately (50%) defoliated tree, leaving 15 kg of foliage. The effects of defoliation were somewhat mitigated by canopy fuel heterogeneity and potential branchwood drying, but these effects were less pronounced than defolation itself. Our study suggests that areas heavily defoliated by western spruce budworm may inhibit crown fire spread and may thus promote