Search

Changing climate and disturbance regimes are increasingly challenging the resilience of forest ecosystems around the globe. A powerful indicator for the loss of resilience is regeneration failure, that is, the inability of the prevailing tree species to regenerate after disturbance. Regeneration failure can result from the interplay among disturbance changes (e.g., larger and more frequent fires), altered climate conditions (e.g., increased drought), and functional traits (e.g., method of seed dispersal). This complexity makes projections of regeneration failure challenging. Here we applied a novel simulation approach assimilating data-driven fire projections with vegetation responses from process modeling by means of deep neural networks. We (i) quantified the future probability of regeneration failure; (ii) identified spatial hotspots of regeneration failure; and (iii) assessed how current forest types differ in their ability to regenerate under future climate and fire. We focused on the Greater Yellowstone Ecosystem (2.9×106 ha of forest) in the Rocky Mountains of the USA, which has experienced large wildfires in the past and is expected to undergo drastic changes in climate and fire in the future. We simulated four climate scenarios until 2100 at a fine spatial grain (100m). Both wildfire activity and unstocked forest area increased substantially throughout the 21st century in all simulated scenarios. By 2100, between 28% and 59% of the forested area failed to regenerate, indicating considerable loss of resilience. Areas disproportionally at risk occurred where fires are not constrained by topography and in valleys aligned with predominant winds. High-elevation forest types not adapted to fire (i.e., Picea engelmannii-Abies lasiocarpa as well as non-serotinous Pinus contorta var. latifolia forests) were especially vulnerable to regeneration failure. We conclude that changing climate and fire could exceed the resilience of forests in a substantial portion of Greater Yellowstone, with profound implications for carbon, biodiversity, and recreation.

Aim: Climate warming and increased wildfire activity are hypothesized to catalyse biogeographical shifts, reducing the resilience of fire-prone forests world-wide. Two key mechanisms underpinning hypotheses are: (1) reduced seed availability in large stand-replacing burn patches, and (2) reduced seedling establishment/survival after post-fire drought. We tested for regional evidence consistent with these mechanisms in an extensive fire-prone forest biome by assessing post-fire tree seedling establishment, a key indicator of forest resilience. Location: Subalpine forests, US Rocky Mountains. Methods: We analysed post-fire tree seedling establishment from 184 field plots where stand-replacing forest fires were followed by varying post-fire climate conditions. Generalized linear mix ed models tested how establishment rates varied with post-fire drought severity and distance to seed source(among other relevant factors) for tree species with contrasting post-fire regeneration adaptations. Results: Total post-fire tree seedling establishment (all species combined) declined sharply with greater post-fire drought severity and wit h greater distance to seed sources (i.e. the interior of burn patches). Effects varied among key species groups. For conifers that dominate present-da y subalpine forests (Picea engelmannii, Abies lasiocarpa), post-fire seedling establishment declined sharply with both factors. One exception was serotinous Pinus contorta, which did not vary with either factor. For montane species expected to move upslope under future climate change (Larix occidentalis, Pseudotsuga menziesii, Populus tremuloides) and upper tree line species(Pinus albicaulis), establishment was unrelated to either factor. Greater post-fire tree seedling establishment on cooler/wetter aspects suggested local topographic refugia during post-fire droughts. Main Conclusions: If future drought and wildfire patterns manifest as expected, post-fire tree seedling establishment of species that currently characterize subalpine forests could be substantially reduced. Compensatory increases from lower montane and upper tree line species may partially offset these reductions, but our data suggest important near-to mid-term shifts in the composition and structure of high-elevation forests under continued climate warming and increased wildfire activity.

Forest managers in the western US are facing more fires and the most extensive bark beetle outbreaks recorded for the region. • Interactions of wildfire and bark beetle outbreaks and their reciprocal influences on fire behavior, bark beetle dynamics, and ecosystem structure are critical research issues in many western coniferous forests. For the effects of bark beetles on fire: o There is increasing evidence that spruce beetle outbreaks have no effect on the occurrence and severity of stand-replacing fires in spruce-fir forests. o The effect of mountain pine beetle infestations on fire occurrence and severity in lodgepole pine forests are unclear; research results are ambivalent. o Effect of bark beetle outbreaks on fire in other forest types are unknown (ponderosa pine, pinyon-juniper) or need more research (Douglas-fir). o Time-since-beetle outbreak is an important factor to consider in the relationship between bark beetle outbreaks and fire risk because fuels change over time. For the effects of fire on bark beetle outbreaks: o There is increasing evidence that Douglas-fir beetle attack rates are higher on fireinjured Douglas-fir trees. o The evidence regarding effects of fire injury in lodgepole pine on mountain pine beetle attack rates is inconclusive, and whether attacks on fire-injured trees lead to subsequent attacks on healthy trees and/or outbreaks is not known. o Effects of fire injury on attack rates of tree-killing bark beetles in other systems (spruce beetle-Engelmann spruce, mountain pine beetle-ponderosa pine, Jeffrey pine beetle-Jeffrey pine) need more research. o Non-aggressive bark beetles such as Ips and wood borers have higher attack rates on fire-injured trees. o More studies are needed to test whether the above trends in attack rates result in changes in bark beetle reproductive success.