Spruce beetle (Dendroctonus rufipennis Kirby) outbreaks are important disturbances affecting subalpine forests of Engelmann spruce (Picea engelmannii), subalpine fir (Abies lasiocarpa), and lodgepole pine (Pinus contorta) in the southern Rocky Mountains. However, little is known about the influences of these outbreaks on overall forest dynamics. We used age—structure analyses and dendrochronological techniques to investigate the effects of a major spruce beetle outbreak on stand composition, dominance, tree age and size structures, radial growth, and succession in subalpine forests in Colorado. This outbreak, which occurred in the 1940s, caused a shift in dominance from spruce to fir and a reduction in average and maximum tree diameters, heights, and ages. The outbreak did not favor new seedling establishment of the seral lodgepole pine. Thus, in seral stands spruce beetle outbreak accelerates succession towards the shade—tolerant tree species. The predominant response to the outbreak was the release of previously suppressed small—diameter spruce (not attacked by the beetle) and subalpine fir (a non—host species). Following the 1940s outbreak, growth rates of released trees remained high for >40 yr. The relative increases in growth rates were similar for both species. Both spruce and fir will continue to codominate the affected stands. The predominance of accelerated growth following a spruce beetle outbreak, instead of new seedling establishment, is a major contrast to the pattern of stand development following fire. In some Colorado subalpine forests the effects of disturbance by spruce beetle outbreaks appear to be as great as those due to fire.

Abstract. Tree size and age structure, tree‐fall and gap characteristics, and regeneration in gaps were studied in Nothofagus‐dominated old‐growth forests in Tierra del Fuego, Argentina. Gap‐phase regeneration has resulted in all‐aged populations for N. pumilio, N. betuloides, and Drimys winteri, and regeneration in gaps appears to be maintaining coexistence between species in mixed stands. N. betuloides fills many gaps via advance regeneration and some individuals persist for > 150 yr in the understory. Multiple periods of release and suppression indicate that N. betuloides may take advantage of several gap events to reach the main canopy. Likewise, Drimys grows well under closed canopy and can rapidly respond to gap formation, sometimes impeding the regeneration of N. betuloides. In contrast, N. pumilio regenerates in gaps mainly from seed or from advance regeneration of small, ephemeral seedlings. Gap turnover times in Fuegian forests were estimated at 300 ‐ 500 yr, although gap formation was highly episodic and possibly associated with regionally extensive windstorms, earthquakes, and stand‐level dieback. 92 % of gaps involved multiple tree‐falls, and at least 53 % involved secondary expansion. Gap and tree‐fall characteristics in Tierra del Fuego were similar to results from northern Patagonia, Chile, and New Zealand; however, we emphasize that regeneration of Nothofagus spp. and Drimys winteri in gaps depends on associated vegetation and varies along both local and regional environmental gradients.

Aim The purpose of this study was to reconstruct early nineteenth-century vegetation and fire regimes to examine the role of fire, topography, and substrate interactions in shaping landscape and regional vegetation patterns. LocationOur study area was the Current River watershed of the Ozark Highlands in south-central Missouri, USA. MethodsWe combined analysis of early nineteenth-century Public Land Survey (PLS) notes and dendrochronology-based fire histories to reconstruct vegetation and disturbance regimes of pine-oak (Pinus-Quercus) woodlands.Three methods were used to display and analyse PLS data within a Geographic Information System (GIS): (1) simple point distributions for each tree species; (2) section line descriptions of each tree species and other coded features (e.g. 'prairie'); and (3) spatial interpolation of the point-tree data.Vegetation patterns were then related to geological parent material, topography, and mean fire-return intervals from 23 sites using correlation and Canonical Correspondence Analysis (CCA). ResultsThe most striking patterns in the early 1800 s were extensive stands of shortleaf pine (Pinus echinata Mill.) and oak-dominated 'barrens' (savanna) in the frequently burned areas south-west of the Current River, and more mesophytic, fire-sensitive species (red oaks (Quercus rubra L., Q. coccinea Muenchh.), maples (Acer rubrum L., Acer saccharum Marsh), eastern red cedar (Juniperus virginiana L.) in a fire shadow north-east of the river. Several kilometre-wide ecotones of pine-mixed hardwood encompassed the major pineries and barrens.Fire-return intervals and relative dominance of several tree species were strongly correlated at both fine (3-64 km 2 ) and coarse (> 100 km 2 ) spatial scales. At fine scales, relative dominance of shortleaf pine increased with increasing fire frequency during 1701-1820. Relative dominance of black oak (Q. velutina Lam.), and to a lesser extent post oak (Q. stellata Wang.), decreased with increasing fire frequency. Shortleaf pine and these xerophytic oak species occurred on similar bedrock types but were strongly differentiated by fire regimes. Main conclusions Fires exerted strong constraints on vegetation composition and patterns.Historical patterns of Native American occupancy in the region are consistent with the reconstructed vegetation and fire histories and suggest that anthropogenic fire regimes played an overriding role in the development of Ozark vegetation in the 1800s.

We examined the use of dendrochronological techniques and stand age structure for detecting past outbreaks of spruce beetle (Dendroctonusrufipennis Kirby) in subalpine forests in the Colorado Rocky Mountains. The methods were based on the observation that following beetle-caused mortality of canopy trees, growth rates of subcanopy Engelmann spruce (Piceaengelmannii (Parry) Engelm.) and subalpine fir (Abieslasiocarpa (Hook.) Nutt.) increase severalfold and remain high for several decades or more. Although inspection of unstandardized individual tree ring-width chronologies is useful for detecting past disturbance, standardized mean chronologies based on ≥20 cores assure adequate sample sizes and improve interpretability. Standardization of ring-width series by fitting the observed data to a horizontal line passing through the mean ring width facilitates detection of past outbreaks better than other standardization models such as polynomial or cubic spline functions. The incorporation of samples from dead-standing trees increases sample sizes and reveals patterns in the early history of a stand that may not be detectable if samples only from live trees are used. Patterns of release frequencies based on nearly all the trees in a stand were helpful in confirming interpretations not always clear from the chronologies alone. Variations in tree population age structures were related to the history of disturbance by beetle outbreak, but age structure alone was not a reliable means of detecting past outbreaks.