Past tree influence and prescribed fire exert strong controls on reassembly of mountain grasslands after tree removal.

Woody-plant encroachment represents a global threat to grasslands. Although the causes and consequences of this regime shift have received substantial attention, the processes that constrain reassembly of the grassland state remain poorly understood. We experimentally tested two potentially important controls on reassembly, the past influence of trees and the effects of fire, in conifer-invaded grasslands (mountain meadows) of western Oregon. Previously, we had reconstructed the history of tree invasion at fine spatial and temporal resolution. Using small subplots (10 × 10 m) nested within larger (1-ha) experimental plots, we characterized the fine-scale mosaic of encroachment states, ranging from remnant meadow openings (minimally altered by trees) to century-old forests (lacking meadow species). Subsequently, we removed trees from six plots, of which three were broadcast burned and three remained unburned (except for localized burn piles). Within each plot, subplots were sampled before and periodically after tree removal to quantify the individual and interactive effects of past tree influence and fire on grassland community reassembly. Adjacent, uninvaded meadows served as reference sites. "Past tree influence" was defined as the multivariate (structural or compositional) distance of subplots to reference meadows prior to tree removal. "Reassembly" was defined as the distance, or change in distance, to reference meadows at final sampling. Consistent with theory, we observed greater reassembly of plant community structure than of composition, as loss of meadow specialists was offset by establishment of disturbance-adapted meadow generalists of similar growth form. Nevertheless, eight years after tree removal, most subplots remained structurally and compositionally distinct from reference meadows. Furthermore, fire had both destabilizing and inhibitory effects: it reduced survival of meadow specialists across the range of encroachment states and, where past tree influence was greater, it stalled reassembly by promoting expansion of a highly competitive native meadow sedge. The slow pace of reassembly, despite abundant open space, suggests strong seed limitation: a condition exacerbated by burning. We present a novel test of the importance of past tree influence and fire for restoration of tree-invaded grasslands, offering insights into how constraints on community reassembly vary along a continuum of tree-altered states.

Tree rings provide no evidence of a CO2 fertilization effect in old-growth subalpine forests of western Canada.

Atmospheric CO2 concentrations are now 1.7 times higher than the preindustrial values. Although photosynthetic rates are hypothesized to increase in response to rising atmospheric CO2 concentrations, results from in situ experiments are inconsistent in supporting a CO2 fertilization effect of tree growth. Tree-ring data provide a historical record of tree-level productivity that can be used to evaluate long-term responses of tree growth. We use tree-ring data from old-growth, subalpine forests of western Canada that have not had a stand-replacing disturbance for hundreds of years to determine if growth has increased over 19th and 20th centuries. Our sample consisted of 5,858 trees belonging to five species distributed over two sites in the coastal zone and two in the continental climate of the interior. We calculated annual increments in tree basal area, adjusted these increments for tree size and age, and tested whether there was a detectable temporal trend in tree growth over the 19th and 20th centuries. We found a similar pattern in 20th century growth trends among all species at all sites. Growth during the 19th century was mostly stable or increasing, with the exception of one of the coastal sites, where tree growth was slightly decreasing; whereas growth during the 20th century consistently decreased. The unexpected decrease in growth during the 20th century indicates that there was no CO2 fertilization effect on photosynthesis. We compared the growth trends from our four sites to the trends simulated by seven Earth System Models, and saw that most of the models did not predict these growth declines. Overall, our results indicate that these old-growth forests are unlikely to increase their carbon storage capacity in response to rising atmospheric CO2 , and thus are unlikely to contribute substantially to offsetting future carbon emissions.

A synthesis of radial growth patterns preceding tree mortality.

Tree mortality is a key factor influencing forest functions and dynamics, but our understanding of the mechanisms leading to mortality and the associated changes in tree growth rates are still limited. We compiled a new pan-continental tree-ring width database from sites where both dead and living trees were sampled (2970 dead and 4224 living trees from 190 sites, including 36 species), and compared early and recent growth rates between trees that died and those that survived a given mortality event. We observed a decrease in radial growth before death in ca. 84% of the mortality events. The extent and duration of these reductions were highly variable (1-100 years in 96% of events) due to the complex interactions among study species and the source(s) of mortality. Strong and long-lasting declines were found for gymnosperms, shade- and drought-tolerant species, and trees that died from competition. Angiosperms and trees that died due to biotic attacks (especially bark-beetles) typically showed relatively small and short-term growth reductions. Our analysis did not highlight any universal trade-off between early growth and tree longevity within a species, although this result may also reflect high variability in sampling design among sites. The intersite and interspecific variability in growth patterns before mortality provides valuable information on the nature of the mortality process, which is consistent with our understanding of the physiological mechanisms leading to mortality. Abrupt changes in growth immediately before death can be associated with generalized hydraulic failure and/or bark-beetle attack, while long-term decrease in growth may be associated with a gradual decline in hydraulic performance coupled with depletion in carbon reserves. Our results imply that growth-based mortality algorithms may be a powerful tool for predicting gymnosperm mortality induced by chronic stress, but not necessarily so for angiosperms and in case of intense drought or bark-beetle outbreaks.

Grassland restoration with and without fire: evidence from a tree-removal experiment.

Forest encroachment threatens the biological diversity of grasslands globally. Positive feedbacks can reinforce the process, affecting soils and ground vegetation, ultimately leading to replacement of grassland by forest species. We tested whether restoration treatments (tree removal, with or without fire) reversed effects of nearly two centuries of encroachment by Abies grandis and Pinus contorta into dry, montane meadows in the Cascade Range, Oregon, USA. In nine, 1-ha plots containing a patchy mosaic of meadow openings and forests of varying age (20 to > 140 yr), we compared three treatments affecting the ground vegetation: control (no trees removed), unburned (trees removed, slash burned in piles leaving 90% of the area unburned), and burned (trees removed, slash broadcast burned). We quantified changes over 3-4 years in soils, abundance and richness of species with differing habitat associations (meadow, forest, and ruderal), and recruitment of conifers. Except for a transient increase in available N (especially in burn scars), effects of burning on soils were minimal due, in part, to mixing by gophers. Tree removal greatly benefited meadow species at the expense of forest herbs. Cover and richness of meadow species increased by 47% and 38% of initial values in unburned plots, but changed minimally in burned plots. In contrast, cover and richness of forest herbs declined by 44% and 26% in unburned plots and by 79% and 58% in burned plots. Ruderal species and conifer seedlings were uncommon in both treatments. Although vegetation was consumed beneath burn piles, meadow species recovered significantly after three years. Long-term tree presence did not preclude recovery of meadow species; in fact, colonization was greater in older than in younger forests. In sum, temporal trends were positive for most indicators, suggesting strong potential for restoration. Contrary to conventional wisdom, tree removal without fire may be sufficient to shift the balance from forest to meadow species. In meadows characterized by historically infrequent fire, small-scale disturbances and competitive interactions may be more critical to ecosystem maintenance and restoration. Managers facing the worldwide phenomenon of tree invasion should critically evaluate the ecological vs. operational need for fire in ecosystem restoration.

Variation in responses of late-seral herbs to disturbance and environmental stress.

Clonal herbs that attain maximum development in late-seral forest are often assumed to have similar responses to disturbance and to be functionally equivalent. However, little is known about the demographic or physiological responses of these plants to disturbance or to the altered conditions of the post-disturbance environment. Following harvest of a mature coniferous forest, we compared abundance, demographic changes, and physiological acclimation of three clonal herbs (Asarum caudatum, Clintonia uniflora, and Pyrola picta) that differ in belowground morphology and leaf longevity. We measured ramet density, leaf area, and demographic variables (survival, clonal growth, flowering, and seedling establishment) before and for two years after harvest, and in adjacent undisturbed forest. Acclimation to increased solar radiation was assessed two years after harvest by measuring leaf mass per unit area (LMA) and chlorophyll a:b ratios of leaves produced in the current year. Although initial declines in abundance were similar, demographic responses indicate that patterns of recovery varied greatly among species. Two years after logging, ramet survival and clonal growth (production of new ramets) of Clintonia were greater in the harvest area than in the forest. Asarum had lower survival in the harvest area, but greater clonal growth, and Pyrola showed no difference in either survival or growth between environments. Only Asarum produced seedlings, although their survival was low in the harvest area. All species had higher LMA in the harvest area, but only Clintonia (with annual leaves) had a higher chlorophyll a:b ratio, suggesting the greatest potential for acclimation to increased light. Our results demonstrate that forest herbs with greater rhizome plasticity and shorter leaf duration have greater potential to acclimate after disturbance than those with rigid architectures and persistent leaves. Thus, species with comparable successional roles can vary substantially in their demographic and physiological responses to disturbance, with potential consequences for long-term recovery.

Structure and dynamics of an ancient montane forest in coastal British Columbia.

Old-growth forests are common in the snowy, montane environments of coastal western North America. To examine dynamics of a stand containing four canopy tree species (Abies amabilis, Chamaecyparis nootkatensis, Tsuga mertensiana and T. heterophylla), we used four stem-mapped, 50 m x 50 m plots. From measurements of annual rings, we obtained ages from basal discs of 1,336 live trees, developed master chronologies for each species, reconstructed early growth rates, and delineated periods of release. The stand was ancient; individuals of all four species exceeded 900 years in age, and the oldest tree exceeded 1,400 years. The four plots differed in the timing of events, and we found no evidence of major, stand-level disturbance. Instead the stand was structured by small-scale patch dynamics, resulting from events that affected one to several trees and initiated episodes of release and relatively rapid early growth. The species differed in age structure and dynamics. A. amabilis and T. heterophylla had a classical reverse-J age structure indicative of stable populations, whereas C. nootkatensis and T. mertensiana appeared to rely on local episodes of increased recruitment, which were often separated by centuries, and were probably related to multiple-tree gaps that occurred infrequently. However, such gaps could be considered normal in the long-term history of the stand, and thus these species with their long life spans can persist. Most individuals of all four species grew extremely slowly, with trees typically spending centuries in the understory before reaching the canopy, where they were able to persist for additional centuries. Thus, the key features of this forest are the very slow dynamics dominated by small-scale events, and the slow growth of stress-tolerant trees.

Structure and dynamics of a nearly steady-state subalpine forest in south-central British Columbia, Canada.

We examined the size, age, and spatial structure of trees in an old Engelmann spruce (Picea engelmannii)-subalpine fir (Abies lasiocarpa) forest based on four stem-mapped, 0.25 ha plots. Dendrochronological techniques were used on basal discs of 1,190 trees to reconstruct age and growth pattern, including dates of rapid growth increases. There were no obvious age cohorts or other evidence of past major disturbance. The abundance of both subalpine fir and spruce decreased rapidly with age, especially beyond the ages of 150 years. Very old trees were present, but rare. The best evidence from tree-ring width patterns for past disturbance was a period of release 100 years ago. However, few of the released trees grew into the canopy, which suggests a disturbance of low intensity. Patch dynamics and gap processes were not pronounced in the stand. Clumping was generally weak and only present at small spatial scales (<5 m) for live trees, and largely non-existent for dead trees; mortality was spatially random in this forest. Although spruce were sparse (5.1% of trees) in the forest relative to fir, which is consistent with predictions that fir will ultimately replace spruce in the absence of disturbance, coexistence seems more likely judging from the age structure and numbers of dead trees. In contrast to almost all spruce-fir forests studied previously, the stand we examined showed no record of major disturbances. Thus this stand falls at the limit of the range of dynamics - from disturbance-structured to near steady-state - encompassed in current thinking about forest ecosystems.

Relative reproductive effort in males and females of the dioecious shrub Oemleria cerasiformis.

We derived an index of reproductive effort (g reproductive tissue per g leaf) from data collected over two seasons on 28 males and 28 females of the dioecious shrub Oemleria cerasiformis. Males produced an average of three times as much flower and flower-stalk tissue as females, but because of their large fruits, females produced four times as much total reproductive biomass. Reproductive effort of both sexes was related to light. Fruit set in females (% carpels producing drupes) averaged 11.2% and was related to spring light levels. Male-biased sex ratios in this species may be related to the greater reproductive effort of females.