No abstract

A B S T R A C TLarge uncertainties surrounding root-specific parameters limit model descriptions of belowground processes and ultimately hinder understanding of belowground carbon (C) dynamics and terrestrial biogeochemistry. Despite this recognized shortcoming, it is unclear which processes warrant attention in model development, given the computational cost of additional model complexity. Here, we tested the sensitivity of four models to adjustments in fine root turnover in forested systems: CENTURY, ED2, MC1, and LANDCARB. In general, faster root turnover rates resulted in lower total system carbon (C) and within model changes ranged from 1% to 38% following 100-year simulations. However, the underlying mechanisms driving these changes differed among models as some expressed lower net primary productivity (NPP) with faster turnover rates and others had similar NPP but large shifts in C allocation away from wood to fine roots. Based on these findings we expect that different model responses to changes in fine root turnover will be determined by (1) whether C is allocated to fine roots as fixed portion of NPP or to maintain a fixed biomass ratio between fine roots and leaves or stems and (2) whether soil nutrient and water uptake is a function of both resource availability and fine root biomass or based on resource availability alone. These results suggest that better constrained estimates of fine root turnover will represent a valuable improvement in many terrestrial biosphere models.

Forest policymakers and managers have long sought ways to evaluate the capability of forest landscapes to jointly produce timber, habitat, and other ecosystem services in response to forest management. Currently, carbon is of particular interest as policies for increasing carbon storage on federal lands are being proposed. However, a challenge in joint production analysis of forest management is adequately representing ecological conditions and processes that influence joint production relationships. We used simulation models of vegetation structure, forest sector carbon, and potential wildlife habitat to characterize landscape‐level joint production possibilities for carbon storage, timber harvest, and habitat for seven wildlife species across a range of forest management regimes. We sought to (1) characterize the general relationships of production possibilities for combinations of carbon storage, timber, and habitat, and (2) identify management variables that most influence joint production relationships. Our 160 000‐ha study landscape featured environmental conditions typical of forests in the Western Cascade Mountains of Oregon (USA). Our results indicate that managing forests for carbon storage involves trade‐offs among timber harvest and habitat for focal wildlife species, depending on the disturbance interval and utilization intensity followed. Joint production possibilities for wildlife species varied in shape, ranging from competitive to complementary to compound, reflecting niche breadth and habitat component needs of species examined. Managing Pacific Northwest forests to store forest sector carbon can be roughly complementary with habitat for Northern Spotted Owl, Olive‐sided Flycatcher, and red tree vole. However, managing forests to increase carbon storage potentially can be competitive with timber production and habitat for Pacific marten, Pileated Woodpecker, and Western Bluebird, depending on the disturbance interval and harvest intensity chosen. Our analysis suggests that joint production possibilities under forest management regimes currently typical on industrial forest lands (e.g., 40‐ to 80‐yr rotations with some tree retention for wildlife) represent but a small fraction of joint production outcomes possible in the region. Although the theoretical boundaries of the production possibilities sets we developed are probably unachievable in the current management environment, they arguably define the long‐term potential of managing forests to produce multiple ecosystem services within and across multiple forest ownerships.

No abstract

Planning for old-growth forests requires answers to two large-scale questions: How much old-growth forest should exist? And where can they be sustained in a landscape? Stand-level knowledge of old-growth physiognomy and dynamics are not sufficient to answer these questions. We assert that large-scale disturbance regimes may provide a strong foundation to understand the spatio-temporal ageing patterns in forest landscapes that determine the potential for old growth. Approaches to describe large-scale disturbance regimes range from scenarios reconstructed from historical evidence to simulation of landscapes using predictive models. In this paper, we describe a simulation modelling approach to determine landscape-ageing patterns, and thereby the landscape potential of old-growth forests. A spatially explicit stochastic simulation model of landscape fire-forest cover dynamics was applied to a 1.8 million-ha case study boreal forest landscape to quantify the spatio-temporal variation of landscape ageing. Twenty-five replicates of 200-year simulation runs of the fire disturbance regime, at a 1-ha resolution, generated a suite of variables of landscape ageing and their error estimates. These included temporal variation of older age cohorts over 200 years, survivorship distribution at the 200 th year, and spatial tendencies of ageing. This information, in combination with spatial tendency of species occurrence, constitutes the contextual framework to plan how much old-growth forest a given landscape can sustain, and where such forest could be located.Key words: landscape management, old growth, spatial simulation modelling, landscape ecology, boreal forest, Ontario, fire regime simulation, natural forest disturbances, stochastic models, age-class distribution Il faut se poser deux questions à grande échelle lorsqu'on planifie pour des forêts anciennes : quelle étendue faut-il prévoir pour elles? et où peuvent-elles subsister dans un paysage? Pour répondre à ces questions, la connaissance de la physionomie et de la dynamique des forêts anciennes au niveau du peuplement ne suffit pas. Nous faisons valoir que les régimes des perturbations de grande échelle pourraient représenter une clé importante pour la compréhension des patrons spatiotemporels de vieillissement qui déterminent le potentiel pour une forêt ancienne dans les paysages forestiers. Les approches de description de ces régimes vont de l'élaboration de scénarios en s'appuyant sur des données historiques jusqu'à la simulation de paysages à l'aide de modèles prédictifs. Dans cette communication, nous décrivons une approche de modélisation produisant des simulations du vieillissement d'un paysage pour en déterminer le patron de vieillissement et, donc, le potentiel de développement de forêts anciennes. Un modèle de simulation stochastique spatialement explicite de la dynamique du feu et du couvert forestier a été appliqué à un paysage d'étude de 1,8 million d'hectares dans la forêt boréale en vue de quantifier la variation spatiotemporelle du vieillissement. Vingt-...