The main rationale in designing iLand is that addressing climate change, natural disturbances and management in simulations of forest dynamics requires consideration of all three dimensions of complexity described in Table 1.
Processes of population dynamics (i.e. structural complexity) are at the core of forest dynamics- forest structure and composition are essentially mediators between the factors climate, disturbances and management- and need thus explicit consideration in iLand. Furthermore, studying effects of climate change requires a realistic consideration of regeneration processes (e.g., migration) and mortality (e.g., as response to climatic extremes). Climate sensitivity is also determining the need for a raised level of functional complexity (i.e. ecophysiology), increasing model robustness in applications under unique environmental conditions. Element cycling and soil processes are, beyond their relevance for ecosystem functioning, increasingly important for questions of ecosystem management, e.g. with regard to the climate change mitigation potential of forest ecosystems. Spatial complexity at the landscape scale is both important with regard to management, where spatial aspects at this scale are increasingly recognized, as well as in relation to natural disturbances. For large parts of the latter particularly strong climate sensitivities are expected, underlining their relevance in studies of climate change impacts and adaptation.
This analysis leads to the conclusion that all three dimensions need a considerable level of complexity in order to study climate – disturbance – management interactions. However, it is also acknowledged that trade-offs between the achievable degrees of complexity remain (Mladenoff 2004). The approach taken in iLand aims at the level of complexity needed to acknowledge ecosystem behavior at the landscape level as complex adaptive system (e.g. with gap dynamics an emerging property of the system, see Grimm and Railsback 2005), to address ecosystem complexities explicitly (cf. Kimmins et al. 2008) in including processes relevant to its dynamics, while taking aggregated and simplified approaches where possible without limiting the overall analysis capacity towards the models intended application (cf. Mladenoff 2004). iLand aims at a balanced representation of relevant ecosystem dimensions, addressing- at the approximate same level of complexity- the structural, functional and spatial dimensions of forest ecosystems. The modeling approach is thus a fusion of different traditional schools and approaches in forest ecosystem modeling at an intermediate to high level of complexity, while accepting a considerable simplification relative to detailed process-specific models. Focus and novelty in modeling will thus particularly lie on the interactions of processes and dimensions, i.e. the feedbacks between population dynamics, ecophysiology and landscape level processes, as a means to analyze the interactions between climate – disturbances and forest management.
Table 15: Relative levels of complexity in iLand
|population dynamics||structure||spatially explicit (coordinates), pattern-oriented ecological field defines an individuals competitve influence on others||4|
|composition||multiple tree species, parameterized for PNW & central Europe||5|
|regeneration||seed distribution, germination, recruitment into the individual-based model structure, influenced by local light availability, climate and soil factors||4|
|mortality||age- and stress-related mortality components||4|
|management||stand to individual level, spatially explicit, implementation of adaptive management through scripting engine||4|
|ecophysiology||primary production||process-based radiation use efficiency approach, stand level radiation interception, individual level competition and physiology, reduction of potential production by environmental modifiers||3|
|allocation||functional balance, parameterized with empirical allometric equations||3|
|element cycling||closed C, and water cycles (N tbd)||3.5|
|atmospheric processes||temperature, frost, vapor pressure deficit, soil moisture, PAR||4|
|belowground processes||dynamic soil model of C cycling, static water holding capacity, nutrient availability (N tbd)||4|
|landscape context||disturbances||wind, fire, insects||4.5|
|management||spatial and temporal flexibility in landscape scale management (spatial extent watersheds to medium landscapes)||4.5|
|interactions||spatially explict seed dispersal, disturbance processes (spatial extent watersheds to medium landscapes)||4.5|