TY - JOUR
T1 - Modeling multiscale effects of light limitations and edge-induced mortality on carbon stores in forest landscapes
AU - Smithwick, Erica A.H.
AU - Harmon, Mark E.
AU - Domingo, James B.
N1 - Funding Information:
We are grateful to Julia Jones, Barbara Bond, Dave Turner, and Warren Cohen for providing helpful comments on early drafts of this paper. The work was supported by NASA-LCLUC (Land-Cover, Land-Use Change Program), an H.J. Andrews Long Term Ecological Research Grant (DEB-9632921), a National Science Foundation Fellowship in Landscape Ecology through Oregon State University, and the Richardson Endowment to the College of Forestry, Oregon State University.
PY - 2003
Y1 - 2003
N2 - Analyses of carbon (C) dynamics at broad scales usually do not consider spatial interactions. The assumption is that C dynamics can be modeled within homogenous (i.e., even-aged) patches and then summed to predict broad-scale dynamics (an additive approach). The goal of this paper is to elucidate the scales over which this additive approach is sufficient to explain observed C dynamics at broad scales. We define emergent "behaviors" (vs. emergent "properties") as those behaviors that cannot be predicted solely from the additive properties of units at a finer scale. We used a forest process model to check for possible emergent behaviors due to pattern-process interactions at multiple levels, from the patch to the landscape. Specifically, using artificial forest landscapes with various spatial structures, we estimated the relative effects of edge-induced, tree mortality (mainly due to wind) and light limitations on C dynamics. Emergent behaviors were observed at all levels examined, indicating that emergent behaviors did not cease as one proceeded from the patch to the landscape level, as we had expected. However, the magnitude of the emergent behaviors depended on the level of spatial interaction considered as well as the type and intensity of the processes included. In all simulations, interactions of light and wind processes resulted in significant emergent behaviors only when parameters controlling wind mortality were set to the highest levels observed in the literature. In one simulation, the magnitude of emergent behaviors differed among the landscapes, indicating that interactions among patches may not be accounted for by an additive correction for edge effects unless spatial interactions are addressed. The implication is that some C dynamics in fragmented landscapes may not be captured at broad-scales using an additive approach, whereas in other cases spatial interactions are small enough to be ignored.
AB - Analyses of carbon (C) dynamics at broad scales usually do not consider spatial interactions. The assumption is that C dynamics can be modeled within homogenous (i.e., even-aged) patches and then summed to predict broad-scale dynamics (an additive approach). The goal of this paper is to elucidate the scales over which this additive approach is sufficient to explain observed C dynamics at broad scales. We define emergent "behaviors" (vs. emergent "properties") as those behaviors that cannot be predicted solely from the additive properties of units at a finer scale. We used a forest process model to check for possible emergent behaviors due to pattern-process interactions at multiple levels, from the patch to the landscape. Specifically, using artificial forest landscapes with various spatial structures, we estimated the relative effects of edge-induced, tree mortality (mainly due to wind) and light limitations on C dynamics. Emergent behaviors were observed at all levels examined, indicating that emergent behaviors did not cease as one proceeded from the patch to the landscape level, as we had expected. However, the magnitude of the emergent behaviors depended on the level of spatial interaction considered as well as the type and intensity of the processes included. In all simulations, interactions of light and wind processes resulted in significant emergent behaviors only when parameters controlling wind mortality were set to the highest levels observed in the literature. In one simulation, the magnitude of emergent behaviors differed among the landscapes, indicating that interactions among patches may not be accounted for by an additive correction for edge effects unless spatial interactions are addressed. The implication is that some C dynamics in fragmented landscapes may not be captured at broad-scales using an additive approach, whereas in other cases spatial interactions are small enough to be ignored.
UR - http://www.scopus.com/inward/record.url?scp=0348161420&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0348161420&partnerID=8YFLogxK
U2 - 10.1023/B:LAND.0000004254.94982.67
DO - 10.1023/B:LAND.0000004254.94982.67
M3 - Article
AN - SCOPUS:0348161420
SN - 0921-2973
VL - 18
SP - 701
EP - 721
JO - Landscape Ecology
JF - Landscape Ecology
IS - 7
ER -