Authors: Mara McPartland*, University of Minnesota - Minneapolis, Scott St. George, University of Minnesota
Topics: Paleoenvironmental Change, Biogeography, Physical Geography
Keywords: dendrochronology, forest ecology, paleoclimate
Session Type: Paper
Start / End Time: 3:20 PM / 4:35 PM
Room: Governors Square 15, Sheraton, Concourse Level
Presentation File: No File Uploaded
It has been widely-established by dendrochronologists that tree ring measurements express significant temporal autocorrelation across multiple years of growth. The origin of this autocorrelation has alternately been attributed to either the legacy effects of abrupt environmental change on tree growth, or to long-term ocean-atmosphere and land-surface processes. In the former case, hereafter referred to as “biological memory” autocorrelation is considered the result of external influences on the long-term accumulation of photosynthates. In the latter case, referred to here as “climate persistence,” autocorrelation is considered the expression of low-frequency variability stemming from the climate system. In applications in dendroecology, biological memory represents the signal of forest growth responses to climate and environment. By contrast in dendroclimatology, temporal autocorrelation is considered to be non-climatic noise and is often removed during chronology development if it does not mimic the low-frequency patterns exhibited by instrumental records. Here, we consider how diverging perspectives on the sources of temporal autocorrelation affect research outcomes in dendrochronology. We advance a conceptual framework for disentangling biological memory from climatic persistence, and discuss the potential effects of differing assumptions regarding the sources of multi-year autocorrelation for analysis in forest ecology and paleoclimatology. We also consider the implications of framing autocorrelation as biological memory versus climate persistence for conducting large-scale analysis of forest productivity in response to climate change. We hope to elucidate the ways in which autocorrelation is conceptualized and treated in tree ring research, and to highlight areas of agreement and tension present in between these two perspectives.