Theses Doctoral

Impacts of Partial Cambial Dieback on Tree-Ring Records from Ancient Conifers

Leland, Caroline Wogan

Tree-ring records from long-lived trees are instrumental for understanding climate variability during the Common Era. Some of the oldest and most valuable conifers used to reconstruct past climate exhibit strip-bark morphology, in which vertical segments of the tree have died in response to environmental stress. This form of localized stem mortality, also referred to as partial cambial dieback, is particularly common on conifers growing in xeric, cold, or exposed environments. Some studies note that strip-bark trees have increasing ring-width trends relative to trees with a fully living stem circumference, but there is substantial uncertainty as to what extent partial cambial dieback can influence tree-ring records and subsequent climate reconstructions. This dissertation explores the environmental drivers of partial cambial dieback on Siberian pine (Pinus sibirica Du Tour) from Mongolia, the effect of cambial dieback on the radial growth and physiology of affected trees, and methods for reducing strip-bark biases in tree-ring records.

Chapter 1 assesses the causes and radial growth impacts of partial cambial dieback on Siberian pine trees growing on an ancient lava flow in central Mongolia. Using a combination of field observations and dendrochronological methods, this chapter demonstrates that strip-bark trees from this site exhibit dieback primarily on the southern side of stems, and that dieback was most common during a cold and dry period in the mid-19th century. Given the directionality and timing of dieback on these strip-bark trees, it is hypothesized that localized mortality events are linked to physiological injuries spurred from solar heating combined with unfavorable climatic conditions. This chapter also reveals that strip-bark trees from this site have increasing radial growth trends relative to trees with a full circular morphology (“whole-bark” trees). Strip-bark trees showed an especially rapid increase in ring widths following the cambial dieback period in the mid-19th century, providing initial evidence that dieback events can lead to increasing ring widths in strip-bark Siberian pine.

Chapter 2 seeks to discern the physiological mechanisms of increasing radial growth trends in the Siberian pine strip-bark trees using stable carbon and oxygen isotopes from tree rings. One simple hypothesis is that strip-bark trees show increasing ring-width trends because radial growth is restricted to a smaller stem area after cambial dieback events. Conversely, some studies have hypothesized that increasing ring widths in strip-bark trees reflect a CO2 fertilization effect on growth that is not readily apparent in whole-bark trees. This chapter finds that strip-bark and whole-bark trees responded similarly to increasing atmospheric CO2 and climate variability in their radial growth and leaf-level gas exchange inferred from tree-ring stable isotopes. However, strip-bark and whole-bark trees showed notably different behavior following documented cambial dieback events. After dieback events, strip-bark trees exhibited an increase in ring widths and an enrichment in stable carbon and oxygen isotopes that was not apparent in whole-bark trees. These results further support the notion that partial cambial dieback leads directly to increasing ring widths in strip-bark trees, and that this response could reflect an increase in the ratio of leaf to live stem area after dieback occurs.

Chapters 1 and 2 demonstrate that partial cambial dieback events and morphological changes impact the radial growth and physiology of strip-bark trees. Therefore, prior to developing climate reconstructions, it is necessary to remove variance associated with these non-climatic, morphological changes in tree-ring series. Chapter 3 outlines two chronology development methods for reducing strip-bark biases in tree-ring records. These methods, applied to Siberian pine and Great Basin bristlecone pine (Pinus longaeva Bailey), successfully reduce a strip-bark bias without removing low-to-medium frequency climate variance inferred from whole-bark trees, which were not impacted by dieback activity. While one approach directly corrects the bias in strip-bark series using a whole-bark chronology as a target, another method is based on the development of a low-percentile chronology, which can be applied to a site collection where the stem morphology of individual trees is unknown. Some limitations and caveats of these methods are discussed in context of the analyzed tree species.

The findings from this dissertation have significantly contributed to our understanding of the radial growth and physiological responses of Siberian pine to partial cambial dieback and environmental changes. This dissertation also provides new methods for removing strip-bark biases in tree-ring chronologies. The conclusions presented here have important implications regarding the potential effects of partial cambial dieback on tree-ring records from other tree species and climate reconstructions derived from them. Continued and detailed study of the causes and impacts of partial cambial dieback on other tree species will be critical for understanding the interactions between ancient trees and their environment, and for improving the reliability of climate reconstructions based fully or partly on strip-bark trees.

Geographic Areas


  • thumnail for Leland_columbia_0054D_15339.pdf Leland_columbia_0054D_15339.pdf application/pdf 9.97 MB Download File

More About This Work

Academic Units
Earth and Environmental Sciences
Thesis Advisors
Cook, Edward R.
Andreu Hayles, Laia
Ph.D., Columbia University
Published Here
July 31, 2019