PP54A-01 INVITED
Whither Dendroclimatology?
As in other fields of paleoclimatology, uniformitarianism is the key principle in dendroclimatology. The assumption that the processes that form tree rings now are the same as those in the past is what allows climate to be reconstructed from tree rings. Recent years have seen declining ring widths in the northern high latitudes coincident with increasing temperatures despite being an ostensibly temperature limited environment. There are several factors that could play into this phenomenon. It could be that that more nuanced statistical or process models are needed to fully understand the climate | growth relationship. Or, there could be exogenous forcings (e.g., global dimming) that contribute to the shift in the climate | growth relationship, and that understanding the nature of those forcings is needed. What can this "divergence problem" tell us about tree growth and climate and does the apparent loss of sensitivity indicate a violation of the uniformity principle? We will present an analysis of simulated and real tree-ring data that attempts to answer these questions and show that while there are basic gaps in our understanding, the careful modeling of climate | growth relations and the proper attribution of error are keys to the making progress on the "divergence opportunity."
PP54A-02
Divergence Effect at Alaska's Treelines a Result of Sub-Population Behavior?
The "divergence effect" (off-set between tree-ring based temperature reconstruction and measured temperatures) in northern forest poses a serious question to tree-ring based climate reconstructions, since it seems to violate the uniformitarian principle of dendroclimatology. Several possible reasons emerge, among them false assumptions about 1) climate data (e.g. which climate parameter can be modeled most effectively), 2) tree-ring data (e.g. shift in climate sensitivity of tree growth) or 3) a truly new and unprecedented phenomenon (e.g. rapid climate warming exceeding the adaptive capacity of trees). Here we test, if undetected emergent sub-population behavior at Alaska's treelines might result in a divergence effect. We reanalyzed seven data sets spanning the entire northern treeline in the Brooks Range, Alaska, which showed populations of trees responding positively, non-significant and negatively to recent warming. There is no effect of standardization (i.e. conventional versus RCS) technique on the grouping and we thus used RCS for the following analysis. Without grouping into sub- populations, a clear divergence between tree-ring based temperature reconstruction and actual climate data emerged. However, if only those trees were used for the climate reconstruction, which showed a consistent positive response to the target temperatures (June/July), tree growth modeled climate data extremely well following the rise in temperature during the last decades (actual climate data 0.53°/decade versus tree-ring based reconstruction 0.44°/decade over 1970-2000 period). This result hints at the possibility that some of the reported divergence might be eliminated (at least at Alaska's treelines) through careful test of sub-population behavior.
PP54A-03
Divergent tree growth along the North East Pacific Rim
Tree growth along the northeast Pacific rim makes up a significant portion of carbon uptake and better understanding the relationship between possible divergent tree-growth and warming temperatures is important for the estimation of future carbon budgets. We review previous reports of divergence in the coastal and near-coastal north Pacific region and focus on key sites where treeline, ring-width series are known to have decreased over the past several decades. Preliminary results examining the spatial pattern of divergence along the North Pacific rim suggest that regions of most significant winter/ spring warming correspond with decreases in tree growth. How this warming is linked to decreased growth at and within a stand of trees is uncertain, but may be related to changes in precipitation and shifts in growing seasons. Clusters of ring-width sites within individual fiords show divergence, while others only kilometers away do not. Examination of these relatively local differences and larger, North Pacific wide spatial mapping of divergence may help in identifying the possible causes. A new collection of tree-ring series from central coastal Alaska, from the Alaskan panhandle and examination of sites from British Columbia along with updating at key sites will be presented along with analyses and discussion of potential drivers of observed changes in tree growth.
PP54A-04
Temperature Meets Tree Physiology: Potential Influence of Different Characteristics of Recorded Temperature Increases in Alaska on the Diverging Growth Responses of White Spruce.
To be useful for temperature reconstructions, tree growth must respond to the same climate parameters today in the same way it has in the past. Recent studies which show tree ring widths from the northern high latitudes are diverging from previous sensitivity to temperature parameters may be partly influenced by new patterns of warm temperature anomalies as well as the limits of possible physiological response by trees. Both factors may be influencing white spruce growth in Alaska. In the 104-year record at UES/Fairbanks, daily maximum temperatures (May:August) have increased only slightly (~0.5 C), while daily minima have increased over 3 C per century, and frost-free growing season length has increased 50% to about 120 days (ca. 150 days at lower temperature thresholds suitable for native species). Total yearly days above freezing have increased by about 20 days. Spring snow and ice dissipation is earlier by about 5 to 6 days. Winters mean temperatures are 2 to 3 C greater and include fewer days below –20, -30, and –40 C. At stations across central Alaska with a shorter 60 to 90-year record (McGrath, Bettles, Talkeetna) trends are similar. In Alaska chronologies, high summer temperatures "kill" growth the following year on many sites. Years with high numbers of days with daily maxima above 70 F (21.1 C) are followed by significant "pointer" years (e.g. 1924, 1940-41, 1958-59, 1974-75, 2005) with small tree rings. White spruce is a determinate growth species, meaning current year growth is disproportionately influenced by reserves accumulated in the previous growing season. In a given year, when maximum moisture stress encountered by a particular tree first exceeds a critical threshold, cell production switches from earlywood (large diameter, thin-walled) to latewood (small diameter, thick-walled) type, which limits further width expansion. These climate trends and tree chronology patterns are consistent with (1) a shorter snow accumulation season and longer evapotranspiration season reducing potential spring melt input, which (2) exhausts early season soil moisture earlier and on previously non-sensitive sites, thus (3) inducing cell type switch earlier in the year and on previously non-sensitive sites, resulting in smaller ring formation and negative sensitivity to previous (and secondarily current) summer-season temperature. The next stage in the warming process would be temperature and moisture combinations near lethal conditions which could injure white spruce so severely that their actual realized growth would underperform previous climatic predictions for a 1 to 3 year period. In addition, outbreaks of insects that reduce tree growth would be released from previous climatic limits and become a more significant factor in determining annual net growth, and thus size of ring formed. Finally, empirical relations of growth versus temperature predictors in existing, negative-responding Alaska white spruce populations suggest that temperature increases of 2.5 to 4 C would result in tree death at low elevations represented by climate stations. Environments suitable for white spruce survival might be found in mountain environments.
PP54A-05
Divergence in ringwidth Chronologies from the Yukon Territory, Canada
A new tree-ring chronology database has been developed across the Yukon and northern British Columbia that includes almost 100 upper elevation white spruce chronologies and over 20 new subalpine fir chronologies that extend into the present century. These include approximately 20 sites along the Dempster Highway that bracket the position of the TTHH site where divergence was initially recognized. Preliminary results indicate a mixed response in regional chronology patterns. The dominant pattern (in both spruce and fir) shows maximum 20th century growth in the 1930—50 interval with lower and/or declining values in the late 20th century although some chronologies in the southwest show highest growth in the late 20th century The strongest correlations of ringwidth series from sites along the Dempster Highway with Dawson temperatures show inverse relationships with summer (JJA) and/or June July temperatures in the two years prior to the growth year although residual chronologies at a few sites do show a summer temperature signal. These results suggest current relationships with climate are complex and may reflect local site and or microclimate conditions as well as larger scale climatic controls. Investigation of changing response to temperatures over the 20th century are ongoing will be reported at the meeting.
PP54A-06
Growth Trend Divergence and Statistical Characterization of Climate-Isotope Relations of White Spruce in the Mackenzie Delta, Northern Canada
Ring-width series from temperature-sensitive white spruce forests in the Mackenzie Delta have shown divergent growth trends during the early- to mid-twentieth century at the subpopulation level. Both positive and negative growth responses with respect to recent climate warming in this region have been observed, with the latter response being more common. When ring-width series from these individuals are combined to form master time-series and compared to local/regional instrumental records, the mean growth response tends to diverge more negatively from late-twentieth century temperature trends as the difference between positive and negative responders increases. Growth trend divergence is a phenomenon that has been documented at many other temperature-limited sites and at various spatial scales. To date, this phenomenon remains poorly understood and has added greater uncertainty to ring-width-based climate reconstructions prior to the twentieth century. As such, there is a need to explore alternative climate proxies such as stable isotope ratios in tree-rings to determine if they too exhibit divergence. Here we examine statistical relations between local climate and annually-resolved stable carbon (δ13C) and oxygen (δ18O) isotope ratios from whole-ring alpha-cellulose of three white spruce trees ( Picea glauca [Moench] Voss) in the Mackenzie Delta to evaluate their potential as paleoclimate indicators. Our δ13C record is positively and most strongly associated with maximum growing season (June-August) temperatures. This relation probably results from temperature induced soil moisture stress causing decreased stomatal conductance and increased 13C assimilation, rather than a direct temperature influence on 13C discrimination. Relative humidity was also significantly correlated with δ13C (negative relation) reflecting the direct influence of atmospheric moisture deficit on stomatal conductance. Overall, δ13C appears to be a proxy for effective moisture in this system. The strongest control on interannual δ18O variability was March-July minimum temperatures (positive relation) reflecting the temperature-dependence of δ18O in precipitation and subsequent use of this water during photosynthesis. Growing season relative humidity was also a significant control on δ18O variability (negative relation) through leaf water evaporative enrichment. Although relative humidity has a significant influence on both isotope ratios at annual resolution, lower-frequency variability in these proxies is dominated by temperature. Furthermore, no divergence between these proxies and their associated temperature parameter has been observed during the late-twentieth century. Both δ13C and δ18O time-series presented here contain a significant amount of climate-related variability and can potentially be used to explore pre-instrumental climate changes in the Mackenzie Delta.
PP54A-07 INVITED
Investigating the "divergence problem" with stable carbon isotopes from tree rings.
The trees of the northern forest have responded in an active way to the anthropogenic increase in atmospheric carbon dioxide levels. They have reduced their stomatal conductance to the extent that the ratio of internal to external carbon dioxide concentrations has remained near constant. The dominant effect is that trees have increased their water use efficiency, and the level of moisture stress has therefore declined. This long-term reduction in moisture stress may, in part at least, explain the widely observed reduction in the climate sensitivity of tree ring width and density measurements. We test this hypothesis using stable carbon isotope, ring width and density chronologies from temperature-sensitive sites in Europe. http://geography.swan.ac.uk/treering/
PP54A-08 INVITED
Process model simulations of the divergence effect
We explore the extent to which the Vaganov-Shashkin (VS) model of conifer tree-ring formation can explain evidence for changing relationships between climate and tree growth over recent decades. The VS model is driven by daily environmental forcing (temperature, soil moisture, and solar radiation), and simulates tree-ring growth cell-by-cell as a function of the most limiting environmental control. This simplified representation of tree physiology allows us to examine using a selection of case studies whether instances of divergence may be explained in terms of changes in limiting environmental dependencies or transient climate change. Identification of model-data differences permits further exploration of the effects of tree-ring standardization, atmospheric composition, and additional non-climatic factors.