PP51C-0659
Radial growth response of white spruce ( Picea glauca) to climatic conditions, Mackenzie Delta, NWT
In Canada, 20th century air temperature warming has been greatest in the western Arctic.The prevailing theory of vegetation response to climate suggests warmer temperatures will result in increased growth of trees. However, a number of dendroecological studies from northwestern North America have highlighted a divergence in growth trends during recent decades. Several possible reasons for this response have been suggested. These include warmer temperatures that are exceeding the physiological threshold of white spruce, increased temperatures contributing to drought stress as precipitation is insufficient to offset increasing water demands or possibly large- scale processes such as changes in stratospheric ozone concentration or the effects of global dimming on photosynthetic rates. Here we document additional evidence for sites within the Mackenzie Delta showing a similar divergence trend with regional growing season temperatures. Additionally not all trees at a single site are responding similarly, with many more trees (~75%) having inverse climate-growth responses (negative responders). A much smaller subset continues to exhibit the traditional climate-growth response for treeline species - warmer temperatures equating to greater growth (positive responders). This study will investigate how environmental conditions influence tree growth within the Mackenzie Delta as the growing season progresses. Radial growth of white spruce trees is measured using automatic, high-resolution sensors (point dendrometers). These measurements are compared with environmental data (ground and air temperatures, precipitation, soil moisture, thaw layer depth and solar radiation) collected in-situ so that the critical factors associated with white spruce growth could be determined. The automatic growth sensors and environmental data were installed and monitored for two consecutive growing seasons at a site located north of Inuvik, NT. Tree cores were collected from 42 trees at the site including each of the instrumented trees to determine growth response over time. Measurements of active layer depth, tree health, competition and reproductive success were collected for each sampled tree to determine if any of these characteristics are related to the observed divergence with summer temperatures. This research has implications for the accurate modeling of future carbon sequestration rates within the boreal forest as well as accurate reconstruction of past climatic conditions and environments.
PP51C-0660
Interspecific Variation in Tree Growth Response to 20th Century Climate Variability in the Circumpolar Boreal Forest
We examined relationships between tree ring-width and climate at 232 existing tree-ring sites around the circumpolar boreal forest (north of 55° N). The data set included ten common boreal species: Larix gmelinii, Larix sibirica, Picea abies, Picea glauca, Picea mariana, Picea obovata, Picea sitchensis, Pinus banksiana, Pinus sylvestris, Tsuga mertensiana. We used moving-window correlation analysis for eight 30-year time windows, lagged by 10 years, to characterize the climate response at each site from 1902-2002. We identified two categories of response to temperature: a browning response characterized by inverse correlations between growth and temperature, and a greening response characterized by positive correlations between growth and temperature. Inverse growth responses to temperature were widespread, occurring to some extent in all species, during all time periods, and in nearly all geographic areas. Although the browning response did occur during all time periods, its frequency increased after 1942, during which time the frequency of the greening response declined. Although the trend in the frequency of browning paralleled the rise in Northern Hemisphere temperature in recent decades, the two were not significantly correlated. Browning was concentrated in five species ( Picea abies, Picea glauca, Picea mariana, Picea obovata and Pinus banksiana), and occurred infrequently in the remaining five species. The genus Picea thus appeared to be particularly prone to experience inverse responses to temperature: the only Picea species included in the analysis in which browning did not commonly occur was Picea sitchensis, which is restricted to the coastal temperate rainforests of southeastern and south-central Alaska. Our analyses identified two possible causes of inverse growth responses. First, browning occurred significantly more frequently in the warmer parts of species" sampled ranges, supporting the hypothesis that direct temperature stress might be a cause of inverse growth responses to temperature. Second, in some species, dry sites were also more likely to experience browning; moisture stress might thus be an additional explanation in some cases.
PP51C-0661
Carbon Isotopes and the Diverging Growth Response of Treeline Trees to Changing Climate in Alaska
One of the underlying assumptions in dendroclimatology is that trees respond to climate today the same way they have responded in the past (uniformitarian principle). Recent studies at northern high latitudes treeline show this assumption may no longer be valid or may be flawed, as tree ring width based temperature reconstructions underestimate recent warming. This "divergence effect" might be due to 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). A recent survey of treeline trees in a longitudinal transect across the Alaska and Brooks Ranges in central and northern Alaska (maritime conditions in the west to more arid conditions in the east), has identified 3 responses of tree ring width to warming temperatures at discrete sites; positive (increased growth), negative (decreased growth) and no significant response. We hypothesize that the trees with decreased growth have shifted from temperature to moisture sensitivity as temperatures have increased without a concurrent increase in precipitation or change in snowpack. But there has been no definitive study confirming this. Contrasting this, white spruce growth on productive sites at low elevation sites in central Alaska is best modeled by mean May through August temperature. On such sites there is no threshold change in the prediction efficiency of radial growth across the range of temperatures (residuals are scale-independent) in the 104-yr Fairbanks record. This suggests that low elevation trees consistently have been limited by temperature-induced moisture stress, whereas treeline trees may have been high-temperature limited irregularly in the past, and are now increasingly so in recent decades. For this study, tree cores were collected from 12 white spruce (Picea glauca) trees in the Alaska Range at a site which contains all three responder types. Ring width was measured to confirm to which responder type each tree belongs. Carbon-13 isotopic content of the annual wood of the rings of these 3 response types (for the past 100 years) is being measured to determine if drought stress can explain the differences. Results are still in the preliminary stage but show promise. The premise is that stomatal conductance is reduced under low moisture conditions as leaf pores shut down to conserve water, resulting in a more limited pool of intercellular carbon dioxide for photosynthesis. The hypothesis is that the resulting photosynthate should therefore be heavier as less discrimination of heavier carbon-13 occurs, resulting in a higher ratio of 13C/12C (del13C) incorporated into the wood. While the stomates are open, the energetically more efficient carbon-12 is preferentially incorporated and this ratio is lower. We hope to begin to identify the processes controlling treeline growth, as no intra-site differences (soil moisture and temperature, depth of the A horizon, tree density, slope, aspect, elevation) were documented on any scale which could explain the 3 different growth responses. Isotopes could provide some of the answers.
PP51C-0662
Tree-Ring Evidence for [CO2] Fertilization of Forests and its Influence on the "Divergence Problem"
The "divergence problem" refers to the observation that increasing temperatures are not matched by increasing radial growth at many northern forest sites. A related, and potentially offsetting, effect may be caused by CO2 fertilization of trees. Increased atmospheric CO2 could increase photosynthetic rates and cause trees to use water more efficiently, thereby increasing overall growth rates relative to climatic limiting factors. Accelerated tree growth in response to elevated atmospheric CO2 has been seen across a range of forest types, however these results have been inconsistent, and are generally based on short-term studies. Long-term studies based on tree-rings have generally been restricted to a few sites, and have produced conflicting results. Here we analyze the global record of tree-ring width for evidence of increasing growth relative to drought, and for changing sensitivity of radial growth to drought. This analysis shows that a small but highly significant proportion of trees exhibit increasing growth relative to drought over the past 130 years. These growth increases cannot be attributed to increasing water use efficiency or elevation effects, and no differences between species were detectable. These results suggest that while CO2 fertilization is occurring at some locations and will influence future forest dynamics, it cannot be expected to offset the effects of increasing temperatures, or to substantially slow the rate of carbon accumulation in the atmosphere. It also suggests that the "divergence problem" may be more widespread than is currently recognized, but its effects are obscured by the offsetting effect of CO2 fertilization.
PP51C-0663
Divergence problem in Japanese tree-ring records
It is a critical issue in tree-ring based climatic reconstructions to search for possible causes of divergence between tree-ring and temperature records. The divergence problem has been evidenced mainly by ring-width and density records from circumpolar northern forest sites. In this study, we compiled recently developed tree-ring chronologies in Japan. We performed running correlation analysis between the ring-width data and local climate records. A decreased temperature sensitivity since 1960s was observed in ring-width data for Hinoki cypress trees in central Japan. It was suggested that the divergence at this location was cased by anthropogenic SO2 emission that increased rapidly until 1970 and by then decreased gradually. On the other hand, a Japanese cedar chronology in north-eastern Japan showed stable response to April temperature and increased sensitivity to February, March and May temperatures. Adding some other forest sites to these, we present some conclusions with regards to the current understanding of the divergence problem in Japan.
PP51C-0664 [WITHDRAWN]
Divergence of Temperature and Wind Signals in Tree Rings at Port Angeles, WA
At Port Angeles, WA some divergence of tree response to temperature may be explained by previously reported response of trees to wind. Near a weather station there two intermixed populations of low elevation Douglas-fir were found, of which roughly half showed positive response of annual ring width eccentricity to integrated, instrumentally measured hourly wind velocity over the same time interval and in the same direction. The other half responded negatively to the same wind along the same measurement transect azimuth. A similar result was obtained for wind when using ring width data from the downwind sides of the trees, where much of the wind signal was retained by modulation of early wood growth. This is similar to the Wilmking et al. description of divergent temperature response of transect mean ring width in high elevation trees on the Alaska Brooks Range; where a third showed positive, a third showed negative and a third had insignificant correlation with temperature. Eccentricity in Port Angeles northeast-southwest and southeast-northwest transects shows significant divergence of both wind and wind-associated temperature response from 1997 to 2001. Alternating power and lag of wind and temperature correlations with eccentricity at certain azimuths suggests that wind and temperature may operate interdependently in affecting eccentric growth. Response power and sign were highly dependent on the tree, transect azimuth and lag. Trees responded with wind azimuth resolution close to 10°, suggesting that accuracy and precision of sampling azimuth are important. Response power on most transects in most trees was usually higher at zero response lag. Eccentricity response to temperature and wind was usually of opposite sign. Northeast and southeast transect mean width showed weak, temperature-driven divergence at one-year lag. Analysis is underway back to 1982. At Port Angeles warmer winds from the Strait of Juan de Fuca flowed from the northern sector. No clear time trends were noted. At Port Angeles warmer winds from the Strait of Juan de Fuca flowed from the northern sector. No clear time trends were noted. Cool season temperature was significantly and positively driven by north sector wind. Nearer the ocean at Quillayute Airport positive trends were seen in April-September. In cool season there, temperature was positively driven by frontal wind from the south-southwest with high significance. This suggests that site information on direction, temperature and seasonality of wind may be important in inferring proxy temperature trends from ring width measurements. The cause of ‘divergent' wind-driven growth eccentricity is currently under experimental investigation. The signal may be in response to changes in static loading resulting from crown anisotropy; with some due to dynamic physiological modification in the stem. Divergence of the signal may relate to structural anisotropy affecting torsion in the stem so that two adjacent trees sway and twist in different directions in response to constant wind direction and velocity. Complex airflow within a stand may result from repositioning of vortices, so that stem oscillation directions in adjacent trees vary divergently with change in wind direction and velocity.
PP51C-0665
The Divergence Problem: Some Statistical Considerations
Cook et al. (2004), following Briffa et al 2001, argued that the 'divergence problem' was a 'uniquely recent phenomenon', which they interpreted as evidence for an anthropogenic cause. This finding was cited by both the NAS Panel (2006) and the IPCC (2007). The statistical basis for their argument is analyzed, together with an assessment of the validity and statistical significance, if any, of these and related results http://www.climateaudit.org/pdf/agu07