Tree-rings contain a wealth of information about climatic conditions affecting the growth and health of trees (Fritts 1976) and are useful indicators of inter-annual to decadal-scale climate variability. Large, thin-walled, light-colored cells are laid down early in the growing season (earlywood) when conditions are favorable for growth. These cells are followed by smaller, thick-walled, dark-colored cells late in the growing season (latewood), which form due to the onset of cooler temperatures and shorter days. An abrupt termination of cell growth occurs at the end of one growing season. The dark-colored latewood cells occur in contrast to the light-colored earlywood of the next growing season, and designates a ring boundary. The two different parameters of the annual tree rings (width and d13C) measured in this study produce distinctly different climatic information.
We will use accepted dendroclimatological techniques (Fritts 1976, Wigley et al. 1983, Cook and Kairiukstis 1990) involving standardization and de-trending of composite tree-ring series and tests of climatic sensitivity. Growth anomalies that may be climatically related (missing ring boundaries from early frosts, consistent narrow rings, coherent stand-wide acceleration of growth) will be used to establish cross-dating and a narrative summary of climate response history will be developed from these observations.
Tree ring-width is the parameter studied in most dendrochronological and dendroclimatological literature. In standard dendroclimatological investigations, great care is taken to restrict sampling of trees to “growth sensitive” sites. Generally this means suitable sampling trees are isolated individuals experiencing no canopy-to-canopy competition on stressful sites where predominately one factor of the environment is limiting (e.g. temperature is presumed to be limiting at treeline). The ring-width record thus can be taken as a pure climatic “signal.” Unfortunately, such sites do not represent entire stands, much less whole forest regions, and whatever ecological information is obtained does not have general applicability. We therefore propose to sample at representative plots within each landscape unit and sample every tree in our plot to obtain area-weighted results representing the actual ecological situation.
We will compare radial growth responses in the tree-rings with available National Weather Service temperature records from the nearest representative stations to the tree-ring sampling locality. Weather stations used in the correlation may not be the physically closest to the tree-ring sampling locality because of mountain and rain shadow effects. The interior Alaska record is recoverable to 1906 (Fairbanks), but many stations only begin in the 1943 to 1947 time period. Thus we will use the past 55 years of climate data, except in central Interior Alaska, where the earlier 20th century record from Fairbanks will be included. We will test radial growth responsiveness to see which station is best correlated to a given stand of trees at a given location. If both data arrays (chronology and climate) exhibit normal distribution, a Pearson correlation will be applied to determine the maximum sensitivity to climatic factors, otherwise a Spearman rank correlation approach will be used. Correlation coefficients will be calculated between the tree-ring chronology and the monthly climate values. We will combine the selected climate factors scoring the strongest correlation (positive or negative) into a normalized index. We will perform sensitivity analysis to determine factors limiting tree growth in the early compared to more recent portions of the record because changes in temperature sensitivity in the mid to late 20th century have been demonstrated for certain high latitude and high altitude treeline trees (Briffa et al. 1998, Lloyd 2003, Wilmking 2003).
Stable isotope content: We will measure the d13C isotope, which is correlated with moisture availability (Francey and Farquhar 1982, Freyer and Belacy 1983). Moisture sensitivity is registered in d13C because stomates remain open for longer periods when moisture is available, allowing for greater exchange of CO2 and thus selection of isotopically lighter carbon. By contrast, during moisture limiting conditions or in arid climates, reduced opportunities to exchange CO2 during prolong periods of stomatal closure results in less discrimination against the heavier d13C, so higher d13C values will result since the carbon that is fixed is skewed to the isotopically heavier form. The combined drought effect of Fairbanks summer temperature and annual precipitation, correlate well with annual d13C content in upland white spruce at Bonanza Creek LTER (Barber et al., 2000).
For d13C analysis, 6 additional samples of tree radii will be obtained from each sampled stand in each landscape unit for tree species present. Multiple radii average out intra-tree variability in isotope content within a tree, and the average of multiple tree samples makes a more representative isotope value on a plot basis (Leavitt and Danzer, 1992). These increment cores will be obtained from the same trees cored for tree-ring analyses, where feasible, in order to maximize the overlap of multiple independent response functions. The trees will be cored near the base, but above the location where the cores for tree-ring analyses were obtained. We estimate that 100 increment cores will be obtained for d13C analysis.
We will prepare wood samples for the past 55 years. This involves slicing the rings year-by-year (4 radii each from 6 trees), grinding the wood (1 g sample), extracting the cellulose (Leavitt and Danzer, 1992), weighing out a standard measurement amount, and submitting to the UAF isotope laboratory. We will plot d13C values by year and analyze the results by statistical comparisons with climate factors and other tree-ring measurements. We will apply the same sensitivity tests through correlation analysis as described for ring-widths, and combine the selected climate factors into a normalized index calibrated by species, landscape unit and region of Alaska. We will also perform a multiple regression to develop a predictive equation if test conditions can be met. The combination of width and isotope analysis offers a powerful set of tools for testing climate sensitivity of trees and tree populations in different regions of Alaska to temperature and precipitation gradients.
Landscape stratification
The first task is to stratify the
landscape using GIS and identify landscape units, which represent the different
site conditions in Interior Alaska.We will use the
Ecoregions of Alaska (Gallant et al., 1995) to broadly stratify interior Alaska
into regions of similar physiography, glaciation, permafrost, soils and
vegetation. The regional precipitation gradient results in moist conditions in
western Interior Alaska and drier conditions further east (Hammond and Yarie,
1996).
Elevation strongly controls the climate within an ecoregion in interior Alaska. Annual precipitation is strongly correlated with elevation. During the summer, the adiabatic cooling rate is approximately 1 degree per 100m (Slaughter and Viereck 1986). We will develop a grid of elevation within each ecoregion as a landscape variable related to climate.
Topography
also strongly controls climate because of the low sun elevation at boreal
latitudes. For example, at 65° N south-facing slope of 10º receives 25 %
greater insolation annually than does a comparable north-facing slope, and 14%
greater radiation than does a flat surface (estimated based on Rouse, 1990).
Topographic shading combined with control of insolation, leads to significant
differences in soil temperature and soil moisture (Slaughter and Long 1974),
permafrost (Dingman and Koutz 1974, Peddle and Franklin 1993), and therefore
forest productivity (Yarie 1983, Viereck and Van Cleve 1984). We will use a
solar insolation model (Dubayah and Rich 1995) that produces a spatially explicit
estimate of growing season insolation as a function of slope orientation,
topographic shading, and seasonal transmittivity. (Figure 3)
We will
then combine the continuous surface variables (elevation and insolation) within
an ecoregion with spectral response pattern from remote sensing data. We will
use 30-meter Landsat ETM+ satellite imagery from 1999-2001 to map candidate
sites that are spectrally similar. Because of disturbance history, successional
trajectories, and legacy factors, the vegetation at similar landscape positions
may differ. However, we expect the vegetation from similar landscape positions
that have similar spectral response patterns to be structurally similar because
factors such as vegetation type, leaf area index, and surface energy are
closely related to spectral patterns (Nemani and Running 1997, Heikkila et al.
2002, Skakun et al. 2003).
We will produce maps of landscape units
within each region that have similar physiographic site conditions (elevation
and insolation) and show similar spectral pattern. These maps will be used to
a) test existing climate-growth relationships; b) guide sampling effort to
establish climate-growth relationships for new site types and/or species; and
c) serve as the basis for extrapolation of results once climate-growth
relationships and area weighted carbon calculations have been established.
Our sampling strategy has to address the different environmental factors at their respective scale in order to comprehensively assess the spatial variability of drought stress in Interior Alaska's boreal forest.
By combining the climate-growth relationships we will develop with the area weighted carbon calculation on a plot basis, we can now develop the tools for an extrapolation into space. Even though, tree density seems to control climate-growth relationships in treeline areas (Wilmking, 2003), we hypothesize, that in Interior Alaska stands are more homogeneous over larger areas (e.g. landscape units). Extrapolation of results will take two steps:
First, we will work first in Central Interior Alaska, since climate-growth relationships there have been partly developed. We will forecast a climate growth relationship for a landscape unit (A) based on a sample from a similar landscape unit (B). We then will test this relationship with newly sampled trees from unit A.
Two possibilities emerge:
1) The relationship holds, we can assume trees on similar landscape units will exhibit similar climate-growth relationships on a scale, which can be differentiated with Landsat-ETM+.
2) The relationship does not hold. Factors acting on smaller scales (tree density, soil moisture regime) exert control over the growth response. We then will use IKONOS imagery (small scale, 1-4m resolution) to map density differences within the landscape unit where the relationship was developed and test, whether the climate-growth relationship is density dependent.
Once we have developed a relationship between landscape unit or density class within a landscape unit to tree growth, we will test this relationship outside of the Central Interior Region. Using an iterative process, we will be able to tie a developed climate-growth relationship to a certain area on a certain scale.
We expect to find a relationship between tree growth and landscape unit or density class. Once we have developed that relationship in all regions, we can extrapolate our plot based results into space using remotely sensed data to cover large areas. Forest stands will be classified by the dominant tree cover—white spruce, paper birch, quaking aspen, mixed-hardwood/spruce, black spruce. We thus can use either landscape units or density classes as the basis, along with allometric equations of tree growth, to extrapolate and calculate above-ground carbon stocks. Allometric equations are already developed for each tree species.
Barber V. A., Juday G. P., Finney B. P. and Wilmking M. (2004) Reconstruction of summer temperatures in Interior Alaska from tree ring proxies: Evidence for changing synoptic climate regimes. Climatic Change, March, XX:XX.
Barber V. A., Juday G. P. and Finney B. P. 2000. Reduced growth of Alaskan white spruce in the twentieth century from temperature-induced drought stress. Nature 405: 668-673.
Briffa, K. R., F. H. Schweingruber, P. D. Jones, T. J. Osborn, S. G. Shiyatov, and E. A. Vagonov. 1998. Reduced sensitivity of recent tree-growth to temperature at high northern latitudes. Nature 391: 678-682.
Cook, E. R. and L. A. Kairiukstis. 1990. Methods of dendrochronology: applications in the environmental sciences. Kluwer, Dordrecht.
Dingman, S. L. and Koutz, F.R. 1974. Relations among
vegetation, permafrost, and potential
insolation in central Alaska. Arctic and Alpine research 6: 37-42.
Dubayah, R. and P. M. Rich. 1995. Topographic solar
radiation models for GIS. International Journal of Geographic Information
Systems. 9(4): 405-419.
Francey, R. J. and G.D. Farquhar. 1982. An explanation of 13C/12C variations in tree rings. Nature 297: 28-31.
Freyer, H. D., and N. Belacy. 1983. 13C/12C Records in northern hemisphere trees during the past 500 years—anthropogenic impact and climatic superpositions. Journal of Geophysical Research. 88: 6844-6852.
Fritts, H. C.
1976. Tree-rings and climate. Academic Press, New York. Heikkila, J. Nevalainen, S. and T. Tokola. 2002. Estimating defoliation
in boreal coniferous forests by combining Landsat TM, aerial photographs and
field data. Forest Ecology and Management. 158: 9-23.
Leavitt, S. W.
and S. R. Danzer. 1992. Method for batch processing small wood samples to
holocellulose for stable-carbon isotope analysis. Analytical Chemistry. 65:
87-89.
Lloyd A.H.,
Fastie C.L. 2002. Spatial and temporal variability in the growth and climate
response of treeline trees in Alaska. Climatic Change 52: 418-509.
Lloyd A.H., Yoshikawa K., Fastie C., Hinzman L. and Fraver M. 2003. Effects of permafrost degradation on woody vegetation at arctic treeline on the Seward Peninsula, Alaska. Permafrost ad Periglacial Processes 14:.
Nemani, R. R., Running, S. W. 1997. Land cover characterization using multi-temporal red, near-IR and thermal-IR data from NOAA/AVHRR. Ecological Applications 7: 79-90.
Peddle, D. R. and S. E. Franklin. 1993. Classification of permafrost active layer depth from remotely sensed and topographic evidence. Remote Sensing of Environment. 44: 67-80.
Skakun, R. S., Wulder, M. A., and S. E. Franklin. 2003. Sensitivity of the thematic mapper enhanced wetness difference index to detect mountain pine beetle red-attack damage. Remote Sensing of Environment. 86(4): 433-443.
Slaughter, C.W. and Long, L.P. 1974. Upland climatic
parameters on subarctic slopes. University of Alaska, Fairbanks, College, AK.
276-280.
Slaughter, C. and L. Viereck. 1986. Climatic characteristics
of the taiga in interior Alaska. Van Cleve, Keith, Chapin, F. S. III, Flanagan,
P. W., Viereck, Leslie A., and Dyrness, C. Theodore. 1986. Forest Ecosystems in
the Alaskan Taiga: a Synthesis of Structure and Function. Ecological Studies
57, 9-21. New York, Springer-Verlag.
Wigley, T. M. L., K. R. Briffa, and P. D. Jones. 1983. On the average value of correlated time series, with applications in dendroclimatology and hydrometeorology. Journal of Climate and Applied Meteorology. 23: 201-213.
Wilmking M., 2003. The treeline ecotone in Interior Alaska
- From theoretical concept to planning application and the science in between.
Ph.D. Dissertation, University of Alaska Fairbanks, School of Natural Resources
and Agricultural Sciences, 130 pp.
Wilmking M., Juday G.P., Barber V.A., Zald H.S.J. (submitted) Recent climate warming forces opposite growth responses of white spruce at treeline in Alaska through temperature thresholds. Global Change Biology.