The agricultural industry needs reliable estimates of the probability of specific growing season weather, including extreme events. Farmers and scientists cannot be certain that climatic variability in future decades will not exceed the extremes of weather observed in this century. Since our instrumental records are relatively short, generally less than 100 years, an alternative source of climatic data is required. Most trees preserve a continuous record of annual climate. Dendroclimatology, the study of tree rings in relation to climate, enables the reconstruction of long climatic histories with annual resolution. On the subhumid plains, the best source of high-resolution proxy climate data is the trees of the uplands and sheltered coulees. The collection of tree samples and analysis of growth rings can significantly improve our understanding of climatic variability.
Proxy temperature data have been derived from tree ring studies at the alpine tree line (Luckman 1993; Luckman et al. 1997) and northern boreal forest treeline (D'Arrigo and Jacoby 1992; Jacoby and D'Arrigo 1989; Szeicz and MacDonald 1995) where tree growth is temperature-limited. Based on seven boreal treeline records across North America, D'Arrigo and Jacoby (1992) reconstructed mean annual temperature back to 1600. Their reconstruction shows substantially reduced temperatures in the early 1600s and in the early 1700s, but the longest and most sustained interval of reconstructed colder temperatures is between about 1820 and1880. Szeicz and MacDonald (1995) derived a proxy summer temperature record extending to 1638, based on white spruce (Picea glauca) from the Yukon and Northwest Territories. Their age-dependent reconstruction of June-July temperatures shows close similarities with the record of D'Arrigo and Jacoby (1992), with low summer temperatures in the early 1700s and through the early-mid 19th century, succeeded by a warming trend which is most marked since the early 1900s.
The oldest living trees in the Canadian Rockies are Engelmann spruce (Picea engelmannii) and whitebark pine (Pinus albicaulis) aged about 700 years (Luckman et al. 1984). Living trees can be cross-dated with dead snags and, together with work on other taxa including subalpine fir (Abies lasiocarpa) and alpine larch (Larix lyallii), have been used to derive proxy climate data. Luckman et al. (1997) presented a proxy summer temperature record extending for almost a millennium (1073 - 1983 AD) for the Columbia Icefield area, based on composite records from Abies, Picea and Pinus. They noted that the last few decades (1961-1990) have had the warmest summers in the last 900 years, except perhaps for the late eleventh centur. A "striking feature of the reconstruction is the severity of conditions during the nineteenth century which contains eight of the 10 coldest years and four of the 10 coldest decades" (Luckman et al. 1997: 382-383). The coldest year in the entire record was 1813. Shorter very cold intervals are also evident just before 1700 and around 1780. The warmest intervals are 1073-1110 and 1921-1980. The complexity of this record shows that a subdivision of the last millennium into a Medieval Warm Period and a Little Ice Age is too simplistic.
Case and MacDonald's (1995) analysis of limber pine (Pinus flexilis) tree rings from lower treeline in the foothills of southwestern Alberta has yielded a proxy drought record extending back to 1505. It shows clusters of drought years in the early 1600s, the 1790 - 1800 decade, in the early 1860s, and again around 1920. The late 18th century drought is the most severe in this record, suggesting that agricultural planning based on the 100-year instrumental record may underestimate the potential intensity and duration of natural droughts on the Canadian Plains (Case and MacDonald 1995: 275). Further south, Meko (1992) compiled a regional proxy climate record for the US Great Plains based on tree rings, extending to 1750. This reconstruction also shows intervals of severe drought in the last 200 years, with the years around 1860 being most extreme. Other droughts are apparent in the 1820s, 1890s, 1930s, and 1950s. This record shows a sustained interval of generally wet years between about 1905 to 1925. Based on this record, it appears that the change to drought conditions in the early-mid 20th century (the Dirty Thirties) was spatially and temporally variable, occurring earlier in the southern Canadian plains than further south. Attempts at regional reconstructions may therefore blur local variability.
The age of the subfossil wood can be determined from local written and oral history and from cross-dating, the fundamental technique in dendrochronology, whereby distinctive series of narrow and wider tree rings are identified and matched among wood samples of varying age. Calendar years can then be assigned to the rings of dead wood. This extends tree-ring chronologies much beyond the life spans of living trees, and allows us to determine the timing of pre-historic events that affected tree growth (e.g., droughts). This use of archaeological wood for dendrochronology requires a ring-width series from living trees that overlap in age with the growth rings from the subfossil wood.
Field work also was conducted in the Sweetgrass Hills of north-central Montana where a reconnaissance survey had revealed significant potential for dendroclimatic research (Sauchyn, 1997b). This coniferous forest is considerably older and more diverse than the spruce-pine forest of the Cypress Hills. The East and West Buttes of the Sweetgrass Hills were traversed on foot to collect cores from a variety of trees and range of elevations. The oldest trees were found in the West Butte within kilometres of Alberta. Sampling was focused on the oldest and most abundant trees: 74 cores from lodgepole pine (Pinus contorta), 74 cores from whitebark pine (Pinus albicaulis) and 59 cores from Douglas fir (Pseudotsuga menziesii). Two cores were extracted from each tree, unless the condition of the tree (e.g., basal rot) or its growth form (e.g., recumbent) prevented the collection of more than one core. White spruce (Picea glauca) was not particularly common or old. Therefore only 16 cores were retained from these trees and only for comparison to white spruce ring widths from the Cypress Hills. Limber pine (Pinus flexilis) is widespread in the Sweetgrass Hills and is a common source of dendroclimatic data, including the record from southwestern Alberta (Case and MacDonald, 1995). Cores from this species, however, were in consistently poor condition.
Our sampling represents the first intensive study of the dendrochronology of the Sweetgrass Hills, but not the first tree coring trip. Scientists from the Laboratory of Tree Ring Research, University of Arizona, spent June 30, 1981 in the Sweetgrass Hills, the final destination on a long survey of the western US plains. Dr. David Meko has given us their tree ring widths from that day of sampling: 24 ring width series from Douglas fir and five from Limber pine.
During the fall and winter months of 1997-98, the tree cores from the Cypress and Sweetgrass Hills were prepared (mounted on wooden blocks and sanded) and measured in our tree ring laboratory. Sampling and processing of the wood followed standard field and laboratory procedures (Stokes and Smiley 1968; Ferguson 1970; Schweingruber 1988). Ring-widths were measured to the nearest 0.001 mm using a Velmex UniSlide traversing table and an AcuRite III digital counter. Ring widths from 13 living trees and 30 samples of subfossil wood are the basis for a new 313-year tree-ring chronology from the western Cypress Hills. Cross-dating among the ring-width series was verified using the program COFECHA (Grissino-Mayer et al. 1993). The ring-width series were processed statistically by removing non-climatic (i.e., mostly age-dependent) growth trends and tree-specific bias (Fritts 1976; Monserud 1986). Standardized ring-width indices were computed using the program ARSTAN (Grissino-Mayer et al. 1993) by fitting a negative exponential curve, horizontal line, or a straight line with negative slope to the raw ring-width series.
Climatic reconstruction is based on the statistical relationship between ring width indices and meteorological data. Various instrumental records exist for the Cypress Hills, but they are relatively short and discontinuous (Sauchyn and Porter 1992). Furthermore, the climate of the Cypress Hills is not as relevant here as the climate of the surrounding subhumid plains, where climatic variability has much environmental and economic significance. Thus the best meteorological records are from Maple Creek, about 50 km northeast of the western Cypress Hills and just 25 km north of the Centre Block. Even though these data extend back to 1884, prior to 1941 only 15 years have complete records. After 1941, there are nine years of missing data. Therefore 45 years of mean monthly data from the period 1941-1994 were used to construct a model of tree growth response to climate.
Standardized ring widths and August to July precipitation are the most strongly correlated variables (r = 0.616, p < 0.01). none of the correlations involving temperature variables were significant, as expected, because the climate is subhumid and the living trees were selected from the driest sites. over most of the interior plains, including the southern boreal forest (Larsen and Macdonald 1995), tree growth is limited by the availability of water. thus these trees preserve a record of precipitation variability.
Regression of August to July precipitation against the standardized ring widths (n = 45 years) produced the model (adjusted r2 = 36.5%)
precipitation = 220.83 * swr + 140.22
where swr is standardized ring width. this growth response model was then used to reconstruct august to july precipitation for the period 1682-1994 (Figure 2).
A correlation between reconstructed and instrumental precipitation is visually evident in Figure 2 from the extreme years of the 20th century: the droughts of 1937 and 1961, and the generally above-average annual precipitation of the first two decades, the period of expanding population and crop production as described above. The most significant results of this reconstruction are the dry periods of the 18th and 19th centuries. These droughts were of longer duration than the most prolonged drought of the 20th century, during the 1930s. The droughts of the early and late 18th century also appear in the dendroclimatic record from southwestern Alberta (Case and MacDonald 1995). The anomalous precipitation of 1691 may just reflect the small sample size (three trees) prior to 1711, although this apparent drought warrants the collection of more old wood to verify its possible significance. Besides these possibly spurious data from 1690-91, the driest 12-month period (242 mm) in the tree-ring record is August, 1793 to July, 1974. The drought centred on 1820 is prominent in tree-ring records from the American plains (Meko 1992). Finally and notably, reconstructed precipitation for August, 1857 to July, 1858 (313 mm) was well below the average (373 mm) for the 313-year tree ring record. The Palliser expedition traversed the southern prairies during 1857-59 (Spry 1995).
Although not as prominent as the other drought phases, there are indications of a succession of relatively dry years around 1890 (Figure 2). This correlates with the recognition of drought at this time from other areas, such as in the salinity record from Moon Lake (Laird et al. 1996) and in documentary sources (Jones 1987: 16-19). The 1890s drought stood out in people's memories, despite its lesser severity than other droughts from the last few centuries. Certainly, the recent occupants of the Palliser triangle have not yet experienced prolonged dry spells like those of the 18th and 19th centuries. Although prairie agriculture has historically adapted to climatic variability (Hill and Vaisey 1995), there is a perpetual adjustment to weather and climate. The relatively short instrumental climatic records, usually less 100 years, in western Canada are an inadequate basis for forecasting the severity and duration of future climatic extremes. The knowledge-base for drought forecasting includes proxy data on the frequency and intensity of pre-settlement climatic extremes, that can be derived from the trees of the uplands and sheltered coulees of the subhumid plains (Stockton and Meko 1975; Meko 1992).
Our preliminary results also were presented to a scientific audience in August, during a special session on 20th century environmental change at the annual meeting of the Canadian Association of Geographers in St. John's, Newfoundland. This special session is the basis for a special issue of the Canadian Geographer to be published in 1998. My submission to this special issue, Recent environmental change on the Canadian plains, was co-authored with Dr. Alwynne Beaudoin of the Provincial Museum of Alberta.
In addition to these presentations and publication, we have created a World Wide Web site dedicated to our AFIF project. This Web site, located at
http://www.uregina.ca/~sauchyn/urdendro
thus far contains a description of the project, this progress report, various images illustrating the field work, and links to other tree ring sites.
The other major activity during the second year of our AFIF project will be to initiate the process of applying our dendroclimatic database to an analysis of the impact of climatic variability on the sustainability and diversification of prairie agriculture. This activity will involve extensive collaboration with project technical advisor, Ted O'Brien, and his colleagues at PFRA. I will suggest that we establish a small advisory committee to develop an approach for applying proxy data on the severity and frequency of climate extremes to the planning of land, water and crop management programs.
Case, R. A., and Macdonald, G. M. 1995 'A Dendroclimatic Reconstruction of Annual Precipitation on the Western Canadian Prairies Since A.D. 1505 from Pinus flexilis James' Quaternary Research 44, 267-275
D'Arrigo, R. D. and Jacoby Jr, G. C. 1992 'Dendroclimatic Evidence from Northern North America' in Climate Since A.D. 1500 ed R. S. Bradley and P. D. Jones (London: Routledge) 296-311
Ferguson, C. . 1970. Concepts and Techniques of Dendrochronology. in Scientific Methods in Medieval Archaeology ed R. Berger (Berkeley: University of California Press).
Fritts, H. C. 1976 Tree Rings and Climate London: Academic Press
Grissino-Mayer, H., Holmes, R., and Fritts, H. C. 1993 International Tree Rings Data Bank Program Library User's Manual. Arizona: Laboratory of Tree Ring Research, University of Arizona, Tucson
Hill, H. and Vaisey, J. 1995. Policies for Sustainable Development. In: Wilhite, D.A. and Wood, D.A. (Editors), Planning for a Sustainable Future: The Case of the North American Great Plains. Proceedings of the Symposium, May 8-10, 1995, Lincoln, Nebraska, pp. 51-62.
Jacoby Jr, G. C. and D'Arrigo, R. D. 1989. Reconstructed Northern Hemisphere Annual Temperature Since 1671 Based on High Latitude Tree-Ring Data from North America. Climate Change 14, 39-59.
Jones, D. C. 1987 Empire of Dust: Settling and Abandoning the Prairie Dry Belt Edmonton: University of Alberta Press.
Laird, K. R., Fritz, S. C., Maasch, K. A., and Cumming, B. F. 1996 'Greater Drought Intensity and Frequency Before A.D. 1200 in the Northern Great Plains, USA' Nature 384, 552-554
Larsen, C. P. S, and Macdonald, G. M. 1995. Relations Between Tree-ring Widths, Climate, and Annual Area Burned in the Boreal Forest of Alberta. Canadian Journal of Forest Research 25, 1746-1755
Luckman, B. H. 1993 'Glacier Fluctuation and Tree-Ring Records for the Last Millennium in the Canadian Rockies' Quaternary Science Reviews 12, 441-450
Luckman, B. H., Briffa, K. R., Jones, P. D., Schweingruber, F. H. 1997 'Tree-ring Based Reconstruction of Summer Temperatures at the Columbia Icefield, Alberta, Canada, AD 1073-1983' The Holocene 7, 375-389
Luckman, B. H., Holdsworth, G., Osborn G. D. 1993 'Neoglacial Glacier Fluctuations in the Canadian Rockies' Quaternary Research 39, 144-153
Luckman, B. H., Josza, L.A. and Murphy, P. J. 1984 'Living Seven-Hundred Year Old Picea engelmannii and Pinus albicaulis in the Canadian Rockies' Arctic and Alpine Research 16, 419-422
Meko, D. M. 1992 'Dendroclimatic Evidence from the Great Plains of the United States' in Climate Since A.D. 1500 ed R. S. Bradley and P. D. Jones (London: Routledge) 312-330
Monserud, R. A. 1986. Time Series Analysis Applied to Tree-ring Series. in Proceedings, International Symposium on Ecological Aspects of Tree-Ring Analysis (US Department of Energy Conference - 8608144) 554-562.
Nemanishen, W. 1995. A Draft Discussion Paper: A PFRA Strategy for Drought Investigations in the Palliser Triangle. PFRA, South Alberta Region.
Palliser, J. 1859. Papers relative to the exploration of that portion of British North America which lies between the northern branch of the River Saskatchewan and the frontier of the United States; and between the Red River and the Rocky Mountains. Greenwood Press, New York.
Sauchyn, D.J. 1995. White Bear Lake Relict Tree Stump Project. Final Report to PFRA, March, 1995.
Sauchyn, D.J. 1997a. Precipitation Variability in Southwestern Saskatchewan, Report to ADD Board #10, March, 1997.
Sauchyn, D.J. 1997b. Precipitation Variability In The Southern Alberta Portion Of The Palliser Triangle, Final Report to PFRA Prairie Resources Division, March, 1997.
Sauchyn, D.J. and Porter, S.C. 1992. Pre-Settlement Precipitation Variability in southeastern Alberta and southwestern Saskatchewan. Final Report to PFRA on an Environmental Sustainability Initiative project.
Schweger, C. and Hooey, C. 1991 'Climate Change and the Future of Prairie Agriculture' in Alternative Futures for Prairie Agricultural Communities ed J. Martin (Edmonton: Faculty of Extension, University of Alberta) 1-36
Schweingruber, F. H. 1988 Tree Rings: Basics and Applications of Dendrochronology Boston: D. Reidel
Smith, D. J., Mccarthy, D. P., Colenutt, M. E. 1995 'Little Ice Age Glacial Activity in Peter Lougheed and Elk Lake Provincial Parks, Canadian Rocky Mountains' Canadian Journal of Earth Sciences 32, 579-589
Spry, I. M. 1995 The Palliser Expedition (Saskatoon: Fifth House Ltd) 2nd edition. 315 pp.
Stockton, C. W., and Meko, D. M. 1975. A Long Term History of Drought Occurrence in Western United States As Inferred from Tree-rings. Weatherwise 28(6), 244-249
Stokes, M. A. and Smiley, T. L. 1968 An Introduction to Tree-Ring Dating Chicago: University of Chicago Press
Stommel, H., and Stommel, E. 1979 'The Year Without a Summer' Scientific American 240(6), 176-186
Szeicz, J. M., and Macdonald, G. M. 1995 'Recent White Spruce Dynamics at the Subarctic Alpine Treeline in Northwestern Canada' Journal of Ecology 83, 873-885
