BACKGROUND: THE ECONOMIC IMPACT OF DROUGHT IN WESTERN CANADA
The Problem: Recurring droughts in western Canada cause major economic hardship, social upheaval, and environmental degradation (Atmospheric Environment Service 1986; Godwin 1988). For example, severe droughts during 1936-1937, reduced agricultural output 38.6% in Saskatchewan (SK), 18.5% in Manitoba (MB), and 1.2% in Alberta (AB), while the value of farm inventories declined $58 million (Wheaton et al. 1992). In 1937 alone, net farm income in Saskatchewan dropped $44.5 million, causing widespread farm abandonment and displacement of a quarter of a million prairie inhabitants (Godwin 1988). Recent drought in 1988 caused direct agricultural losses of $1.8 billion, or 0.4% of Canada's real gross domestic product (GDP; Wheaton et al. 1992). Production of major grain crops declined 29% from 1987 levels, while specialty crops such as mustard, dry peas, lentils and canary seed declined 40%. Drought also reduced profits of cattle producers through increased fees to transport animals to moist northern pastures and increased prices for feed and hay. The magnitude of these agricultural losses make recurrent droughts the largest source of risk, uncertainty and hardship in the western Canadian economy.
Droughts also account for the single largest proportion of crop insurance premiums and pay outs in the prairie provinces (Susan Crump, AB Agriculture Financial Corp., pers. comm.). For example,droughts represent 53% of past crop insurance pay outs in Manitoba (Doug Wilcox, MB Crop Insurance Corp., pers. comm.; see letters). In Saskatchewan, drought-related insurance pay outs totalled over $196 million during 1990-1996, while annual drought pay outs in MB and AB averaged$40 million and $104 million, respectively. Severe and multi-year droughts can greatly exceed the average annual pay-outs and stress the economic reserves of crop insurance corporations. This uncertainty destabilizes the economy and increases crop insurance premiums, further reducing the profitability of farms.
Droughts have widespread economic and environmental impacts beyond the agriculture sector. For example, Manitoba Hydro lost $26.4 million in net income as a direct consequence of the 1988drought when it was forced to purchase and consume green house gas-producing fossil fuels to meet consumer energy demands. Similarly, drought mitigation costs for water-related services in MB and SK totalled $14.3 million in 1988 (Prairie Farm Rehabilitation Administration (PFRA), MB Water Services Board, SK Water Corp., all unpub. data). Grain transportation also declined 41% that year, affecting 600-900 employees (Wheaton et al. 1992). Droughts reduced forest production and increased fire loss. For example, fire-fighting costs during the 1988 drought were $24 million in MB and $33.9 million in SK (Wheaton et al. 1992). Costs to other private and public sectors were also substantial, with a 50% loss of wetlands, a 16% decline in duck populations, and an increase in waterfowl disease contributing to over $5 million in lost hunting revenue in MB and SK (Wheaton et al. 1992). Clearly, recurrent and severe droughts impact virtually all aspects of the western Canadian economy and environment. Accurate estimates of the duration, intensity, and periodicity of future droughts are essential for a stable economy and sound resource management in western Canada (Godwin 1988; Roots 1988; Wheaton et al. 1992). Unfortunately, despite the high economic and social costs of drought, and the potential savings that could be derived from better drought prediction, there are few avenues presently available to predict drought occurrence. Consequently,we seek funding to develop and apply a novel technology to estimate the frequency, intensity and duration of droughts in the main agricultural regions of the Canadian prairies.
Drought Prediction: Drought prediction requires anticipation of the climatic fluctuations that produce unusually dry conditions for an extended period of time (Maybank et al. 1995). Drought prediction is complex because the various components of climate that produce droughts span a broad range of behaviours from completely unpredictable (stochastic) to highly cyclic. As well, the periods of such cycles vary from days (boundary-layer turbulence), to weeks (synoptic- scale weather) and years (oceanic-atmospheric linkages; Maybank et al. 1995). In the Canadian prairies,shifts in climate and droughts arise from the interplay among three major air masses: warm dry flow from the Pacific, cold dry Arctic air, and moist tropical air from the Gulf of Mexico (Bryson 1966). Although controversial, recent studies have related prairie drought episodes to a wide variety of repetitive causes, including sunspot and magnetic solar cycles (Laird et al. 1996a) and surface temperatures of the North Pacific Ocean (Namias & Cayan 1981; Trenberth et al. 1988; Bonsal et al. 1993; Latif & Barnett 1994). However, despite a general understanding of the factors that control western Canadian climate and drought occurrence, the long-term history of the region is poorly known (Barnosky et al. 1987). Instrumental records of climatic factors related to drought (e.g.,precipitation, evaporation, temperature) usually extend back fewer than 100 years and are often too brief to characterize the long-term patterns of drought occurrence. In addition, recent paleoclimatological records from the edges of the plains suggest that the last 100 years is not representative of the full range of regional climatic conditions (Case & MacDonald 1995; Laird et al. 1996a). Evidence of more intense droughts prior to 1900 AD warns that present resource planning based on instrumental data underestimates the intensity and duration of droughts in western Canada.
Some paleo climate records indicate that prairie droughts occur at regular intervals that may be used as a basis for drought prediction. For example, based on an analysis of tree-rings, Weakly (1965)concluded that major droughts occurred on a 21-yr cycle, with each episode lasting 13 yrs. Similarly, Laird et al. (1996a) identified significant climatic cycles of 13-18 yr by analyzing sediments of a US prairie lake. These strongly recurrent patterns of drought could be used to alleviate uncertainty in agricultural and resource management if they were recorded in western Canada. Unfortunately, standard paleoclimatology methods are inadequate for drought prediction in agriculturally-important regions of the prairies. A major problem is the lack of suitable trees for tree-ring analyses on the agricultural plains (Fritz et al. 1994). The longest drought records (487years) are from tree-ring analyses from the eastern foothills of the Rocky Mountains, AB (Case &MacDonald 1995), a site which shows idiosyncracies in the timing, intensity and duration of droughts due to local orthographic precipitation (Reinelt 1970). Similarly, most tree-ring records from the US plains are from river-valleys (e.g., Will 1946; Meko 1992) that are under different climatic and hydrologic control than upland agricultural districts (Fritz et al. 1994). Pollen studies have also proven inadequate for high-resolution studies on the prairies because climatic interpretations are obscured by low taxonomic resolution of fossil records from grasslands (family level only, 'grasses'). To circumvent these problems, we will use recently developed paleolimnological techniques to quantitatively infer climatic change from the sedimentary records of moderately saline prairie lakes (Fritz et al. 1994; Laird et al. 1996a; Laird et al. 1996b). These records will be used in novel time series models to predict future drought occurrence.
Return to
THE PALEOLIMNOLOGICAL APPROACH:
We will analyze sediments from hydrologically-closed lakes on 1-3 year intervals to estimate the frequency, duration and intensity of droughts in the Canadian prairies during the past 800 years. In the semi-arid Canadian prairies where evaporation exceeds precipitation, lake-level and lakewater salinity are mainly controlled by temperature and precipitation balance. These are the same factors that control drought frequency (measured as the Palmer, Bahlme-Mooley, or Versatile Soil Moisture Budget index; Palmer 1965; Bahlme & Mooley 1980; Akinremi et al. 1996). During warmer or drier periods, lake levels drop in closed basins and dissolved salts concentrate. Conversely, cool wet climates result in high lake levels and dilution of brine concentrations. The strong osmotic stress imposed by lakewater salinity is the main determinant of the species composition in saline lakes(Hammer et al. 1983; Hammer 1990). Therefore, with knowledge of the salinity optima and tolerance of aquatic organisms, it is possible to accurately reconstruct past variations in concentrations of dissolved substances (and hence climatic agents) from an analysis of fossil remains. Such fossil reconstructions are increasingly used for US prairie lakes (Fritz 1996; Fritz et al. 1991, 1994; Laird et al. 1996a) and central British Columbia (Smol & Walker NSERC Strategic0133979), but are rare in the Canadian prairies (reviewed in Vance 1997; Wilson et al. 1996, 1997),and have never been conducted at the high temporal resolution proposed here.
This project builds upon extensive documentation of the salinity optima and tolerance of diatom communities in lakes (Fritz et al. 1991, Cumming et al. 1995, Wilson et al. 1996, Laird et al. 1996a,1996b) to reconstruct recent changes in drought on the Canadian prairies. Diatoms are common and abundant members of the algal flora of inland saline lakes whose distribution is strongly related to lakewater salinity (Cumming et al. 1995; Gasse et al. 1995; Wilson et al. 1994; Fritz et al. 1991;Fritz et al. , 1993), and whose fossil species composition allows quantitative reconstruction of past lake salinity (Fritz et al. 1991, 1994; Laird et al. 1996a, 1996b). One of the novel aspects of this project is that our diatom-inferences will be verified with 50-100 yr instrumental records of drought,climate change, and crop yields, and compared with independent analysis of biochemical fossils from algae and bacteria, important additional paleoindicators of climatic change (Pienitz et al. 1992). As well, we will use new nonparametric analyses and point-process models of paleoclimatic data from diatoms (Chen et al. 1996) to predict the occurrence and characteristics of droughts over the next 5-50 years.
Return to
OBJECTIVES:
Return to
Main Page or Table of Contents.STUDY AREA:
This study will reconstruct drought occurrence during the past 800 years in three semi-arid agricultural regions of western Canada. Precise study locations will be determined in consultation with our user sector during the first workshop, and will be based on preliminary sampling during the first 6 months of the project. In all cases, sites will be selected to ensure that lakewater salinity closely reflects changes in effective moisture (Evaporation-Precipitation; E-P) and that sediment accumulation rates are sufficient to provide high-resolution records at intervals of 1-3 years. In general, high-resolution climatic reconstructions will be conducted in small (<0.5 km2), relatively deep (5-10 m), hydrologically- closed basins (no surface flow), with undisturbed sediments and present-day lakewater salinity of 1-5 g l-1 total dissolved solids (TDS). Under these conditions,groundwater recharge and lake salinity is primarily controlled by precipitation (Fritz 1996), and diatom species composition is highly sensitive to climatically-driven changes in salinity (Fritz et al.1993; Laird et al. 1996a). As well, we will select lakes to lie within 100 km of long-term climatological stations (50-100 yr) to verify paleoclimatic reconstructions with real-time observations.
In consultation with our user sector, we have selected a preliminary set of study regions which encompass the main agricultural, climatic and socio-economic zones of the Canadian prairies. At each site, we have identified a principal study lake, as well as a series of alternates. In Manitoba,study lakes lie in the prairie pothole region south of Riding Mountain National Park (50o 35'N, 100o13'W), an area of black soil and mixed grassland that supports cereal cultivation (40-85% of area). The main lake, known as FRB No. 28 (50o 34.7'N, 100 o 12.8'W), has an appropriate surface area(10.1 ha), maximum depth (6.1 m), and modern salinity (2.3 g TDS l-1) and is maintained by surface runoff and groundwater alone (Fedoruk 1971). Alternate lakes include Horseshoe Lake (50o 38.7'N,100o 24.2'W), FRB No. 23 (50o 33'N, 100o 03.5'W), and Beaufort Lake (50o 32'N, 100o 11.3'W), all small (10-113 ha), relatively deep (5.5-10 m), and moderately saline sites (1.6-8.0 g TDS l-1). All lakes lack hydrological control structures, are undisturbed by fisheries manipulations, and exhibit thermal stratification indicating tranquil depositional environments.
In Saskatchewan lakes, we have identified two lakes that lie within the dark brown soil zone. Crater Lake (50o59'N 102o00'W) is a small closed site (4.6 ha; 6 m depth) with populations of Ruppia(Christiansen Associates, Saskatoon, SK; pers. comm.), a plant characteristic of saline lakes(Husband and Hickman 1985), whereas Lenore Lake (52o30'N, 104o59'W) offers a small, protected embayment (<100 ha; 6 m) of moderate salinity (6.3 g TDS l-1). Although Humbolt Lake (52o09'N,105o06'W), Wakaw Lake (52o40'N, 105o35'W) and Fishing Lake (51o51'N, 103o33'W) are all larger sites (10-29 km-2; 8-20 m), previous studies have demonstrated that both lake-level and salinity (3.5-4.0 g TDS l-1) change in response to precipitation (Hammer 1978) and that fossil diatoms are well-preserved (Fritz et al. 1993). All catchments are 90-95% agricultural.
Our Alberta lakes are located in the black Chernozem soil zone, an agricultural region southeast of Edmonton. In particular, Mink Lake (53o31'N, 114o14'W) is of an appropriate size (0.4 km2), depth(7 m), and salinity (1.2 g TDS l-1) for paleoclimatic reconstructions. Alternate lakes include Camp (53o05'N, 111o32'W) and Miquelon (53o15'N, 112o55'W), both comparatively large (1.4-7.8 km-2), shallow (4 m) lakes with appropriate modern salinities (1.2-5.4 g TDS l-1). All sites are hydrologically-closed and lack control structures.
Exact sites for comprehensive climate reconstruction will be selected following analyses of preliminary cores collected in each region during Jan-May 1998 (Milestone 1). All cores will be examined for evidence of depositional hiatus (hardpan, grain size, plant roots), fossil preservation status (diatoms, pigments), and the presence of annual laminae (varves) that might improve the temporal resolution of our climatic reconstructions. Results will be presented to the user sector during the first workshop in June 98 (Milestone 2) to determine final site selection for collection of master cores during summer 1998 (Milestone 3). Final site selection criteria will include the quality of the sedimentary record, proximity to long-term climatic data, and the needs of the user sector.
Return to Main Page or Table of
Contents.
METHODS:
A multiple indicator approach using diatom frustules and pigments from bacteria and algae will be used to measure historical changes in lakewater salinity and drought in western Canada during the past 800 years. These paleoclimatological reconstructions will provide input data for point- process and nonparametric time series models to predict future drought occurrence in the prairie Provinces. Space limitations do not allow us to detail our field and lab protocols, but all aspects of sediment coring and sectioning, microscopic analysis, quality assurance and quality control ( QA/QC)guidelines, etc. will follow established protocols which have been subjected to peer and agency reviews. Diatom procedures have been rigorously developed and tested by Dr. Hall as a member of Dr. J.P. Smol's laboratory at Queen's University, ON (e.g., Hall & Smol 1992, 1996). Fossil pigment analyses use state-of-the-art high performance liquid chromatography (HPLC) systems(Leavitt et al. 1989), fully calibrated with authentic US-EPA standards. These techniques have all been developed for high- resolution paleoclimate and lake management work. All fossil reconstructions will be verified and calibrated against historical climate, crop, and land-use data provided and interpreted by our user sector (e.g., Atmospheric Environment Services, Prairie Crop Insurance Corps., Canadian Wheat Board, Canadian Wildlife Service, Ducks Unlimited, etc.; seed eaters of support). Below we describe briefly the approaches and indicators we will use for drought reconstruction and prediction.
Siliceous Algal Indicators (diatoms): The mainstay of most fossil analyses continues to be diatoms from lakes. These organisms have highly distinctive cell walls (frustules) made of biogenic silica('glass') that preserves for 1000s of years in lake sediments. Diatoms are abundant members of the algal flora of inland saline lakes whose distribution is strongly related to lakewater salinity, pH and conductivity, all climatically- sensitive characteristics of closed basins (Fritz et al. 1991; Fritz et al.1993; Wilson et al. 1994; Cumming et al. 1995; Gasse et al. 1995). Consequently, analysis of the fossil diatom species composition allows quantification of past trends in lake salinity and associated chemical characteristics (Fritz et al. 1991; Cumming & Smol, 1993; Fritz et al., 1993; Wilson et al.,1994; Laird et al. 1996a, 1996b).
Quantitative inferences of lakewater salinity and drought will be estimated from fossil records using the diatom-inference salinity model developed by Dr. Fritz specifically for use in the prairies (Fritzet al. 1993, 1994; Laird et al. 1996a, 1996b). This model quantifies diatom distributions among 66lakes spanning a wide range of salinity (0.65-270 g TDS l-1) and brine types (Fritz et al. 1993). All paleosalinity reconstructions will be based on weighted-averaging regression and calibration (Lineet al. 1994) using state-of-the-art statistical programs (WACALIB v. 3.3) to estimate salinity optima of diatom taxa (Cumming & Smol 1993; Fritz et al. 1993; Wilson et al. 1994, 1996).) and to determine realistic error estimates based on computer-intensive bootstrapping procedures (Line et al. 1994). The predictive ability of these diatom-based salinity models is strong and highly significant (R2 = 0.8-0.9; Cumming & Smol 1993; Fritz et al. 1993; Wilson et al. 1994, 1996),particularly in the salinity range of our study regions (1-20 g TDS l-1; Fritz et al. 1993; Cumming et al. 1995).
The power of using diatom species composition as a monitor of environmental change has been recognized by the United States Environmental Protection Agency (US-EPA) nation-wide EMAP program, which is using diatoms in their water quality biomonitoring. Our diatom salinity inferences are firmly placed in the "species approach", whereby correct and documentable taxonomic identification are the most important considerations of quality assurance/ quality control. Consequently, all of our species identification will be conducted by Dr. Roland Hall, an expert on diatom taxonomy who has trained with Dr. J.P. Smol, a leading authority on the use of diatoms in environmental inference, and a frequent collaborator of Leavitt's. Additionally, one of the originators of drought reconstruction using fossil diatoms, Dr. S. Fritz (Dept. Earth and Environmental Sciences, Lehigh University, PA) will collaborate on this project at no cost. Routine analyses will include scanning electron microscopy to confirm difficult species identifications. Photomicrographs of new and problematic taxa will be catalogued and archived, along with reference samples from all lakes, at the Canadian Museum of Nature, Ottawa. Finally, we will continue to develop and evaluate new multivariate statistical methods to determine the reliability of transfer functions for reconstruction of past water chemistry (Hall & Smol 1992, 1996; Hall et al.under review).
Fossil Pigments: Chlorophylls (Chls) and carotenoids from algae, photosynthetic bacteria, and higher plants often preserve in freshwater sediments long after all morphological structures have disappeared (Brown 1969). Unmodified algal carotenoids are typically present throughout the post glacial history of lakes in the northern hemisphere (reviewed in Sanger 1988). Biogeochemical fluxes of pigments have been extensively studied in a wide variety of lakes (Leavitt & Carpenter1990; Hurley & Armstrong 1991) and the processes that regulate pigment deposition are particularly well documented (reviewed in Leavitt 1993). Many of these pigments are unique to specific phytoplankton groups and allow independent assessment of the past production dynamics of green algae, cryptophytes, diatoms and chrysophytes, and cyanobacteria (Leavitt & Findlay 1994). Our previous analysis of carotenoids in 113 saline lakes from British Columbia, SK and the Yukon demonstrates that fossil pigments are well-preserved, sensitive to changes in lakewater salinity, and provide an independent confirmation of diatom inferences (Pienitz et al. 1992; Vinebrooke et al.under review). Because most pigments are deposited at the lake bottom in direct proportion to the abundance of algae (Leavitt 1993; Leavitt & Findlay 1994), analyses of fossil pigments also allows estimation of past algal populations, including many that cause severe water-quality problems. Consequently, an important offshoot of this project is that we will be able to measure any reductions in water quality that arise during droughts, a factor that regularly reduces livestock and fish production on the prairies (Prairie Farm Rehabilitation Association, unpubl. data; Barica, 1987). All pigment analyses use high-performance liquid chromatography (HPLC) and photodiode array analysis - the method of choice for many studies of local, regional, and global changes in water quality and algal community composition (Millie et al. 1993). Our HPLC system is calibrated with authentic standards provided by US-EPA. Leavitt (1993) provides a literature review documenting the assumptions and power of the pigment approach.
Time Series Modelling and Data Analyses:
An important and unique feature of this proposal is the use of paleoclimatic data to predict future drought occurrence in the Canadian prairies. Our approach uses standard time series analyses to estimate the frequency of past drought events, as well as to determine if the incidence of drought is increasing or declining. However, because the validity of these analyses is sensitive to assumptions of data properties that are difficult to validate (see Chatfield 1989 for full discussion), we are also using two innovative approaches adapted specifically for climatological analyses by our statistical expert, Dr. G. Chen. Our nonparametric approach avoids commonly violated assumptions about time series, whereas the point-process approach will further allow direct construction of predictive models of drought frequency, duration and intensity.
First, standard time series analyses will be used to quantify the empirical correlation structure of the paleosalinity data and to improve our understanding of both the random fluctuations and the systematic components of paleoclimate. In particular, the time domain analysis (Box & Jenkins,1976; Hipel & Mcleod 1994) will provide a compact description of past climate, confirm the existence of increasing and/or decreasing trends, and provide point-wise predictions that can be used to forecast future trends and extreme salinity. Additionally, frequency domain analysis (Brillinger1975; Priestley 1981; Brockwell & Davis 1991) will be used to determine the pattern of past drought frequency. Second, we will analyze the paleoclimate data using stochastic point-process models to gain more insight into the mechanistic nature of the salinity time series (Cox & Miller 1965;Leadbetter et al. 1983; Cox & Isham 1985; Berman 1992). Although popular in physical studies, the point- process approach is new to paleoecology. Instead of using the observed correlation structure to describe paleoclimate, we will directly model the process by which salinity changes over time(Rodriguez-Iturbe et al. 1987, 1988). In principal, this amounts to directly modelling the factors that lead to formation of the fossil record. Critically, the point-process approach allows us to make probability statements about future droughts or floods in a fixed time period (e.g., 5 yr, 10 yr, etc.)as well as the frequency of drought recurrence. Finally, to avoid problems resulting from unverifiable assumptions of fossil data properties, we will also use nonparametric time series modelling (Tong 1990; Friedman 1991; Lewis & Stevens 1991) specifically adapted for use with environmental and climatic data (Chen et al. 1996). Briefly, the procedure produces an analytic spline model to provide a more robust description of fossil data. In particular, the nonparametric spline model is usually superior at describing extremes of drought occurrence, and so provide a better measure of drought intensity. Additionally, this nonparametric approach is capable of directly incorporating parallel data (e.g., instrumental data, fossil pigments) to improve model predictions and allow us to account for directional climatic change due to global warming.
Overall program: We will collect sediment cores from each of the three target lakes and nine backup lakes using a variety of coring techniques to assure undisturbed sediments including standard freeze-coring (Leavitt & Findlay 1994), gravity-coring (Hall & Smol 1996), and vibra-coring techniques developed by the Geological Survey of Canada specifically for high-resolution paleoclimatic studies(Last & Vance 1996). Sediments will be photographed, x-rayed, and sectioned in 5 mm intervals to provide temporal resolution of 1-3 yr. Each sample will be analyzed for fossil diatoms and pigments. The age of sediments from each core will be estimated using standard 210Pb (150-200years) and 14C technologies (400-1000 years). Because ancient coal beds often cause over-estimates of sediment age in prairie lakes, we will use 14C accelerator mass spectroscopy techniques to precisely determine ages of discrete sedimentary materials (plant macrofossils, pollen, charcoal). Changes in fossil diatom species assemblages will be used to measure past changes in salinity,conductivity, and pH (Fritz et al. 1991; Cumming et al. 1995; Laird et al. 1996a, 1996b). Point-process, nonparametric and traditional time series analyses will be used to quantify past drought occurrence, frequency and intensity, and to predict future occurrence. Future drought probabilities will be used by crop insurance actuaries to stabilize and reduce premium rates to farmers, and by government agencies and Crown corporations to improve policy and planning for sustainable economic and environmental development in western Canada. Final project results will be submitted for publication as a monograph for use by our user sector and the general public (see CPRC letter of support).
Return to Main Page or Table of
Contents.
ANTICIPATED SIGNIFICANCE OF THE WORK:
This project will measure the intensity,frequency, and duration of droughts over the past 800 years in western Canadian prairies and use these data to predict drought occurrence during the next 5-50 years. Accurate estimates of drought likelihoods are essential for a stable economy and sound resource management in western Canada(Bauer 1988; Wheaton et al. 1992). Improved drought prediction will benefit farmers by providing a sound rationale for selecting agricultural practices (crop, tillage, rotation) that protect against drought impacts (Ripley 1988). Similarly, knowledge of drought probabilities will reduce the uncertainty in crop insurance calculations that lead to unstable, and therefore high, insurance premiums which often persist decades after droughts (P. Johnson, SK Crop Insurance Corp., pers.comm.; S. Crump, AB Agri. Financ. Corp., pers. comm.). Because moisture availability and crop yield are highly correlated (Mack 1988), improved drought prediction will also aid long-range planning by grain marketing and transportation companies, resulting in lower costs to producers and ultimately the general public (Paul Bullock, Canadian Wheat Board, pers. comm.). Similarly, uncertainty in drought occurrence has forced Manitoba Hydro to maintain a $400 million reserve to withstand revenue losses that would accrue during a multi-year drought (H. Surmanski, Generation Systems Studies, Manitoba Hydro, pers. comm.). Improved drought predictions will improve the efficiency of lake and reservoir management, increase economic flexibility, and ultimately reduce costs to the public (Warkentin, 1988). Overall, the project will provide predictive environmental technologies that will directly benefit farmers and crop insurance corporations (Paul Johnson, SKCrop Insurance Corp., pers. comm.), drought relief agencies (Ted O'Brien, PFRA Regina, pers.comm.), and agricultural planning agencies (Paul Bullock, Dir., Weather and Crop Surveillance, Canadian Wheat Board, pers. comm.) and power generation sectors (H. Surminski, Generation System Studies, Manitoba Hydro, pers. comm.). If incorporated into premium determinations, the effect of this project could be felt within two years of project completion and should continue at least ten further years. Based on 1990's averages, crop insurance pay outs arising from drought will have totalled over $1.8 billion during this period.
Return to Main Page or Table of
Contents.
RESEARCH PERSONNEL:
Leavitt heads the Limnology Laboratory at the University of Regina, a group of seven graduate students, postdocs and other scientists who use paleoecology and limnology to study environmental change in Canada and overseas. The Limnology Laboratory is the sole Canadian laboratory dedicated to the analysis of fossil pigments in lake sediments, a mainstay of the present proposal. Regina's state-of-the-art facilities include walk-in cold rooms and freezers for sample processing, a high-capacity freeze-drier, a climate-controlled analytical room, light and electron microscopes,and a computer-driven liquid chromatography system for pigment analyses. Leavitt has authored over 20 articles and 50 public and professional presentations on the use of fossils in environmental and management studies, including six publications on climatic change. He has conducted research on paleolimnological environmental assessment on four continents with over 20 national and international collaborators (see PDF100). Following an NSERC fellowship with Prof. D. Schindler (1991-1993, U. Alberta), Leavitt accepted the post of Assistant Professor of Biology at the University of Regina. Since then, he has won the 1995 Kalium award for 'outstanding professor in the Faculty of Science or Engineering' at Regina, was granted tenure after 3 years, and has been successful in 4 NSERC grant competitions.
Research Associate Dr. Roland Hall is a former NSERC scholar (PGS, PDF) with Dr. P. Dillon FRSC and is a diatom and paleostatistics expert. Hall has authored 10 peer-reviewed publications on the use of fossil diatoms and weighted-averaging diatom inference models, including a co-authored book on the relationship between diatoms and lake salinity (Cumming et al. 1995). Hall's expertise has led to invited presentations to the American Society of Limnology and Oceanography and the Canadian Association of Geographers. Hall trained under Dr. J.P. Smol FRSC at the Queen's University Paleoecological Environmental Assessment and Research Lab (PEARL). Along with other PEARL members, Hall was co-recipient of the 1993 North American Lake Management Society prize for "outstanding research in lake restoration, protection and management".
Dr. Gemai Chen is an expert in applied statistics and actuarial science and forms the pivotal link between the climatic reconstructions and the user sector. Following his PhD from Simon Fraser University in 1991, Chen was hired by the Statistical Consulting Unit at the University of Waterloo and over 3 years provided statistical expertise to over 30 projects in business, engineering, social,natural, and medical sciences. He has wide experience with climatic time series (see PDF100). Chen became an Assistant Professor of Statistics at the University of Regina in August 1994 and was granted tenure after 2.5 yr. His ability to develop and apply high-level statistical methods resulted in Chen being made a member of the Statistical Consulting Committee of the Statistical Society of Canada (1995-1996). Time series analysis of environmental change now forms the main thrust of Chen's research program and has led to the development of innovative non-parametric and point-process time-series modelling techniques integral to the quantitative prediction of droughts (e.g.,Chen et al. 1996).
Return to Main Page or Table of
Contents.
TRAINING OF HIGH QUALITY PERSONNEL:
This project will train three PhD students in modern paleolimnological techniques, statistical analyses of time series, and studies of the economic impact of drought on agriculture and resource management. In addition, this project will train Dr. Hall in novel time series analyses of fossil records and further develop his expertise with algal indicators of environmental change. Finally, this project will train a research technician in modern paleoenvironmental techniques. All eight previous technicians hired by Leavitt have gone on to graduate studies, two of them receiving NSERC scholarships (see Leavitt PDF100).
Return to Main Page or Table of
Contents
PROJECT MANAGEMENT AND COLLABORATION:
Drs. Leavitt and Chen will provide scientific leadership, direct research operations, and co-ordinate interactions between labs and user sector. In addition, Leavitt will direct the research of two PhDstudents who will complete analyses of fossil diatoms (2 lakes) and pigments (3 lakes), while Chenwill supervise one PhD student who will complete the time series analyses and drought predictions. Biweekly meetings between Leavitt's and Chen's labs will be used to coordinate all research activities. Research Associate Dr. Hall will co-ordinate core collection (with Dr. Don Lemmen,GSC-Calgary), analyze diatoms from the SK lake, assure uniform and verifiable diatom taxonomy(QA/QC), and train one PhD student in diatom taxonomy. Hall will also perform all salinity reconstructions, and maintain the project database, Website, and user-group information services in association with the 1/2-time research technician.
A hallmark of this project is the close collaboration and communication among University research personnel and the diverse user sector. All activities will be made available to the user group and public using a World Wide Web site dedicated to this project and maintained at the University of Regina. The site will be updated quarterly, and news releases made available to specialized (The Western Producer, PFRA newsletters) and public print media (Regina Leader Post, Winnipeg Free Press, Edmonton Journal, etc.). Communication among team members will be maintained via regular phone contacts and as well as email lists. In addition, we will hold three workshops in Regina to solicit inputs from the user sector and to help broadcast our results. Workshops will beheld in June 1998 to select sites for climatic reconstructions (Milestone 1), in Aug. 1999 to present results from the Saskatchewan site (Milestone 6), and in Sept. 2000 to present the final report on past and future prairie droughts (Milestone 9).
Detailed interactions have been planned among University and user members. Drs. Hall and Chenwill travel to the respective provincial Crop Insurance Corporations, Canadian Wheat Board, PFRA,and provincial agriculture agencies to help compile historical information on past droughts, crop yields and climate, and to plan time series analyses. We will also help each user group apply our findings to improve their drought planning capabilities. Dr. Lemmen (Geol. Survey of Can.) will provide additional equipment and technical expertise during collection of master cores. Parks Canada will provide logistical support for the field team, as well as local and regional climate data. Dr. Murkin (Ducks Unlimited) and Dr. Bob Clark (CWS) will provide lake-level and wildlife data. Officials at SK Water and MB Hydro will provide long-term data on hydrologic processes. Drs.Herrington and Hunter (Env. Can. Atmos. Serv.) play key roles in this project by providing instrumental climate data from western Canada, as well as expertise in climate data interpretation. Mr. Abouguendia (Canada-SK Grazing & Pasture Technology Program) will advise us on implementation of our results for ranch and farm planning, and will disseminate project information to the cattle ranchers. Finally, the Canadian Plains Research Center will provide logistic support for our workshops and substantial editorial contributions to the production of a summary monograph.
Return to Table of Contents or Give Us Your Comments
See References