7 |
Long-term Ecological Research - A Pedological Perspective |
| DARWIN W. ANDERSON | |
| Saskatchewan Institute of Pedology, University of Saskatchewan, Saskatoon, Canada S7N 0W0 |
Paul B. Sears (1935), in his work Deserts on the March, wrote of the just-developing field of ecology and ecologists:
His [an ecologist's] work involves analysis, of course, but only as a means to a final synthesis and interpretation. When he enters a forest or a meadow he sees not merely what is there, but what is happening there. To him there is afforded a glimpse of continuity, integration and destiny which is indispensable to management and control in any real sense.
Despite its being written more than 50 years ago, amid the spectre of the dust-bowl that had ravaged agriculture and the land resources of the Great Plains, this message is as appropriate today as it was then.
Soil scientists may be criticized in many ways for not fulfilling this vision of the ecologist as they pursued pragmatic and generally successful (at least temporarily) solutions to problems of crop production (Nikiforoff, 1959), or assembled the descriptive data required to characterize the world's soils. Soil scientists, however, have initiated long-term studies, mainly dealing with crop rotations and the monitoring of soil erosion, that have provided valuable material to those working to synthesize, organize, and predict in today's environment. Earlier practical approaches to soil mapping (Ellis, 1932) that recognized soils at different levels of detail have elements of hierarchical theory which are consistent with more refined treatments of today. Other studies have monitored systems which are highly variable both in time and space, such as saline soils, with some success. The objective of this chapter is to discuss briefly some principles that are appropriate to studying ecosystems, and illustrate their application (or their being ignored) in several soils-related, long-term studies, and to present some current ideas on monitoring, understanding, and modeling mainly agricultural ecosystems in the Great Plains region of the United States and Canada.
There is little doubt that soil scientists, in common with scientists in general, have the technology to measure accurately a multitude of properties and to evaluate processes on a sample experimental unit or point location. The key question is how to extend or extrapolate the findings to the remainder of the universe for which they are appropriate, and not to impose them where they are not. Soil scientists, through the carrying out of soil surveys, have been attempting to do that for several decades. Interestingly, an early soil-classification system for use in soil survey has many elements of hierarchical theory, despite the system's development as a pragmatic solution to soil survey and land management problems (Ellis, 1932). The basic class dealing with soil individuals was an associate, or a particular combination of soil horizons and accessory features, especially drainage, on a specific kind of geological material. Associates were organized at the next higher level to form associations, basically the various interrelated soil associates within a landscape. An associate, by necessity, could occur in more than one association and was a real and functioning portion of the association; in this manner, this classification differs from taxonomic systems in that higher integrative levels represented real bodies, not constructs in the minds of the classifier. Associations, in turn, were organized into combinations based on major physiographic and landform properties, resulting in an approach consistent with recent ideas on landscape (Swanson et al., 1988). The most general level or soil zone integrated the regional effects of climate and vegetation, but grouped all soil within a region. This kind of hierarchical system or variants thereof have been and are being used as valuable elements in land management.
Some soil maps have been criticized as not realistically describing the soils of particular areas. Much of this criticism is well founded, and can usually be traced to methods of generalization rather than aggregation as scale decreases, coupled with overuse of typical or modal individuals to represent large areas where this individual may or may not occur, and a too-rigid application of taxonomic limits when locating soil boundaries on a map (Witty and Arnold, 1987). Soil maps are most useful when the complexity of the soil is organized with the paradigm of hierarchical systems. These systems have elements of both spatial and functional scale, with lower levels representing units that are relatively small with a quick time response, and high levels representing units that are larger and slower (Urban et al., 1987).
Soil systems represent this well. A particular lowland may have saline or sodic soils because of regional piezometric systems that move soluble components back to the surface (Henry et al., 1985) or limit the leaching of weathering products (Keller and van der Kamp, 1988). Water in large, regional groundwater flow systems may have residence times of years to centuries and millennia (van der Heijde, 1988). Within the lowlands, however, not all soils will be saline. Saline soils occur mainly on slight convexities, where runoff limits the water available for leaching, with non-saline soils in the effectively more moist concave areas that receive the runoff (Sandoval et al., 1964). Certain layers of the soil, particularly near the surface, may be affected by what happened yesterday, with soluble salts leaching after rains and moving upwards during dry periods.
Some researchers propose that objective and numerical models based mainly on geostatistics are the preferred method of extrapolating point data to larger universes (Kachanoski, 1988). These methods, that describe elements of ecosystems, particularly landscape and soil properties, using techniques such as autocorrelation, spectral analyses, and kriging, permit the determination of spatial relation or pattern of single factors, or groups of related factors. The methods probably are most valuable where the factors occur at random and a pattern is not intuitively evident. These statistical methods also have promise as a test of biased and intuitive methods. A recent study of paired cultivated and native soil areas used systematic sampling along transects and statistical analyses such as correspondence analysis, autocorrelation, and non-parametric methods to evaluate relationships between landscape, soil properties, erosion, and crop yield (Moulin, 1988). The sampling method and analyses indicate the considerable importance to crop yield of calcareous soils on planar, lower slopes, organic matter content and composition, degree of erosion, and other factors. These calcareous, lower slope soils account for only a small proportion of the land, and have usually been ignored in biased or stratified sampling methods.
Several features of ecosystem function and study are related to elements of time. Despite an all-too-frequent view of soils as relatively static bodies that have developed gradually over very long time spans, most soil scientists now view soils as dynamic entities with rapid changes in some properties, and very gradual effects in others. The second concept considers that the present characteristics of a soil represent an integration of a large number of temporally variable processes that may not always move soil formation in the same direction. Many soil properties may be the consequence of episodic events that can initiate a non-linear or threshold response which results in considerable change in properties over short time spans. Soil formation, for example, is in a dynamic balance with soil erosion on uplands, and it is likely that erosion events are episodic, whereas the continued formation of new soil at the weathering front is much more gradual.
Soils, in common with many ecosystems (Rolling, 1985), have variables that operate at slow, intermediate, and fast rates, and it is important to recognize the nature of the variable(s) studied. Variables such as soluble salts are highly dynamic, varying over a season and reaching tentative equilibrium in a few years, whereas organic matter levels have a time dimension of decades to centuries, with carbonate and, in particular, clay weathering having a scale of millennia in semi-arid climates (Anderson, 1977). Interestingly, considering organic matter processes to represent a medium time-scale is, in itself, an integration. At another level of detail, it is possible to differentiate fast (mainly microbial) processes, intermediate components where turnover is dampened by physical sorption to clay, and slow or chemically stabilized humus components in soils (Jenkinson and Rayner, 1977; Anderson, 1979).
Related to time is the concept of threshold-controlled responses or mechanisms. Threshold-controlled responses are considered to be important to the understanding of the development of above-ground ecosystems and geomorphic surfaces, but have only recently been discussed in soil science (Stoner and Ugolini, 1988). A threshold-controlled response occurs when a low-frequency, high-intensity impact or combination of impacts impinge on a system, resulting in a new course of development for a previously stable system. Stoner and Ugolini (1988) found that a large and rare rainstorm (740 mm over 4 days) initiated an intense or catastrophic leaching event, resulting in a considerable surge in subsurface translocation in Spodosol soils of Alaska. Formerly stable horizons were disrupted, and normally insoluble organo-metallic bodies were dissolved and moved downwards in the profile in association with suspended particulate matter to form unusually deep B horizons enriched in both sesquioxides and humus. Such episodic pulse mechanisms are important in other soils. Translocation in Luvisol or Alfisol soils occurs mainly when unusually heavy precipitation initiates leaching into relatively dry soil (Rowitt and Pawluk, 1985). Solonetzic soils, which combine strongly leached surface horizons with subsoils enriched in soluble salts and sodium, may represent a balance between intense leaching (at snowmelt or following intense storms) through larger pores and fissures, balanced by upward fluxes due to capillary rise in much finer pores (Anderson, 1987).
The highly dynamic and temporal variability of many soil processes indicates the need for continuous and long-term monitoring of soil processes. These studies must recognize the possibility of rare and intense threshold-controlled responses, in that they may be critical to understanding the system.
7.4 LONG- TERM STUDIES OF SOIL SALINITY
Saline soils occur generally throughout the sub-arid, semi-arid, and sub-humid regions of the Great Plains of North America. These soils are a natural phenomenon that are the result of high salt content in soil parent materials containing shales of marine origin, the build-up in the soil of the products of present-day weathering (Keller and van der Kamp, 1988), or the redistribution of soluble components due to groundwater flows on the discharge of salt-carrying water from near-surface or glacial aquifers and deep bedrock aquifers (Henry et al., 1985).
Several studies have examined the effect of agriculture on the areal extent and severity of saline soils. Ballantyne (1963), in assessing an apparent increase in saline land during a wet period, the 1950s, in sub-humid south-eastern Saskatchewan, presented evidence for increasing salinity in lower slope soils; the most strongly affected soils occurred on planar, very gentle slopes slightly above and adjacent to depressions that held temporary ponds during wet periods. The saline land had no growth of the intended wheat crop, and salt concentrations that increased towards the soil surface. Ballantyne (1963) considered that the build-up of salts in lower slopes represented a downslope movement of salts within the soil, resulting from the saturation of the upper slope soils. The extra water entering the upper slope soils was considered a result of four years of much-above-average precipitation.
A study in the semi-arid region of southern Alberta reported increasing salinity in agricultural regions, with Solonetzic soils in lower areas most strongly affected in that salts had moved upwards from a normally saline C horizon to re-salinize B and A horizons. Formerly non-saline Dark Brown soils on gentle slopes had also become more saline. The increase in salinity evident in comparing aerial photographs from 1951 and 1962 was attributed to an increased incidence of wet years with greater than the mean precipitation, but implicated the practice of bare fallow and land-use changes on adjacent rangeland.
A more comprehensive assessment involving annual sampling of 64 sites in Saskatchewan (Ballantyne, 1978) indicated the highly variable and dynamic nature of soil salinity. There were always changes in salt concentration for individual soil profiles that were opposite to the average change in the area. Yearly variations were not related to any single factor such as cropping practice, topography, or type of soil profile. The largest annual changes in salinity (30% increase in soluble salts) occurred in soils under low knolls in bare fallow fields.
Despite these studies that indicate the complexity and dynamic nature of soil salinity, the most often quoted estimates of increases in salinity due to agriculture were based on comparisons of aerial photographs of the same areas taken ten years apart (van der Pluym et al., 1981). Estimates of soil salinity were based on visible evidence (light-colored areas) or restricted vegetative growth, and indicated that salinity was increasing at a rate of 10% annually. The total area of land that is affected is not known accurately, with estimates for the prairie region of Canada ranging from about 2 million to 5.4 million ha. Combining the 10% rate of increase with 5.4 million ha gives an annual rate of increase which is probably at least an order of magnitude too high, and indicates the need for long-term studies to monitor soil salinity and the factors that control its occurrence. The temporally variable nature of soil salinity, and assessments that are based mainly on proxy data such as the severity of crop growth restriction, make accurate assessments of areal extent difficult and generally unreliable.
This study involved the monitoring of saline soils in an area where predicted and actual increases in artesian pressure or piezometric level of the groundwater had occurred. The study was situated in a local basin, with a semi-arid climate and Brown (Aridic Haploboroll) soils occurring on mixed glacial till, glaciofluvial, and alluvial deposits. The study was initiated because the filling of a reservoir constructed to provide cooling water for a thermal power station had resulted in a consistent rise of 2-4 m in the piezometric head within land just below the reservoir (Figure 7.1 ). The response was most rapid on the groundwater directly connected to the reservoir through the Empress gravel, but gradually affected other stratigraphic units, including the surficial glacial till deposits (Figure 7.2).
Soil salinity was monitored by sampling in early August, at marked sites at seven different locations, with 26 individual sites in total. There were variable changes in soil salinity in the 1979 to 1983 period. There was no change or a slight decrease in salinity in the 0- to 30-cm depth at 12 sites, a variable response at six, and a gradual increase in salinity at eight sites. Results for a silty alluvial soil with a piezometric head at ground level and a water table at less than 2 m depth indicate a marked increase in salinity as shown by increases in the electrical conductivity (EC) of saturation extracts of the soil (Figure 7.3(a)) and the soluble sodium concentration (Figure 7 .3(b)). EC and soluble sodium values were highest in surface layers, and generally decreased with depth, indicating a pronounced upward flux of soluble constituents that is driven by artesian pressure and evaporation of soil water at the surface.
Other sites, usually Solonetzic soils, that had little or no salts in surface layers but saline subsoils had salt move upwards, particularly in 1980 and 1981 when summer precipitation was well below average. These salts were flushed out of the soil in 1982 by several heavy rains (Figure 7.4 ). Sampling in 1983 showed that the salts were moving back towards surface layers. These kinds of dynamic fluctuations were evident at several sites, particularly those with low to moderate concentrations of salts, and indicate the dynamic nature of salt movement in soil. Upward fluxes of salts were greatest in silty soils with high water contents, consistent with earlier observations of upward movement of sodium by convection (carried along by moving water) and diffusion or moving within the soil solution in response to a concentration gradient (Merrill et al., 1983).
Figure 7.1 Generalized hydrogeology and salinity at the Coronach soil salinity study area
Figure 7.2 Changes in the piezometric level of main aquifers following the filling of a reservoir upstream.
Figure 7.3 Changes in (a) soil salinity and (b) soluble sodium in a saline Brown soil with time
Figure 7.4 Changes in (a) soil salinity and (b) soluble sodium in a saline Brown soil with time
Aerial photographs taken each year within a few days of 10 July indicate the gradual but highly variable increase in saline soils, particularly in silty and sandy materials at low elevations within the basin (Figure 7.5). Those soils that were most severely affected include Solonetzic soils, which are soils that normally have saline subsoils because of upward fluxes of soluble constituents due to artesian pressure. Rego or A/C soil profiles, where strong accumulations of calcium carbonate and, occasionally, soluble salts are considered to be indicative of artesian influences in the past, were also strongly salinized. Field checking, mainly in areas of limited crop growth, using a hand-portable, non-contacting conductivity meter (EM38, Geonics Ltd) indicated that previously non-saline Orthic Brown soils that had developed under moderate leaching regimes were becoming saline by about 1983 to 1988. The soils that were most strongly affected had sandy to silty clay loam textures, and occurred on slight convexities or very gentle slopes at low elevation, similar to those observed by Ballantyne (1978). Soils in concave areas, despite their occurring at slightly lower elevations and with supposedly stronger artesian fluxes, often were much less saline. These differences, similar to those observed by Sandoval et al. ( 1964 ), were attributed to runoff from sloping or convex lands that results in a reduced leaching potential, coupled with enhanced leaching in runoff-collecting concave areas, that counterbalanced upward fluxes due to artesian pressure.
Figure 7.5 Color infra-red aerial photographs of the Coronach salinity study area for 1982, 1984, 1986 and 1988. The black line indicates the extent of saline soils as determined with the EM salinity meter
The results of this study are not directly applicable to other areas as most land has not had the increase in artesian pressure that occurred in the study area. There are, however, principles that can be applied to long-term studies of soil salinity. At the start of this study, the investigators had a more mechanistic view of salinity which considered that salinity in the area was due to artesian pressure, that it would uniformly worsen as pressure increased, and that the process would be unidirectional and gradual. This work, despite its shortcomings, has indicated the highly complex and dynamic nature of salinity and the factors that control it.
The dynamic nature of soil salinity indicates that monitoring salinity at different depths of the soil requires more than annual sampling and careful records of precipitation at the monitoring sites. Sampling the soil by removing cores will alter the permeability of the soil, and eventually the salt concentration. The best technique may be to insert electrodes to measure electrical conductivity in situ, rather than removing soil cores. The non-contacting, portable salinity meters are of limited value, but could be used if soil moisture was also monitored.
A conceptual model that describes the factors and indicates processes (Figure 7.6) is based partly on our experience in the area. This model, that considers landscape and regional hydrological processes that are relatively slow, and changes in the soil that may be rapid, is presented as an initial step to modeling soil salinity. Sub models will be required to describe processes at greater detail, such as models for salt flow (Childs and Hanks, 1975), salt-transport and chemical equilibria (Robbins et al., 1980), and upward fluxes by convection or diffusion (Merrill et al., 1983). The basic approach, however, is to key on the processes critical to change, namely the relative balance between downward flux (leaching) and upward flux (capillary rise or artesian pressure), in order to predict the impacts of agricultural practices, including irrigation, and long-term effects due to change in climate. Changes in climate due to increased greenhouse gases that result in more moist winters and drier, warmer summers may result in marked changes in salinity. Recharge to the groundwater is mainly depression-focused and could be enhanced by the accumulation of snowmelt in depressions, resulting in higher artesian pressure after some years or decades. Higher rates of evapo-transpiration and less rainfall in summer would decrease downward fluxes and slightly increase upward fluxes near the surface, resulting in an increase in saline land over time. Evaluations of the effects of changes in artesian pressure on soil salinity will be dependent on long-term records of piezometric levels in major aquifers. This kind of record is available in Saskatchewan, a result of far-thinking hydrologists who began keeping systematic records more than 25 years ago (Meneley et al., 1979).
Figure 7.6 A conceptual model describing the factors and processes that affect soil salinity
The soil science research that has been ongoing for long periods of time includes crop rotation studies. The studies at Rothamsted were begun more than a century ago, and records, soil analyses, and soil samples have been used in recent times to evaluate processes such as the turnover of organic matter (Jenkinson and Rayner, 1977; Johnston, Chapter 6, this volume). The designation of rotation studies as ecological research may be questionable in that the studies included monocultures under highly managed conditions in which response rather than process was measured. However, several aspects of these studies indicate their value to discerning gradual and long-term changes to soils.
Of several long-term rotations in Western Canada, three will be discussed in this chapter. One study, which was made to compare the effects of green manures and farmyard manures on yield and soil properties, was begun at the University of Manitoba in 1919 on clayey Black soils (Poyser et al., 1957), and has more detailed soil sampling and yield records available from 1930 until the late 1960s. The records show organic carbon and nitrogen in the soil decreased steadily from 1930 to 1955, with smaller losses on rotations where manure was added or legumes grown. Crops grown following manures generally yielded higher than crops given other treatments, with interactions between enhanced nitrogen supply and the moisture-depleting effects of green manures showing up as reduced yields after green manures during dry periods. A later evaluation showed the organic matter-depleting effects of more frequent bare fallow, and indicated that soil fertility and yields could be sustained or increased over time in continuous cereal rotations, provided that adequate manure or chemical fertilizer was applied (Ridley and Hedlin, 1968). These rotations were discontinued in about 1970, partly because the rotations being studied were no longer considered pertinent to Manitoba agriculture, and the land was required for athletic fields. Since many of the long-term effects on soils were not evaluated, the value of the studies was diminished.
Continuous wheat and fallow-wheat rotations were begun in 1912 on Dark Brown, loamy soils at the Agriculture Canada Research Station at Lethbridge, Alberta. Recent soil analyses have provided data that furthers our understanding of the effects of bare fallow, fertilizer, and crop rotation on soil organic matter (Janzen, 1987). The application of inorganic nitrogen fertilizer in continuous wheat or wheat-wheat-fallow rotation significantly increased organic carbon and nitrogen in the soil, with an even more marked increase in labile or mineralizable forms of carbon and nitrogen. This effect was attributed to increased production where nitrogen was applied, resulting in greater return of residues to the soil, greater microbial activity, and more efficient organic cycling of nitrogen. Applying ammonium nitrate fertilizer at 45 kg N ha-1, however, decreased the pH of surface soils from 7.2 to 6.9, indicating the potential for acidification problems on less buffered and more acidic soils.
A five-year rotation that included forage and cereal and a two-year wheat-fallow rotation were set up at Breton, Alberta, in 1929. Soils are Gray Luvisols on loamy glacial till, with a pH of about 6.0 and an organic carbon content of 1.2% (McGill et al., 1986). Organic carbon and nitrogen contents of the Ap horizon have increased in rotations containing forages, and yields of cereal grains are higher following forages. These soils have surface horizons that form compact crusts upon drying after heavy rains and are difficult to till; soil tilth was improved as indicated by larger and more stable soil aggregates in the forage rotations (Toogood and Lynch, 1959). Chemical fertilizers, mainly ammonium sulphate, ammonium nitrate, and ammonium phosphate compounds, were applied at rates to supply 11 kg N ha-l and 9 kg S ha-l annually. Despite a natural gradient in pH across the systematically arranged plots, detailed evaluations indicated that the chemical fertilizers had increased acidity and easily extractable forms of aluminium and manganese, and decreased exchangeable calcium (McCoy and Webster, 1977). Manures had no effect on soil reaction. The decrease in pH from approximately 6.0 to about 5.2 resulted in reduced proportions of legumes in forage stands and yields, and was remedied by applying lime to the soil.
A later study examined both the short-term dynamics of microbial biomass and the longer-term processes affecting organic matter (McGill et al., 1986). The five-year cereal-forage rotation contained 38% more total nitrogen, but 117% more microbial nitrogen, than the wheat-fallow soils. Manured plots had twice the microbial nitrogen of chemically fertilized or control plots. The average quantity of biomass appears to be controlled by long-term additions of carbon, whereas seasonal fluctuation in environmental conditions, particularly moisture, appears to control short-term biomass dynamics. Average turnover rates of the biomass were 0.2 to 3.9 years, being 1.5 to 2 times faster in the two-year wheat-fallow rotation. Average carbon additions are insufficient to account for annual turnover in the two-year rotation, indicating that native soil organic matter may still be decreasing. These studies use modern-day methods but are made possible by the legacy of experimental material initiated more than five decades ago.
The rotation studies can be criticized because they were systematically rather than randomly arranged, thereby limiting statistical analysis (Ridley and Hedlin 1968), because soil samples representative of initial conditions often were not taken, and because response rather than process was measured. In addition, it is difficult to allocate research funds to continue rotations that are no longer pertinent to the region. Despite these limitations, these rotation studies have provided much of the data that evaluate long-term trends, and have provided stored soil samples that can be analyzed by modern methods (Jenkinson and Rayner, 1977; Anderson and Paul, 1984). Many of the findings are currently being used to develop and validate simulation models that describe long-term processes such as organic matter turnover. In other instances, the measurement of surprises, such as the acidification of soil by even low additions of nitrogen fertilizer to poorly buffered soils, has provided findings the originators had not envisaged.
The Great Plains region of Canada includes the former grasslands that are now mostly converted to farmland, and the boreal forest and wetlands of the northern portion. Both areas require long-term ecological study to address problems of both local and global perspectives.
The sustainability of agriculture is the question of greatest concern in the prairie region, with a perception of increasing soil deterioration due to erosion, loss of organic matter and fertility, and increasing salinization. Questions related to the severity and extent of soil deterioration and current economic costs to farmers and society have not been adequately resolved and require more study (van Kooten and Furtan, 1987). There is, however, a consensus that soil deterioration is a serious problem, and concerted action is required to address the problem. Knowledge of the effect on soil quality of recent, conservation-oriented tillage systems such as stubble-mulch tillage is lacking, although studies of farmers ' fields in North Dakota (Bauer and Black, 1981) and the aforementioned rotation studies (Janzen, 1987) suggested that soil quality is not deteriorating under good crop management. Climatic change due to the greenhouse effect may make agriculture impossible in the prairie region and extend the agricultural frontier in the north, and may have complex effects on conditions such as soil salinity, as discussed earlier in this chapter.
Determining long-term changes in soils will require a number of carefully selected and adequately sampled sites, with annual monitoring of climate and land management. These monitoring sites should be established as soon as possible, to provide some data in the medium term on processes such as organic matter dynamics, salinization, and erosion. The real value, however, of these sites may be decades in the future, when they will provide the baseline data required to evaluate the changes and surprises that undoubtedly will occur.
The area of concern is approximately 500 000 km2, and it certainly will not be possible to locate long-term study sites on all combinations of land, climate, and agricultural use. Simulation models are envisaged as the means to address change and potential change in these areas. The models will build on the findings gathered by long-term rotation studies that were discussed earlier, utilize research that has examined soil development and response using an environmental gradient approach, and be extrapolated to the region within a hierarchical paradigm based mainly on soil and landscape maps. Models may be tested by comparing simulations to present conditions in soils that have been managed under known rotations by fam1ers. An example of such a model is the Century model, that simulates the effects of climate and texture on production and organic matter quality and quantity (Parton et al., 1987). This model integrates three sub models: soil and decomposition, nitrogen, and plant production. The variables required include climate (temperature and moisture), soil texture, and nature of plant residues. This model was developed based on the findings of soils along environmental gradients (Anderson 1979), and relies on the results of a wide variety of studies, both basic and more applied.
Other models will be required. Current models of soil erosion, such as the Universal Soil Loss Equation, for example, do not work well in Western Canada because of differential effects of both wind and water erosion on soils within a landscape (Kiss et al., 1986) and the complex nature of landscapes with both erosion and deposition occurring in the same field (Johnson, 1988). Current efforts are to relate erosion losses to landform and the composition of soils. Erosion is estimated by determining the current amounts of 137Cs in soils and relating that amount to the background levels. The 137Cs in the soil resulted mainly from atmospheric testing of nuclear weapons about 1960. These studies have identified the kinds of landscapes most susceptible to erosion, the landscape regimes where erosion is most severe, and areas of deposition (Martz and de Jong, 1987; Pennock and de Jong, 1987). The objectives of this research are empirically based models that can be used to rate various combinations of landform, soils, and climate for susceptibility to erosion under various management alternatives, and to predict the effects of conservation practices on erosion. Soil salinity models generally are not available, but may be possible by integrating hydrological, weathering, leaching, and landscape sub-models as depicted earlier (Figure 7.6).
The Boreal Forest is of considerable importance, not only because of possible impacts due to acidic precipitation or climatic change, but because of its significance to the global atmosphere, particularly methane additions (Matthews and Fung, 1987). Northern wetlands appear to contribute significantly to methane generation, and the strong recovery of beaver (Castor canadensis) population in recent decades may be an important factor in the increase (Nisbet, 1988). Wetlands altered by beaver activity produce orders of magnitude more methane than beaver-free wetlands. The marked increase in beaver population is indicated (Figures 7.7 and 7.8) by the construction of 11 beaver ponds along a stream in the Boreal Forest where no ponds existed in 1950 and three existed in 1960. Each beaver pond forms a wetland of at least one hectare, and influences land upstream by reducing drainage.
Figure 7.7 Aerial photograph showing a stream with no evident beaver wetlands in 1950 (upper photo) and three wetlands in 1960 (lower photo)
A significant drying of the climate in Boreal regions could result in a lowering of the watertable in wetlands and substantially increased rates of aerobic decomposition to release CO2, coupled with reduced methane evolution (Moore and Knowles, 1989). The significance of these kinds of wetlands, plus the bogs and fens of this region, to methane and CO2 cycles is important. The annual flux of methane was estimated at 0.1 to 0.6 g C m-2 which, although low compared to other wetlands, becomes substantial because of the large area of subarctic peatlands (Moore and Knowles, 1987).
Other questions of concern are the rates of sulphur emission from anaerobic wetlands and from plants under saline or high sulphur stress. The sulphur cycle of Boreal regions is incompletely understood. Most sulphur is retained by Luvisol soils in organic form, mainly in organic surface layers but also as organic sulphates that appear to have leached into subsoils (Schoenau and Bettany, 1987). The acidic but only moderately weathered Luvisol soils contain low total amounts of sulphur, indicating somewhat inefficient storage mechanisms for sulphur received in precipitation. Some well-humified peats within concave areas in lands dominated by sandy soils have very high sulphur contents, and much sulphur in relation to carbon. This anomalously high concentration of sulphur indicates that these bogs may be a sink for sulphur within those landscapes. Similarly, estimates of losses of phosphorus during the formation of Luvisolic soils are generally larger than the amount of phosphorus in surface waters (St Arnaud et al., 1988). This suggests that phosphorus lost from mineral soils may be retained in the organic soils of wetlands.
Figure 7.8 An aerial photograph taken in 1985, showing the same stream as in Figure 7.7, with eleven beaver wetlands evident
This chapter has dealt mainly with long-term studies of the nature of soil, with no review of the several ecological research efforts in which soil scientists have participated with other scientists. The long-term crop rotation studies have provided key data beyond the original and practical objectives of the scientists who established them, and still remain a valuable resource for those requiring records, analyses, or samples that were taken decades ago. Improved data to determine the direction, severity, and extent of soil-deteriorating processes such as salinization, erosion, and decline in soil organic matter are required so that research efforts can be appropriately directed to solve the problems of most serious long-term concern. Some long-term monitoring sites are required, but the main impetus should be to build on those data already available from detailed studies, to construct simulation models that key on the variables most critical to long-term processes within ecosystems, and to extrapolate these models to larger universes within a hierarchical paradigm, based mainly on soil, landscape, and climatic maps.
The author gratefully acknowledges the Saskatchewan Power Corporation for allowing use of data from the Coronach study, and D.R. Cameron for his contribution to the hydrogeology at Coronach. Careful reviews of the manuscript by D.F. Acton and J.R. Bettany are appreciated.
Anderson, D.W. (1977). Early stages of soil formation on glacial till mine spoils in a semi-arid climate. Geoderma, 19, 11-19.
Anderson, D.W. (1979). Processes of humus formation and transformation in soils of the Canadian Great Plains. J. Soil Sci., 30, 77-84.
Anderson, D.W .( 1987). Pedogenesis in the grassland and adjacent forests of the Great Plains. Adv. Soil Sci., 7, 53-93.
Anderson, D.W. and Paul, E.A. (1984). Organo-mineral complexes and their study by radiocarbon dating. Soil Sci. Soc. Am. J., 48, 298-301.
Ballantyne, A.K. (1963). Recent accumulation of salts in the soils of southeastern Saskatchewan. Can. J. Soil Sci., 43, 52-58.
Ballantyne, A.K. ( 1978). Movement of salts in agricultural soils of Saskatchewan 1964 to 1975. Can. J. Soil Sci., 58, 501-509.
Bauer, A. and Black, A.L. (1981). Soil carbon, nitrogen and bulk density comparision in two cropland tillage systems after 25 years, and in virgin grassland. Soil Sci. Soc. Am. J., 45, 1166-1170.
Childs, S.W. and Ranks, R.J. (1975). Model of salinity effects on crop growth. Soil Sci. Soc. Am. Proc., 39, 617-622.
Ellis, J.R. (1932). A field classification of soils for use in the soil survey. Sci. Agric., 12, 338-345.
Henry, J.L., Bullock, P.R., Rogg, T.J. and Luba, L.J. (1985). Groundwater discharge from glacial and bedrock aquifers as a soil salinization factor in Saskatchewan. Can. J. Soil Sci., 65, 749-768.
Holling, C.S. (1985). Simplifying the complex: The paradigms of ecological function and structure. European J. Oper. Res., 30, 139-146.
Howitt, R.W. and Pawluk, S. (1985). The genesis of a Gray Luvisol within the Boreal Forest region. II. Dynamic pedology. Can. J. Soil Sci., 65, 9-19.
Janzen, R.R. (1987). Effect of fertilizer on soil productivity in long-term wheat rotations. Can. J. Soil Sci., 67, 165-174.
Jenkinson, D.S. and Rayner, J.R. (1977). The turnover of soil organic matter in some of the Rothamsted classical experiments. Soil Sci., 123, 298-305.
Johnson, R.R. (1988). Putting soil movement into perspective. J. Prod. Agric., 1, 5-12.
Kachanoski, R.G. (1988). Processes in soils - from pedon to landscape. In Rosswall, T., Woodmansee, R.G. and Risser, P.G. (Eds) Scales and Global Change, SCOPE 35. John Wiley, Chichester, 153-178.
Keller, C.K. and van der Kamp, G. (1988). Rydrogeology of two Saskatchewan tills, II. Occurrence of sulfate and implications for soil salinity. J. Hydrol., 101, 123-144.
Kiss, J.J., de Jong, E. and Rostad, H.P.W. (1986). An assessment of soil erosion in west-central Saskatchewan using cesium-137. Can. J. Soil Sci., 66, 581-590.
Martz, L.W. and de Jong, E. (1987). Using cesiurn-137 to assess the variability of net soil erosion and its association with topography in a Canadian prairie landscape. Catena, 14, 439-451.
Matthews, E. and Fung, I. (1987). Methane emission from natural wetlands: Global distribution, area, and environmental characteristics of sources. Global Biochemical Cycles, 1, 61-86.
McCoy, D.A. and Webster, G .R. ( 1977). Acidification of a Luvisolic soil caused by low-rate, long-tern applications of fertilizers and its effects on growth of alfalfa. Can. J. Soil Sci., 57, 119-127.
McGill, W. B., Cameron, K.R., Robertson, J.A. and Cook, F.D. (1986). Dynamics of soil microbial biomass and water soluble organic C in Breton L after 50 years of cropping to two rotations. Can. J. Soil Sci., 66, 1-20.
Meneley, W.A., Maathuis, R., Jaworski, E.J. and Allen, V.F. (1979). SRC Observation wells in Saskatchewan, Canada. Saskatchewan Research Council Report No. 19, Saskatchewan Research Council, Saskatoon, Canada.
Merrill, S.D., Doering, E.J., Power, J.F. and Sandoval, F.M. (1983). Sodium movement in soil-minespoil profiles: Difussion and convection. Soil Sci., 136, 308-331.
Moore, T.R. and Knowles, R. (1987). Methane and carbon dioxide evolution from subarctic fens. Can. J. Soil Sci., 67, 77-82.
Moore, T.R. and Knowles, R. (1989). The influence of water table levels on methane and carbon dioxide emissions from peatlands. Can. J. Soil Sci., 69. In press.
Moulin, A.P. (1988). Spatial variability of soil properties in hummocky terrain as affected by cultivation. PhD Thesis, University of Saskatchewan, Saskatoon, Canada.
Nikiforoff, C.C. (1959). A re-appraisal of the soil. Science, 129, 186-196.
Nisbet, E.G. (1989). Some northern sources of natural atmospheric methane: production, history and future implications. Can. J. Earth Sci., 26, 1603-1611.
Parton, W.J., Schimel, D.S., Cole, C.V. and Ojima, D.S. (1987). Analysis of factors controlling soil organic matter levels in Great Plains Grasslands. Soil Sci. Soc. Am. J., 51,1173-1179.
Pennock, D.J. and de Jong, E. (1987). The influence of slope curvature on soil erosion and deposition in hummocky terrain. Soil Sci., 144, 209-217.
Poyser, E.A., Hedlin, R.A. and Ridley A.O. (1957). The effect of farm and green manures on the fertility of blackearth-meadow clay soils. Can. J. Soil Sci., 37, 48-56.
Ridley, A.O. and Hedlin, R.A. (1968). Soil organic matter and crop yields as influenced by the frequency of summerfallowing. Can. J. Soil Sci., 48, 315-322.
Robbins, C.W., Wagenet, R.J. and Jurinak, J.J. (1980). A combined salt transport-chemical equilibrium model for calcareous and gypsiferous soils. Soil Sci. Soc. Am. J., 44, 1191- 1194.
Sandoval, F.M., Benz, L.C., George, E.J. and Mickelson, R.H. (1964). Microrelief influences in a saline area of Glacial Lake Agassiz: I. On salinity and tree growth. Proc. Soil Sci. Soc. Am., 28, 276-280.
Schoenau, J.J. and Bettany, J.R. (1987). Organic matter leaching as a component of carbon, nitrogen, phosphorus, and sulfur cycles in a forest, grassland, and gleyed soil. Soil Sci. Soc. Am. J., 51,646-651.
Sears, P.B. (1935). Deserts on the March. University of Oklahoma Press, Norman, OK.
St Arnaud, R.J., Stewart, J.W.B. and Frossard, E. (1988). Application of the 'pedogenic index' to soil fertility studies, Saskatchewan. Geoderma, 43, 21-32.
Stoner, M.G. and Ugolini, F.C. (1988). Arctic pedogenesis: 2. Threshold-controlled subsurface leaching episodes. Soil Sci., 145, 46-51.
Swanson, F.J., Kratz, T.K., Caine, N. and Woodmansee, R.G. (1988). Landform effects on ecosystem patterns and processes. BioScience, 38, 92-98.
Toogood, J.A. and Lynch, D.L. (1959). Effect of cropping systems and fertilizers on mean weight diameter of aggregates of Breton Plot soils. Can. J. Soil Sci., 39, 151-156.
Urban, D.L., O'Neill, R.V. and Shugart, H.H., Jr (1987). Landscape ecology. BioScience, 37,119-127.
van der Heijde, P.K.M. (1988). Spatial and temporal scales in groundwater modelling. In Rosswall, T., Woodmansee, R.G. and Risser, P.G. (Eds) Scales and Global Change, Scope 35. John Wiley, Chichester, 195-224.
van der Pluyrn, H.S.A., Paterson, B. and Holm, H.M. (1981). Degradation by salinization. In Agricultural Land: Our Disappearing Heritage. A Symposium, Proc. 18th Alberta Soil Science Workshop.
van Kooten, G.C. and Furtan, W .H. (1987). A review of issues pertaining to soil degradation in Canada. Can. J. Ag. Econ., 35, 33-54.
Witty, J.E. and Arnold, R. W .( 1987). Soil taxonomy: An overview. Outlook on Agriculture, 16, 8-13.
|
|
|
The electronic version of this publication has been
prepared at |
