Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A comparison of the effects of 20 and 30 years of grazing on grassland soil properties in southern British… Evans, Christian Welby 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2011_spring_evans_christian.pdf [ 12.06MB ]
Metadata
JSON: 24-1.0071583.json
JSON-LD: 24-1.0071583-ld.json
RDF/XML (Pretty): 24-1.0071583-rdf.xml
RDF/JSON: 24-1.0071583-rdf.json
Turtle: 24-1.0071583-turtle.txt
N-Triples: 24-1.0071583-rdf-ntriples.txt
Original Record: 24-1.0071583-source.json
Full Text
24-1.0071583-fulltext.txt
Citation
24-1.0071583.ris

Full Text

A Comparison of the Effects of 20 and 30 Years of Grazing on Grassland Soil Properties in Southern British Columbia by Christian Welby Evans B.A., The University of British Columbia, 1998  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in The Faculty of Graduate Studies (Soil Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) January, 2011  © Christian Welby Evans, 2011 i  Abstract Although numerous studies have been conducted on rangeland soil quality in Alberta, Manitoba, Saskatchewan, as well as in the Great Plains, there has been little documentation of the response of soil properties to time of grazing and stocking rate treatments for the grasslands of the southern interior of BC. In the Lac Du Bois range in Kamloops, BC, the current moderate stocking rate of 2 AUM ha -1 was based on a desired available forage use of 50%. Livestock is moved up and down an elevation and productivity gradient over the grazing season so that pastures located midway up the gradient are grazed either in the spring or fall. The objectives of this study were to determine (1) the effects of spring and fall grazing treatments on selected soil properties after periods of 20 and 30 years and (2) the effects of 0 and 2 AUM ha -1 grazing rates on selected soil properties after periods of 20 and 30 years. The spring grazing treatment led to greater soil bulk density, mechanical resistance, pH, as well as lower polysaccharides and CEC relative to the fall grazing treatment. The grazing rate of 2 AUM ha 1  led to greater soil mechanical resistance and pH, as well as lower soil polysaccharides and LOMF  relative to the ungrazed control. After 30 years of grazing, soil bulk density was greater in the 0-7.5-cm depth under the 2 AUM ha-1 treatment relative to the ungrazed exclosure in spring-grazed but not in fall-grazed areas, indicating that this stocking rate, when used for spring grazing, has led to soil compaction. Rangeland managers in the southern interior of BC should consider adjustments of time of grazing and stocking rate recommendations when these have been solely on vegetation responses, and should consider including soil properties in rangeland health assessments.  ii  Table of Contents Abstract......................................................................................................................................................ii Table of Contents......................................................................................................................................iii List of Tables..............................................................................................................................................v List of Figures...........................................................................................................................................vi List of Equations.....................................................................................................................................viii List of Symbols and Abbreviations...........................................................................................................ix Acknowledgements....................................................................................................................................x Chapter 1- General Introduction................................................................................................................1 1.1 British Columbia Rangelands..........................................................................................................1 1.1.1 Rangeland Management..........................................................................................................3 1.1.2 British Columbia Grasslands...................................................................................................5 1.1.3 Rangelands in British Columbia Grasslands...........................................................................9 1.2 Soil Quality......................................................................................................................................9 1.2.1 Soil Quality Concept..............................................................................................................10 1.2.2 Soil Quality Indicators...........................................................................................................14 1.2.3 Soil Quality and Rangeland Health.......................................................................................18 1.3 Grazing Effects..............................................................................................................................20 1.3.1 Grazing Effects on Soil Chemical Quality............................................................................20 1.3.2 Grazing Effects on Soil Biological Quality...........................................................................30 1.3.3 Grazing Effects on Soil Physical Quality..............................................................................32 1.3.4 Grazing Effects on Vegetation...............................................................................................37 1.4 Summary of General Introduction.................................................................................................40 1.5 Study Objectives and Hypotheses.................................................................................................42 Chapter 2- Grazing Effects on Selected Grassland Soil Properties in Southern British Columbia after a Period of 20 and 30 Years........................................................................................................................44 2.1 Introduction...................................................................................................................................44 2.2 Materials and Methods..................................................................................................................46 2.2.1 Site Description.....................................................................................................................46 2.2.2 Sampling and Analyses..........................................................................................................46 2.2.3 Statistical Analyses................................................................................................................49 2.3 Results and Discussion..................................................................................................................49 2.3.1 Grazing Effects on Soil Chemical and Biological Properties...............................................49 iii  2.3.2 Grazing Effects on Soil Physical Properties..........................................................................59 2.4 Summary of Grazing Effects on Soil Physical, Chemical, and Biological Properties..................69 2.5 Conclusions...................................................................................................................................71 Chapter 3- General Conclusions and Recommendations for Future Research........................................72 3.1 General Conclusions......................................................................................................................72 3.2 Recommendations for Future Research........................................................................................74 References................................................................................................................................................76 Appendix A- Correlations........................................................................................................................90 Aggregate MWD and Size Fraction Correlation Matrix....................................................................90 Soil Quality Indicator Correlation Matrix..........................................................................................91 Appendix B- Pasture and Soil Series Descriptions..................................................................................92 Appendix C- ANOVA Table.....................................................................................................................94 Appendix D- Additional Figures..............................................................................................................95 Appendix E- Additional Tables..............................................................................................................100 Appendix F- Plates.................................................................................................................................101  iv  List of Tables Table 1.1: Selected soil properties and supported functions....................................................................15 Table 1.2: Soil properties suggested for inclusion in an MDS.................................................................16 Table 1.3: Soil quality assessment framework for the Lac du Bois rangeland in Southern Interior of BC ..................................................................................................................................................................20 Table 2.1: Grazing effects on cation exchange capacity (CEC) in the 0-7.5-cm depth as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC...............49 Table 2.2: Time of grazing and stocking rate treatment interaction effects on cation exchange capacity (CEC) in the 7.5-15 and 15-30-cm depths as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC...................................................................................50 Table 2.3: Grazing effects on exchangeable cations as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC...................................................51 Table 2.4: Grazing effects on soil reaction as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC...........................................52 Table 2.5: Grazing effects on available phosphorus as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC........................53 Table 2.6: Grazing effects on total soil carbon, total soil nitrogen, and carbon-to-nitrogen ratio as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. ..................................................................................................................................................................55 Table 2.7: Grazing effects on light organic matter fraction (LOMF) as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC...........................................56 Table 2.8: Grazing effects on soil polysaccharides as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC.................................57 Table 2.9: Time of grazing and stocking rate treatment interaction effects on soil bulk density as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC...........................................................................................................60 Table 2.10: Grazing effects on soil bulk density as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC.................................60 Table A.1: Aggregate size fraction correlation coefficients. All data from 0-7.5-cm depth. n=16 for all correlations...............................................................................................................................................91 Table A.2: Correlation matrix for selected soil quality indicators. Data were averaged by exclosure location. n=48 for all correlations. ..........................................................................................................92 Table C.1: ANOVA Table for completely randomized split-plot design with subsampling for assessment of grazing effects on grassland soil quality after 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC...................................................95 Table E.1: Grazing effects on above-ground biomass as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC.................................................101  v  List of Figures Figure 1.1: Total cattle population and breakdown by farm size in British Columbia in selected years...2 Figure 1.2: Occurrence of Grasslands in British Columbia ......................................................................6 Figure 2.1: Root biomass under spring and fall grazing treatments at 0-7.5 cm, 7.5-15 cm, and 15-30 cm as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32).....58 Figure 2.2: Root biomass under 0 and 2 AUM ha-1 treatments at 0-7.5 cm, 7.5-15 cm, and 15-30 cm as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32).....58 Figure 2.3: Soil mechanical resistance under spring and fall grazing treatments as determined 20 years (i.e., 1998) (a) and 30 years (i.e., 2008) (b) after establishment of an experiment on rangeland north of Kamloops, BC. The dotted line represents the critical upper limit for root growth. Error bars represent the standard deviation of the mean (n=32). Significant differences (p<0.10) were marked by an asterisk. ..................................................................................................................................................................62 Figure 2.4: Soil mechanical resistance under ungrazed and grazed treatments as determined 20 years (i.e., 1998) (a) and 30 years (i.e., 2008) (b) after establishment of an experiment on rangeland north of Kamloops, BC. The dotted line represents the critical upper limit for root growth. Error bars represent the standard deviation of the mean (n=32). Significant differences (p<0.10) were marked by an asterisk. ..................................................................................................................................................................63 Figure 2.5: Aggregate mean weight diameter (MWD) under time of grazing (a) and stocking rate (b) treatments as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32). Significant differences (p<0.10) were marked by an asterisk...........................................64 Figure 2.6: 2-6 mm, 1-2 mm, 0.25-1 mm and <0.25 mm aggregate size fractions under spring and fall grazing treatments as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32). Differences were considered significant at p<0.10 and marked by an asterisk.................65 Figure 2.7: 2-6 mm, 1-2 mm, 0.25-1 mm and <0.25 mm aggregate size fractions under ungrazed and grazed treatments as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32).............................................................................................................................................66 Figure 2.8: Relationship between polysaccharides and 2-6 (a), 1-2 (b), 0.25-1 (c) and <0.25 (d) mm aggregate size fractions as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC.............................................................................70 Figure B.1: The layout of the long-term grazing study in the Lac du Bois Rangeland north of Kamloops, BC..........................................................................................................................................94 Figure D.1: Monthly precipitation (a) and monthly average temperature (b) at Kamloops airport, BC in 1997-1998 and 2007-2008 compared to 1971-2000 averages. The asterisks indicate sampling dates in 1998 and 2008. ........................................................................................................................................96 Figure D.2: Logarithmic regressions based on 4 extractions compared to 12 extractions on the same sample for the 0-7.5, 7.5-15, and 15-30-cm depths. The vertical axis measures root mass density in vi  grams (g cm-3), while the horizontal axis represents the number of sequential 10-minute extractions..96 Figure D.3: Total daily rainfall at Kamloops airport, BC, January 1997-June 2008...............................97 Figure D.4: Aggregate water content prior to wet-sieving analysis under spring and fall grazing and 0 and 2 AUM ha-1 treatments as determined 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32). Differences were considered significant at p<0.10 and marked by an asterisk.......................................97 Figure D.5: Relationship between total carbon and total nitrogen as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC..................98 Figure D.6: Relationship between total carbon and mechanical resistance as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC. Data for soil mechanical resistance were averaged over the 0-7.5, 7.5-15, and 15-30-cm depths for the purposes of the comparison......................................................................................................................98 Figure D.7: Relationship between soil mechanical resistance and root biomass as obtained on the longterm grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC. Data for soil mechanical resistance were averaged over the 0-7.5-, 7.5-15-, and 15-30-cm depths for the purposes of the comparison......................................................................................................................99 Figure D.8: Time of grazing and stocking rate treatment effects on bareground, litter, rock, and species cover as determined 32 years (i.e., 2010) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=64). Differences were considered significant at p<0.10 and marked by an asterisk. ................................................................100  vii  List of Equations Equation 1: Soil quality index equation...................................................................................................13 Equation 2: Soil quality element equation...............................................................................................13 Equation 3: Soil function equation...........................................................................................................13 Equation 4: Mean weight diameter equation...........................................................................................48  viii  List of Symbols and Abbreviations ANPP AU AUM BC BG CEC DPC LOMF MAP MDS MWD PP SQI  above-ground net primary productivity animal unit animal unit month British Columbia Bunchgrass (biogeoclimatic zone) cation exchange capacity desired plant community light organic matter fraction mean annual precipitation minimum data set mean weight diameter Ponderosa Pine (biogeoclimatic zone) soil quality index  ix  Acknowledgements I'm very grateful to a number of institutions and individuals who provided financial support for this project in the forms of scholarships and bursaries, including the National Science and Engineering Research Council, The BC Cattlemen's Association, Derek Vallis, Gilmour and Marjorie Clark, The Faculty of Land and Food Systems, the Faculty of Graduate Studies, and the Alma Mater Society of UBC. I am indebted to my supervisor, Dr. Maja Krzic, for her many edits and suggestions, and for her tireless and generous help and guidance. My gratitude also goes out to my committee, consisting of Dr. Les Lavkulich, Dr. Klaas Broersma, and Dr. Roy Turkington, as well as to my External Examiner, Dr. Andrew Riseman, and my M.Sc. Defense chair, Dr. Keith Adams, for their time and expertise. Plant species identification and cover estimates on the Lac du Bois Range would not have been possible without the special help of Dr. Don Thompson. I appreciate Dr. Reg Newman and Brian Wallace for their time and ideas on my data set, and I'm grateful to Dr. Tony Kozak for his advice on statistics. Special thanks go out to Andy Jakoy, Rachel Strivelli, Brian Wallace, and Melissa Iverson for their help during the week of sampling in Kamloops in June, 2008.  Christian Evans January 10, 2011  x  Chapter 1- General Introduction The production of food from rangeland agroecosystems is a globally significant phenomenon that occurs on half the world's land area and on every continent except Antarctica (Holechek et al., 2004). Range health assessment in British Columbia (BC) has been largely based on plant indicators (Fraser, 2007; Delesalle et al., 2009), and hence there are few studies examining grazing effects on ecosystem sustainability from the perspective of soil quality in BC. Rangelands present a unique opportunity for the assessment of sustainable land management practices because they typically occur in relatively dry regions where irrigated agriculture is limited (Gifford and Hawkins, 1978) and where detrimental effects of poor management can rapidly manifest. Historically, sustainable food production systems have resulted from a trial-and-error approach, while currently the creation of sustainable management programs is based more on the scientific analysis of ecosystem sustainability. Range management is founded on the principles of sustainable use and soil conservation (Task Group on Unity in Concepts and Terminology Committee, 1995), and has as its goal the development of sustainable practices using environmental quality indicators. The contention that “ranching is one of the only ways that our society produces food in intact ecosystems” (BC Grasslands Conservation Council, 2009) must be supported with environmentally sensitive management based on credible assessments of ecosystem health, including soil quality.  1.1 British Columbia Rangelands Rangelands can occur on almost all land that is not ice, bare rock, bare soil, or under intensive agricultural or urban development. Land grazed by domestic animals is the most dominant land type and comprises 50% of the world's land area (Holechek et al., 2004). Despite the low precipitation, sparse vegetation cover, and variable soil conditions of semi-arid rangelands (Gifford and Hawkins, 1978), these areas have served historically as a habitat for grazing and browsing animals. Sustainable range management makes use of limited rangeland resources with a goal to provide both forage for grazing animals as well as services such as air and water purification and open space for recreation in perpetuity (Vallentine, 1989). There are over 10 million hectares of rangeland in BC, 85% of which are on Crown land. Of this area, 8 million ha are forested rangelands that are also used for timber production, and 1.3 million ha are located on the grasslands of the Southern Interior and northeast BC (Wikeem et al., 1993).  1  Economic Significance of Ranching in British Columbia As of 2009, there were about 4000 cattle ranches in BC, producing over 120 million kg of beef (from bulls, cows, calves, steers, and heifers). The annual value of live animal exports totalled $62 million, while beef exports were just over $4 million, ranking the beef cattle industry as the fifth largest BC farm commodity. There were 1,270 members in the BC Cattlemen's Association in 2008, and it was estimated that the BC Beef Cattle Industry employed 8,700 people (British Columbia Cattlemen's Association, 2009). Cattle production in BC from 1976 to 2006 is summarized in Figure 1.1 (Statistics Canada, 2009) and shows the continuing importance of this industry to the provincial agricultural economy. It also shows that the number of small- to medium-sized operations has declined while larger operations have increased. This suggests that smaller ranches have been absorbed into larger ones or that the intensity of production has increased.  900000  Head of Cattle  800000 700000  Small (1-77 animals) Medium (78-272 animals) Large (273->1,128 animals) Total  600000 500000 400000 300000 200000 100000 0 1976  1981  1986  1991  1996  2001  2006  Year  Figure 1.1: Total cattle population and breakdown by farm size in British Columbia in selected years  History of Ranching in British Columbia Prior to the eighteenth century, the grasslands of the southern BC Interior were grazed by indigenous elk, deer, and bighorn sheep. The earliest recorded use by non-native animal species began when horses were brought to BC in the early 1700s from the United States. The native grassland at that time was 2  sufficient to sustain them throughout the year except in severe winters. The gold rush on the Fraser River in 1858 drew thousands of people north, bringing cattle with them. They arrived from the USA via the Okanagan, Columbia, and Kootenay Valleys, and some cattlemen settled to raise beef to sell in the markets of Vancouver and Victoria. The development of the Southern Interior cattle industry was facilitated by the construction of the Cariboo wagon road in 1863 and the trans-Canada Canadian Pacific Railway in 1885 (Bawtree, 2005). By the beginning of the 1900s, overgrazing was becoming an issue. Animals had usually grazed only in the open grasslands, but as cattle herd populations increased, summer grazing shifted to the forested areas and there was an increase in hay production and fencing of pasture to preserve grass production. Cattle grazed in season while horses grazed year-round (Bawtree, 2005). Tisdale (1947) remarked on the changed character of grassland vegetation due to overgrazing, and also cited older sources (Grant, 1872, and Macoun, 1876) that documented grassland depletion. Overgrazing was especially severe on lower-elevation grasslands due to their proximity to valley-bottom settlements and the fragility of their vegetation. During the First World War, beef production was encouraged and grazing on the grasslands intensified. From 1915 to 1917, there was an increase of cattle numbers in BC from 100,000 to 190,000 head, which brought the issue of overgrazing to public attention. The Kamloops Range Research Substation was formed in 1935 with government assistance. Cattle numbers remained steady during the Depression, but sheep numbers increased as sub-alpine and alpine areas were opened up to sheep grazing. Cattle numbers increased again during the Second World War, from 101,000 in 1940 to 156,000 in 1945. At that time, cattle were allowed to graze on public land for six months, and for a further one to six months on private land, depending on the location. At the present time, the grazing of the province's 212,000 head of cattle (British Columbia Cattlemen's Association, 2009) on public lands is strictly regulated (Bawtree, 2005).  1.1.1 Rangeland Management Ranches are agroecosystems that receive few inputs, and thus ranch financial viability depends heavily on managing for long-term environmental sustainability. The range manager uses the grazing system as a “tool for distributing the effects of grazing spatially or chronologically in order to conserve range resources” (Dormaar et al., 1997). Under continuous grazing, animals are allowed onto the same area for the entire year, and under season-long grazing, animals are allowed on specific areas at specific 3  times of the year (e.g., onto relatively productive pasture in the winter). Both grazing systems focus more on animal requirements than plant requirements and can lead to overgrazing, especially of preferred areas. Seasonal-suitability grazing is a more sophisticated extension of season-long, where both vegetation and livestock requirements are taken into account. Animals are rotated through different pastures over the growing season in rotational grazing, which enables more efficient use of forage and incorporates short periods of rest between grazing events. Longer rests (e.g., one year) are part of rest-rotation grazing systems, but benefits to livestock and vegetation are not clear (Campbell and Bawtree, 1998). Deferred-rotational grazing makes use of at least two pastures and allows soil and vegetation characteristics to improve (e.g., time for grasses to grow and set seed) before undergoing disturbance (Holechek et al., 2004). Deferring of the start of grazing and division of grazing pressure between the spring and fall seasons can help maintain or improve rangeland condition, and is the most appropriate system for semi-arid grassland rangelands (Campbell and Bawtree, 1998). Spring grazing is generally considered to be more damaging because new vegetation has had little chance to establish root systems and leaves and is thus highly susceptible to long-term damage (Laycock and Conrad, 1967; Naeth et al., 1991; Gayton, 2003). A sustained reduction in plant productivity may have negative consequences for soil properties, such as reduction of soil organic matter. Depending on the amount of snowmelt in the spring, the soil can be moist and consequently susceptible to compaction when grazed. In fall-grazed pastures, a lack of grazing disturbance in the spring ensures higher infiltration of snowmelt and/or precipitation, less overland flow and erosion, more water storage, and higher productivity. These conditions support accumulation of more aboveand below-ground biomass, increased levels of soil organic matter, and increased seed production. In the fall, pastures are generally drier and more resistant to compaction than in the spring, especially in regions with pronounced summer drought. The intensity of grazing is measured as the stocking rate, commonly expressed as animal unit months per hectare (AUM ha-1). This represents the amount of forage removed by one 455-kg “animal unit”, (such as a cow with her calf) in a month, which is 273 kg of air-dried matter per month (Holechek et al., 2004). The use of a generic “animal unit” (AU) enables comparisons between species through an appropriate conversion factor (e.g., a mature bull is 1.3 AU while a ewe is 0.2 AU). The limitations of a common unit are that it does not specify what kind and how many animals per stocking rate (since disturbance is a function of both of these factors) (Evans, 1998). Stocking rate is the major factor influencing range degradation (Bilotta et al., 2007). The ecological relationship between grasslands, 4  grazing animals, and fire suggests that a certain level of disturbance is necessary to maintain ecosystem health (Willms et al., 2002), and low stocking densities can promote biodiversity and nutrient cycling (Bilotta et al., 2007). The stocking rate is often designed to make use of a certain proportion of annual net primary productivity of forage (e.g., 50%). While the carrying or grazing capacity of a ranch is measured as a long-term average, stocking rates can differ between sites and between years depending on range productivity (Holechek et al., 2004). There has been a lack of upper stocking rate limits due to the economic inefficiency of maintaining an overgrazed field in the long term (Biondini et al., 1998), but most recommended stocking rates fall in the range of 0.8-2.4 AUM ha -1, depending on site productivity (Dormaar et al., 1989).  1.1.2 British Columbia Grasslands Extensive grasslands occupy the centre of the North American continent, bordered by the Rocky Mountains to the west, the boreal forest to the north, the deciduous forest to the east, and extending south of the US border as the Great Plains as far as Northern Mexico. In Canada, contiguous grassland comprises the majority of the prairie provinces of Alberta, Saskatchewan, and Manitoba. Further west, isolated areas of grassland occur in BC. Low precipitation, summer drought, and disturbances such as natural grazing and periodic fires limit the success of long life-history strategies such as those of trees and favour the growth of grasses. Grasses are better able than trees to tolerate the extremes of temperature and low soil water conditions due to their low above-ground biomass, extensive root systems, and critical moisture dependency only in the spring and early summer (Carder, 1970). Precipitation in grasslands varies from 320 to 900 mm, and is the major environmental factor controlling plant productivity (Pieper, 2005). Species distribution across the relatively level prairies is influenced by precipitation and topographical patterns (Pieper, 2005), and a number of prairie types have evolved. Tall-grass prairie occurs in the eastern portion of the prairies, where precipitation is more abundant; short-grass prairie occurs in the west, directly in the dry rain-shadow region east of the Rocky Mountains; and mixed-grass prairie occurs between these two zones (Pieper, 2005). In BC, the vegetation in the southern part of the province is typical of the Palouse prairie, which extends as far south as Utah. In comparison to the larger grasslands of the Great Plains, those in BC are unique both in their meagre distribution and their vegetation. In BC, grasslands are largely limited to the Chilcotin, upper Fraser, Thompson, Nicola, Okanagan, Similkameen, and Kettle regions (Figure 1.2), with some additional 5  areas found in the southeast in the upper Columbia and upper Kootenay regions and in the northeast in the Peace River district (Tisdale, 1947).  Figure 1.2: Occurrence of Grasslands in British Columbia source: The Grasslands Conservation Council of BC  The bulk of BC grasslands are located in the Southern Interior east of the Coast Mountain Range. The “rain shadow” effect of these mountains produces the climatic conditions which earned this region the name of “First Dry Belt”. Mean annual precipitation (MAP) in this region is low (e.g., 270 mm in Kamloops) and bimodally distributed, falling in December-January and June-July (Figure D.1). Far from the moderating influence of the ocean, temperatures can range widely (Carder, 1970) from a January mean temperature of -4.2°C to a July mean temperature of 21.0°C (Government of Canada and Meteorological Service of Canada, 2010). BC grasslands occur on the Interior Plateau, a formation dating from the Tertiary period when it consisted of a level plain. Subsequent uplift and erosion of the plain has resulted in the formation of hilly terrain punctuated by wide river valleys (Tisdale, 1947). The deepening of the valley floors to 600 6  to 900 meters below the level of the plateau has created an elevation and precipitation gradient that strongly affects the types of soils and plant communities. Grasslands in this area typically occur at an elevation of 300 to 1200 metres above sea level (Gayton, 2003), in the Bunchgrass (BG), Ponderosa Pine (PP), and Interior Douglas-fir (IDF) biogeoclimatic zones. British Columbia Grassland Vegetation and Soils The most limiting resource for plant growth in grasslands is water (Wikeem et al., 1993). Many xeric varieties of vegetation found nowhere else in Canada exist in the Southern Interior (Tisdale, 1947) due to the dry summer season. BC grasslands are dominated by bunchgrasses. Bunchgrasses grow in clumps, separated by a crust of lichens and mosses, and grow less quickly than their prairie counterparts. Heavy grazing and trampling by bison east of the Rocky Mountains selected for resilient, sod-forming grasses (Mack and Thompson, 1982), while in BC light grazing pressure enabled the evolution of less vigorous bunchgrasses. These species, typical of the Palouse Prairie ecosystem, moved into lands freshly exposed from de-glaciation about 10,000 years ago from eastern Washington, eastern Oregon, Idaho, and northern Utah (Tisdale, 1947). In the Southern Interior, three dominant plant associations exist along the elevation gradient from valley floor to the forested upland plateau. At lower elevations, where moisture is most limiting, the most common vegetation is bluebunch wheatgrass (Pseudoroegneria spicata (Pursh) A. Löve syn. Agropyron spicatum) and basin big sagebrush (Artemisia tridentata Nutt. ssp. tridentata). This zone, which extends from the valley floors to 700 m above sea level, is known as the Lower Grasslands (Tisdale, 1947). The Middle Grasslands (700-850 m) are characterized by the association of bluebunch wheatgrass and Sandberg's bluegrass (Poa secunda J. Presl). Above 850 m, bluebunch wheatgrass and rough fescue (Festuca campestris Rydb.) dominate the Upper Grasslands. Soils in the Southern Interior are formed from parent materials eroded from basalts, shales, sandstones, limestones, and granites. The advance of the Cordilleran ice sheet ground down and redistributed this material over the landscape. The recession of the ice sheet deposited a loose layer of ablation till over more compact basal till, and finer materials were picked up by wind and redistributed as an aeolian cap up to 30 cm thick in some places. This material provides an excellent growing medium for grasses (van Ryswyk and McLean, 1989). The sloughing of material from grass roots as they die and regenerate adds substantial amounts of organic matter to the soil, forming a chernozemic Ah horizon, the diagnostic horizon for Chernozemic 7  soils (Mollisols in the US soil classification system). In the Lower Grasslands, limited moisture and plant growth have formed Brown Chernozems (Aridic Haploborolls) with about 1% organic matter in the surface horizon. In the Middle Grasslands, slightly more favourable conditions have enabled the genesis of Dark Brown Chernozems (Typic Haploborolls), with a surface horizon containing up to 4% organic matter content, or even Brunisols (Inceptisols) where trees have encroached on the grassland. Higher up in the Upper Grasslands, more vigorous plant growth has allowed Black Chernozems (Udic Haploborolls) to form, with up to 10% organic matter content in the Ah horizon (van Ryswyk and McLean, 1989). Post-glaciation, North American Chernozems developed at rates decreasing from 4-5 mm yr-1 to <0.05 mm yr-1 after 3,000-10,000 years (Alexandrovskiy, 2007). In cooler and drier climates, pedogenesis proceeds at a slower rate; it is imperative to protect a resource which takes so long to form. As the Cordilleran ice sheet melted and retreated, primary plant succession occurred and the present soils formed. Dominant species in plant communities shifted from fast-growing weedy species producing large quantities of seeds to longer-lived plant species which invested more resources in growth and competition for light and soil resources (Grime, 1977). Under some circumstances, soils may become depleted of nutrients and the ecosystem may develop into a stressed state where hardier and smaller plants out-compete high-biomass competitors; however, in relatively young soils (~10,000 years), the amount of biomass is more likely to be affected by climatic limitations than by plantinduced soil nutrient limitation. Within intact grasslands, plant communities proceed through various seral stages as one species after another comes to dominate a changing ecosystem according to the ratio of light to nutrients (Tilman, 1985). Periodic disturbances, both natural and anthropogenic, also influence the type of vegetation. For example, the grasslands were traditionally managed by Native Peoples by burning (Gayton, 2003). Suppression of fires first by the discontinuation of traditional land management practices in 1880 and then by the BC Forest Service beginning in the 1940s has led to encroachment on grasslands by trees (Gayton, 2003). Severe disturbance (e.g., intense burning or overgrazing) opens up new niches for unpalatable weedy plants and pushes the plant community “backwards” through seral stages, potentially reducing the amount of biomass production (Gayton, 2003). Unchecked disturbances not only physically remove palatable biomass but also cause a decline in the range's ability to produce forage. Evaluating rangeland plant productivity or plant species distribution can document these ecological changes. In addition to their value in supporting agroecosystems, grasslands exhibit high biodiversity. 8  Grasslands represent only 1.4% of the land area in BC, but they contain more than 30% of wildlife species at risk (Grasslands Conservation Council of British Columbia, 2010). Beginning with grazing pressure from cattle brought in during the 19th-century European settlement of the region, BC grasslands have endured a number of challenges to their long-term sustainability, including inappropriate land management, invasive plants, forest encroachment, urban development pressure, and abusive recreation practices.  1.1.3 Rangelands in British Columbia Grasslands Climatic factors such as recurrent drought, prolonged periods of cold and heat, and high winds contribute to the occurrence of extremes in ecosystems such as temperate grasslands (Carder, 1970) and limit the suitability of these ecosystems for many types of agriculture other than low-intensity management operations such as ranching. Forage production can range from 250 to 2700 kg ha -1 in BC grasslands depending on biogeoclimatic zone, site, soil type, range condition, and weather (Wikeem et al., 1993). Bluebunch wheatgrass, needle-and-thread grass (Stipa comata var. Comata Roshev.), prairie Junegrass (Koeleria macrantha (Ledeb.) J.A. Schult. f.), and crested wheatgrass (Agropyron cristatum (L.) Gaertn.) can produce as much as 30% of their annual production in fall regrowth depending on soil water content and temperature, so some rangeland is used to produce a second round of forage after initial spring grazing and a rest period. Close to the valley floor, the drier Bunchgrass zone (comprising the Lower and Middle grasslands) is suitable for late fall, winter, and early spring grazing. Further up the elevation gradient, the Ponderosa Pine zone (Upper grasslands) provides early spring and late fall grazing. In the coolest and wettest zone, the Interior Douglas-fir zone (Upper grasslands), higher forage production provides for late spring, summer, and early fall grazing (Wikeem et al., 1993). Cattle are moved up and down this gradient over the grazing season. Range management in BC is varied due to the diversity of vegetation and climate across the province's rangelands, as well as by the overlapping multiple uses of rangelands, including timber (in forested rangelands), recreation, watershed quality and capacity, and the maintenance of biodiversity and wildlife (Wikeem et al., 1993).  1.2 Soil Quality Soil, in its roles as a physical matrix, nutrient storehouse, and biological habitat, is crucial to the growth of almost all types of vegetation, and certainly to the vast majority of plants that directly or indirectly 9  comprise the human diet. Recognition of the critical part soil plays in the biosphere makes a strong case for the use of soil quality as an indicator of ecosystem sustainability (Carter, 1996; Carter et al., 1997). However, there have been challenges to the widespread adoption of the concept of soil quality. As opposed to air and water, which function directly to sustain human health, soil has an indirect influence. Along with the complexity due to soil's multiplicity of functions, this creates a more tenuous link between soil quality and human and environmental health (Carter, 1996). As a result, it has been more difficult to raise awareness in policy circles and among the public of the importance of soil quality. Attempts to define soil quality generate important questions: Quality for whom? For what purpose or use? For how long? Soil is both a natural body and a useful material, and therefore a definition of soil quality depends on the desired function and the soil's ability to sustain that function.  1.2.1 Soil Quality Concept Soil Health vs Soil Quality In defining soil quality, a distinction must be made between soil health and soil quality. Soil health refers to the soil's natural ability to perform its ecosystem functions independently of any planned use (Carter et al., 1997). A measure of soil health is the soil's resistance to deterioration and its resilience following disturbances (including employment for use). All other things being equal, in the absence of anthropogenic disturbance, a given soil can be assumed to be in dynamic equilibrium with its environment and operating up to its potential functional level; that is, it can be said to be healthy. On the other hand, soil quality is an assessment of a soil's usefulness for a particular task to which value has been assigned (Carter et al., 1997), and therefore contains a bias toward function. Where the task is to function as a natural soil by recycling nutrients, storing and purifying water, or resisting erosion, soil health and soil quality are equivalent. However, where the desired function of the soil differs from its natural function, definitions of soil quality and soil health will differ. For example, it is possible to have a soil with high quality and low health (e.g., agricultural soils with high productivity but low ability to accept water, leading to surface water quality issues) or one with low quality and high health (e.g., when contemplating crop production using coarse-textured forest soils on steep slopes). Inherent vs Dynamic Quality The word “quality” is commonly used to define a state. However, some states are more persistent than  10  others, and so one must distinguish between inherent and dynamic soil quality (Seybold et al., 1999). Inherent soil quality is based on properties which result from Jenny's (1980) soil-forming factors (i.e., climate, topography, parent materials, biota, and time), such as texture, arrangement of horizons, or parent material. These properties are less likely to change in the short-term (Brejda et al., 2000), and are used to classify and map soils. On the other hand, dynamic soil properties, such as bulk density, organic matter content, soil structure, or pH (Arnold et al., 1990), are more sensitive to external pressures and thus can be used to track management effects or even reconstruct a history of management. A soil's inherent properties influence the range over which its dynamic properties can vary (Norfleet et al., 2003). For example, a soil's water-holding capacity depends on transient properties such as organic matter content as well as more permanent properties such as particle size distribution, soil structure, and mineralogy. Since the amount of organic matter itself also depends on these inherent soil properties, it is largely the inorganic fraction of the soil that controls water availability (Norfleet et al., 2003). The distinction between inherent and dynamic soil quality is highlighted by Norfleet et al. (2003) in their comparison of the concept of soil quality to the practice of pedology. Both concepts focus on soil qualities or properties, but also differ in that pedology considers large-scale landscapes over long time periods while soil quality is usually concerned with “point-scale” effects of short-term management. Pedology can be helpful for assessing a soil's initial suitability for a given task based on its inherent properties. Following that, the severity of management effects on the soil's dynamic properties will decide the continuing ability of that soil to function as desired. Properly assessing suitability for use based on both inherent and dynamic soil quality means that management systems can be made more efficient and less expensive by assigning to a soil uses for which it is naturally suited (Norfleet et al., 2003). When assessment of a soil is not done or done poorly, the potential for sustainable use decreases. The Function Bias of the Soil Quality Definition Management goals form the basis of an assessment of soil quality (Andrews and Carroll, 2001). The goals of sustainable management are subject to the soil's ability to support them. There are some functions for which a given soil is not suited without alteration of its dynamic properties, although inherent properties are unchanging or very slow to change. For example, with regard to the use of soil as an engineering medium (e.g., as a foundation for a house), texture and particle density determine its 11  maximum bulk density, but the degree of compaction depends more on the extent of alteration of properties such as organic matter content, porosity, aggregation, soil water content. Any one soil in a given state can function for only a limited number of desired uses: agriculturally productive land is a poor choice for a building foundation unless the topsoil is removed and the subsoil compacted. Areas which are used for multiple purposes, such as rangelands (Herrick and Whitford, 1995), run the risk of becoming suitable for only the more intensive of those uses (e.g., grazing rather than forage production, or recreation rather than food production). As the severity of disturbance (i.e., the intensity of use) increases, maintaining soil quality requires a shift in management resources (Carter et al., 1997). Correctly matching a desired use to a suitable soil depends not only on the correct identification of the desired function(s) but also on the soil's ability to sustain those functions. Boundaries of the Soil System Soil at any scale contains, and is contained within, other ecosystems. Its ability to accept, store, and release nutrients, water, and gases connects it to every higher and lower level. The soil's ecological interactions cannot be ignored, and due to the endless connections to other ecosystems they must be to delimited. Thus the delineation of ecosystem boundaries is required as part of the definition of soil quality. Establishing boundaries, which may be expanded or reduced to suit particular purposes, focuses attention on management effects contained within. When the scope of an evaluation of soil quality is widened to include the ecosystem, the real effects of management become clearer (Karlen et al., 1997). For example, an analysis of soil quality at the site level would not reflect as many of the consequences of poor management as an analysis at the landscape level. Soil Quality Definition While there have been widely accepted definitions for water and air quality for some time, a definition for soil quality has been more elusive (Doran and Parkin, 1994), owing to the variety of complex functions the soil performs for the ecosystem as well as for human societies. Proposed definitions for soil quality vary (Doran and Parkin, 1994; Harris and Bezdicek, 1994; Larson and Pierce, 1994), but most propose that the concept of soil quality consist of a soil's ability to function within its ecosystem boundaries, to interact positively with the environment external to that ecosystem, to support sustainable and healthy production of plants and animals, to function as an environmental buffer enhancing water and air quality, and maintain suitability for an intended use. 12  The concept of soil quality is defined relative to the user (Carter et al., 1997) and interpretations of acceptable levels of soil function are subjective. For example, desired functions are different for rangelands, intensive agriculture, or recreation. A soil quality description must represent the aggregate of all desired functions. Soil Quality Index Soil quality assessment must consider a complex range of issues and components. For example, Doran and Parkin (1994) offer a qualitative model which illustrates the relationship between the components of soil quality. It relates an overall soil quality index (SQI) to six soil quality elements (SQEx): SQI = f {KSQ E1 , KSQ E2 , KSQ E3 , KSQ E4 , KSQ E5 , KSQ E6 }  (Equation 1)  where: K= weighting factor; SQE1= food and fibre production; SQE2= susceptibility to erosion; SQE3= groundwater quality; SQE4= surface water quality; SQE5= air quality; and SQE6= food quality. These soil quality elements depend on that status of various soil functions: SQ Ex = f {SF 1 , SF 2 , SF 3 , SF 4 , SF 5}  (Equation 2)  where: SF1= the ability to hold, accept, and release water to plants, streams, and the subsoil; SF2= the ability to hold, accept, and release nutrients and other chemicals; SF3= the ability to promote and sustain root growth; SF4= the ability to maintain suitable soil biotic habitat; and SF5= the ability to respond to management and resist degradation. Finally, soil functions are defined as a function of soil properties, e.g.: SF 1= f {infiltration rate , water retention}  (Equation 3) 13  Doran and Parkin's proposed model is not simple – not every soil function is applicable to each soil quality element – and yet it only approximates the complexity of the soil system. This example of conceptual equations illustrates the association of soil quality, soil quality elements, soil functions, and soil properties. For the purposes of my study, a simpler hierarchical model for soil quality assessment has been used. This model consists of soil functions that are dependent on various soil processes, which are in turn that are supported by various soil properties. The final group in this association are the indicators of soil properties.  1.2.2 Soil Quality Indicators Continuous monitoring of air and water can be conducted at one point and general conclusions drawn about the particular watershed or instrument tower footprint being measured due to these compounds' properties as fluids: they mix well and can travel long distances. Soil, however, is stationary and heterogeneous, and so continuous monitoring is not common (Carter et al., 1997). Monitoring must be frequent enough to detect changes in soil quality, but infrequent enough to be affordable. The selection of quality indicators pertinent to soil properties must be based on proven links to desired or ecosystem functions. Characteristics of Useful Indicators The assessment of soil quality depends on the functions the soil is expected to perform (including both ecosystem functions and desired uses), and these functions, in turn, are supported by various soil processes. Each soil process depends on one or more soil properties. Doran and Parkin (1994) suggested that soil properties useful for soil quality monitoring and assessment must: 1. be sensitive to management and able to show changes in a relatively short time; 2. be properties for which measurement methodologies exist; 3. have accessible (and ideally long-term) data sets; and 4. be properties for which pedotransfer functions can be defined. A pedotransfer function is a mathematical relationship between a more easily measured variable and one which is less easy to measure. In the event that a desired property cannot be directly measured, it can be calculated from other obtainable measurements (Karlen and Stott, 1994). The status of each property, in turn, can be inferred from measuring soil quality indicators. According  14  to Doran and Parkin (1994), ideal indicators should: 1. integrate soil physical, chemical, and biological properties and processes; 2. be accessible to many users and applicable to field conditions; 3. be sensitive to variations in management and climate; and 4. where possible, be components of existing soil data bases. It is crucial to select indicators that are well-correlated to desired soil functions. Indicators must be sensitive enough to show management effects in the short and long terms, but not so sensitive that they react continuously to the variability of a natural system (Larson and Pierce, 1991; Herrick et al., 2002). Similar to Doran and Parkin's weighted soil quality index, Karlen and Stott (1994) developed a weighted indicator approach by identifying indicators “critical” and “important” to a particular soil function. Larson and Pierce (1991) related selected soil properties to Doran and Parkin's (1994) five soil functions (Table 1.1). Table 1.1: Selected soil properties and supported functions.  Function Accept hold and release nutrients Accept hold and release water Promote growth Provide a suitable soil biotic habitat Resist degradation  Soil Structure      Texture      Organic matter       pH      Nutrient supply     Minimum Data Set Although descriptions of soil properties (e.g., colour, texture, structure, consistence) have long been used by soil managers, there has been a shift over the last 100 years towards analytical measurements of soil properties (Harris and Bezdicek, 1994). These are standardized and can be replicated, enabling comparisons through time and among different locations. However, analytical measurements are generally more costly, and therefore efforts have been made to identify the minimum amount of information needed to inform management decisions, known as the minimum data set (MDS). The MDS defines the set of properties that will be monitored in order to assess soil quality, and so it is essential that the indicator variables included be representative of the desired functions (Doran and Parkin, 1994; Gregorich et al., 1994; Larson and Pierce, 1994). Due to the wide variety of soils and functions, there is little consensus on what properties should form 15  part of an MDS (Larson and Pierce, 1991, 1994). Larson and Pierce (1994) suggest selecting properties that are reproducible and easy to measure, and which support the processes and functions pertinent to management goals. Relevant properties can be determined by identifying the smallest number of soil quality indicators which explain the largest proportion of total soil data variability. Subsequently this MDS is tested against indicators of successful management (e.g., phosphorus runoff potential, metal contamination, or net profitability of the ranching operation) (Andrews and Carroll, 2001). Studies by research scientists have suggested various properties for inclusion in an MDS (Table 1.2). Table 1.2: Soil properties suggested for inclusion in an MDS.  Property Texture  Reference Larson and Pierce, 1991; Doran and Parkin, 1994; Gregorich et al., 1994; Martin et al., 1998  Soil water  Doran and Parkin, 1994; Andrews and Carroll, 2001  Bulk density  Arshad and Coen, 1992; Doran and Parkin, 1994; Gregorich et al., 1994; Karlen et al., 1997; Martin et al., 1998; Andrews and Carroll, 2001  Soil structure  Larson and Pierce, 1991; Karlen et al., 1997; Andrews and Carroll, 2001  Soil organic matter  Arshad and Coen, 1992; Larson and Pierce, 1991; Doran and Parkin, 1994; Gregorich et al., 1994; Karlen et al., 1997; Martin et al., 1998; Brejda et al., 2000; Andrews and Carroll, 2001  pH  Arshad and Coen, 1992; Larson and Pierce, 1991; Doran and Parkin, 1994; Gregorich et al., 1994; Karlen et al., 1997; Martin et al., 1998; Andrews and Carroll, 2001  Phophorus  Martin et al., 1998; Andrews and Carroll, 2001  Cation exchange capacity  Arshad and Coen, 1992; Martin et al., 1998; Andrews and Carroll, 2001  Rezeai et al. (2006) determined an MDS for assessing rangeland soil quality for semi-arid grasslands (476 mm MAP) in Iran. Soil chemical and physical properties were correlated and regressed against three indicators of range productivity: current year production (biomass), herbaceous plant production, and utilizable forage. Two properties, soil profile effective thickness (the equivalent soil depth consisting only of <2-mm particles) and percent total nitrogen, explained most of the variation in the indicators of rangeland productivity. Two other properties, grade of pedality (i.e., hardness of aggregates) and the thickness of the soil above the B horizon, accounted for a smaller proportion. While some proposals for an MDS aim for global relevance, their authors acknowledge that regional variability will demand different methods of determination as well as other soil quality indicators 16  (Larson and Pierce, 1991). Once a regionally tailored set of properties has been determined, critical ranges for each property must also be defined. Reference Values and Critical Limits The character of treatment effects (i.e., amelioration or degradation) on soil quality must be determined based on a comparison of an indicator value to a reference. The optimum reference level is the level of the indicator value in an undisturbed portion of the ecosystem (Arshad and Martin, 2002). If this is not possible, the reference level may be estimated based on soil resource inventories (i.e., previous studies) or arrived at by consensus of stakeholders and experts (Carter, 2002a). The reference value is located between critical (or threshold) limits. These limits define a range over which soil function can be sustained (Carter, 2002a). Critical limits should be defined in order that statistically significant differences be properly assessed for their effect on soil quality – i.e., does a significant change in indicator value have consequences for soil quality? Defining critical limits requires research about the relationships between particular properties and the functions they support as well as site-specific conditions which structure the relationship. Critical limits for soil properties which relate strongly or are exclusive to one particular function are easier to define than those for multi-functional properties (e.g., soil organic matter). Critical limits for soil organic matter would be site-specific and individual ranges would need to be specified for each associated function (Carter, 2002b). Arshad and Martin (2002) suggested that a 15% change in soil organic matter levels (region not specified) seemed a “reasonable” signal that quality had been significantly increased or decreased. Critical limits are important for soil quality assessment because as they are approached, the relationship between soil properties and processes can become non-linear. Doran and Parkin (1994) used linear combinations of indicators and processes in their soil quality index, and while such an approach may be useful for an ecosystem which is gradually changing without significant disturbances, ecological theory suggests that some ecosystems are also structured by infrequent but catastrophic events (e.g., acid rain, or, in the case of grasslands, fire). Moreover, the cumulative effects of disturbance may be simultaneously expressed when critical limits are reached, an outcome not predictable by linear models (Herrick et al., 2002).  17  1.2.3 Soil Quality and Rangeland Health As forage production is the most obviously crucial resource for range managers, assessments of range health have long focused on plant indicators, and only recently has soil quality become a basis for the evaluation of range condition. Historically, range health assessment was based on the view that plant secondary succession could be controlled positively by environmental variables such as precipitation or negatively by disturbances such as grazing (Pyke et al., 2002). Due to some limitations of this approach, including the lack of a role for other disturbances such as fire, in 1994 the National Research Council suggested that evaluation of soil stability, watershed function, integrity of nutrient cycles and energy flows, and the presence of functioning recovery mechanisms form part of range health assessments (Pyke et al., 2002). The following year, the Society for Range Management's Task Group on Unity in Concepts and Terminology recommended three foci for the assessment of range health: the ecological site, defined as a type of land which differs from others in the vegetation it supports and in its response to management, and which corresponds with the phases of the US soil taxonomic system; the site conservation rating, which is a measure of the protection afforded by the vegetation of a site against erosion; and the desired plant community (DPC), which consists of vegetation which meets the range use plan's objective for the site (e.g., vegetation which does not protect against erosion would not comprise the DPC) (Pyke et al., 2002). The new focus for the concept of range evaluation was to be the soil (Task Group on Unity in Concepts and Terminology Committee, 1995). In BC, the two most commonly used methods for assessing rangeland health are the BC Ministry of Forest and Range's Uplands Assessment Method (Fraser, 2007) and the BC Grassland Conservation Council's Grassland Assessment scoresheet (Delesalle et al., 2009). The Uplands Assessment Method was found to correlate more strongly with soil data due to its emphasis on soil and site stability (Newman et al. 2007). The shift in rangeland health assessment philosophy and the results of Newman et al.'s study both promote soil quality as an integral component of such assessments. Rangelands are often situated in regions with climates unfavourable for rain-fed agriculture or lacking irrigation infrastructure (Gifford and Hawkins, 1978). Rangelands present a challenge for soil quality assessment because they typically exhibit a much greater variability than cropped land. Management inputs are usually minimal, and without the homogenizing effect of tillage, assessment may be confounded by the effects of shrub distribution, animal compaction along tracks, and localized effects of livestock feces and urine (Herrick and Whitford, 1995). Unlike in intensively managed agricultural land where the soil's natural productivity is deliberately enhanced, in rangelands the concepts of soil 18  health and soil quality are near in definition and indicators should correlate with ecosystem function (Herrick et al., 2002). When appropriate and representative indicators are not selected, it may result in failure to properly assess rangeland soil quality. In addition to this criteria, Herrick et al. also suggested three more guidelines for indicator selection for rangeland soil and vegetation monitoring: 1. Indicator selection should change with site-specific resource concerns and inherent soil qualities; 2. The spatial variability of indicator response should be taken under consideration when interpreting indicators so that they better represent ecological processes; and, 3. The interpretation of indicators should be done with an awareness of dynamic, non-linear ecological processes. Soil Quality Indicators for Rangelands Despite unfavourable environmental conditions such as periodic drought and temperature extremes, rangeland managers aim to maintain the range in as natural a condition as possible while still providing forage for livestock. Desired soil functions should support this management goal, and thus two main functions for rangeland soil are proposed: (1) to provide a suitable habitat for plant growth and (2) to partition water at the surface. Plant growth depends, for example, on soil physical processes such as gas exchange and water storage and release, and chemical processes such as nutrient storage and release and pH buffering. Adequate water partitioning depends on soil texture and structure that allow for accepting, storing, and releasing water, among other properties. For rangelands, the sets of functions associated with soil health and soil quality are similar. For my study, the key functions of the rangeland ecosystem, along with their supporting processes, properties, and indicators, are summarized in Table 1.3.  19  Table 1.3: Soil quality assessment framework for the Lac du Bois rangeland in Southern Interior of BC  Function  Process capacity to accept, hold, and release nutrients  Property cation exchange capacity organic matter texture soil reaction  Indicator cation exchange capacity  total carbon and nitrogen % sand, silt, and clay pH Provide suitable available phosphorus and other habitat for plant provision of adequate nutrients soil fertility status growth nutrient to plants root biomass organic matter total carbon and nitrogen capacity to accept, bulk density and particle porosity hold, and release water density texture % sand, silt, and clay organic matter total carbon and nitrogen mean weight diameter and size distribution of water-stable maintenance of stable soil aggregate stability aggregates soil structure polysaccharides Partition water at the root biomass soil surface clay content % clay organic matter total carbon and nitrogen capacity to accept, bulk density and particle porosity hold, and release water density texture % sand, silt, and clay  1.3 Grazing Effects 1.3.1 Grazing Effects on Soil Chemical Quality Soil Organic Matter Soil organic matter is the most important indicator of soil quality (Larson and Pierce, 1991). Its widespread effects on a suite of soil properties including cation exchange capacity, soil fertility, and soil water storage, as well as its multi-functional nature (Table 1.1) guarantee it a place within a data set. Soil organic matter is influenced primarily by climate and vegetation patterns (Bauer et al., 1987; Schuman et al., 1999) and secondarily by the effects of grazing and/or herbivory. Maximum decomposition of organic matter occurs at 37°C, and so rangeland denuded of its protective cover of vegetation may experience soil organic matter losses due to high soil temperatures and rates of 20  decomposition, especially under sufficiently moist conditions (Snyman and du Preez, 2005). The effects of secondary influences (i.e., grazing) can depend on conditions including the amount and distribution of vegetation or the amount of soil organic matter. Soil organic matter consists largely of carbon, oxygen, hydrogen, and nitrogen. From a soil fertility perspective the elements of interest are carbon and nitrogen, and levels of these two elements can be used as indicators of the quantity of soil organic matter. The quality of soil organic matter (i.e., the mineralization potential) is indicated by the carbon-to-nitrogen ratio. In some semi-arid regions, the presence of soil carbonates may confound the use of total soil carbon as a proxy for soil organic matter and consequently organic carbon is a better choice for calcareous soils. However, soil carbonates tend to accumulate below a certain depth (the leach layer) and sampling from above this layer can minimize the influence of carbonates on the apparent levels of soil organic matter. Disturbance due to grazing can cause plants to allocate more carbon and nitrogen to above-ground tissues than roots (Detling et al., 1979), thereby decreasing below-ground stores (Schuman et al., 1999; Ganjegunte et al., 2005). Although nitrogen is another major component of organic matter, its availability is subject to restrictions which differ from those of carbon. Carbon losses due to herbivory can be offset by increased photosynthesis, but lost nitrogen must be replaced either through biological fixation of atmospheric nitrogen in the soil (Schuman et al., 1999; Willms et al., 2002) or by nitrogen deposition. Total soil nitrogen in ungrazed pasture may be lower due to conditions which favour denitrification, such as under increased grass cover where lower soil temperatures can result in more water-filled pores and anaerobic environments. Nitrogen can also be lost from the soil-plant system through volatilization from live and senescing plant tissue (Bauer et al., 1987), which is more abundant in ungrazed areas. Patchy deposition of livestock feces and urine may lead to high variability of soil nitrogen content (Petersen et al., 1956; Banerjee et al., 2000). A change in the carbon-to-nitrogen ratio can indicate a change in the quality of soil organic matter. The ratio of carbon to nitrogen in soils controls whether net mineralization (carbon-to-nitrogen ratio <25) or immobilization (carbon-tonitrogen ratio >25) of nitrogen occurs. However, the carbon-to-nitrogen ratio of Chernozemic soil organic matter is generally in the vicinity of 11 (Brady and Weil, 2007). The carbon-to-nitrogen ratio is also related to the degree of aggregation as it denotes the quality of the organic materials in the soil; generally, organic material with a low carbon-to-nitrogen ratio enhances aggregate formation (Bird et al., 2002). The effects of grazing (both time of grazing and stocking rate treatments) on soil organic matter 21  quantity and quality are variable. In a global review of rangeland literature comprising 236 sites, Milchunas and Lauenroth (1993) found that responses of soil organic matter, soil carbon, and soil nitrogen to grazing were “nearly equally divided between negative and positive”. Naeth et al. (1991) reported that after 12 years of cattle grazing, total organic matter was highest under a light (1.5 AUM ha-1) autumn treatment as compared to heavy (4.4 AUM ha-1) autumn, June, or ungrazed treatments on a loamy Black Chernozem in Alberta (422 mm MAP), supporting the hypothesis that late-season grazing has fewer negative effects on soil quality than early-season. At a moist (550 mm MAP) site in the same study, Naeth et al. (1991) found that 36 years of continuous cattle grazing on a Black Chernozem under a variety of stocking rates (1.2, 1.6, 2.4, and 4.8 AUM ha -1) significantly decreased the amount of organic matter in the Ah horizon by 13-19% compared to ungrazed exclosures. The soil organic matter in the Ah horizon (from 0-10 to 0-15 cm) at this site may have been less affected by grazing disturbance due to high levels of organic matter (~8 kg m -2 below-ground organic matter in the control plot). A change in species composition or a decrease in plant productivity due to disturbance can negatively affect contributions to soil organic matter from roots and litter (Snyman and du Preez, 2005) as well as the dynamics of carbon cycling between plants and the soil. In a long-term (44 years) study in southwestern Alberta (550 mm MAP) on a Black Chernozem, Dormaar and Willms (1998) found that total soil carbon was significantly lower on heavily (2.4 AUM ha-1) and very heavily (4.8 AUM ha1  ) grazed pastures than on lightly (1.2 AUM ha-1) grazed pastures and in grazing exclosures. Under  grazing pressure, the thickness of the Ah horizon decreased from 22 to 8 cm and the colour lightened from dark brown to brown. The authors suggested that this was caused by reduced organic matter inputs to the Ah horizon due to species composition change or possibly by compaction. On the other hand, Schuman et al. (1999) found that total soil carbon to 30 cm was 21 and 22% higher after 12 years of season-long cattle grazing at 0.6 and 2 AUM ha -1 relative to ungrazed pasture on a sandy loam Brown Chernozem (Aridic Argiustoll) at a semi-arid (384 mm MAP) high-elevation (1950 m) site in Wyoming. This was attributed to an increase in cover of blue grama (Bouteloua gracilis (H.B.K) Lag. Ex Steud.), which allocates more carbon to below-ground parts than other species characteristic of the mixed-grass prairie. In other cases, grazing treatments had no significant effect on soil organic matter. Johnston et al. (1971) reported no significant differences in organic matter content on a Black Chernozem at a fescue grassland site (550 mm MAP) after 17 years of continuous cattle grazing at light, moderate, heavy, and very heavy (1.25, 1.67, 2.5, and 5 AUM ha-1, respectively) stocking rates. Mapfumo et al. (2000), on a 22  Dark Brown Chernozem (Typic Haplustoll) in central Alberta (400 mm MAP), found no significant differences for total soil carbon after 3 years between light, medium, and heavy (16, 20, and 37 cowdays per season) grazing treatments. In a multi-site study on an Alberta Chernozem, Dormaar et al. (1977) found that total carbon (and nitrogen) were not significantly affected by 22 years of continuous cattle grazing at 5 AUM ha -1 relative to an ungrazed control on a Black Chernozem at a fescue grassland site (500 mm MAP). This contrasted with a Solonetz at a mixed-grass prairie site (310 mm MAP) which showed increased soil carbon under grazing pressure. They attributed this difference of grazing response to the fact that the Chernozem had ten times the level of soil carbon of the Solonetz. Soil nitrogen correlates strongly with the organic carbon content, and thus similar effects of stocking rate may be observed. Schuman et al. (1999) reported results for nitrogen similar to those they reported for carbon: total soil nitrogen content to 30 cm was 29 and 12% higher under light and heavy grazing relative to ungrazed exclosures. However, other studies found that trends for soil carbon due to grazing pressure were not echoed by soil nitrogen. Dormaar and Willms (1998) found that soil nitrogen, unlike carbon, increased under heavy and very heavy grazing treatments. Mapfumo et al. (2000) found no significant differences in soil carbon, but found significant increases (from 0.043 to 0.068 g 100 g -1) of soil nitrogen in the top 15 cm with grazing intensity. Frank et al. (1995) reported a decrease in total soil nitrogen of 25% over 75 years under both light (0.4 AUM ha -1) and heavy (1.1 AUM ha-1) cattle grazing relative to the ungrazed control on a fine silt Dark Brown Chernozem (Haploboroll) in North Dakota. However, a concomitant decrease in carbon was not observed, which the authors attributed to an increase in the short-rooted species blue grama and increased manure deposition. On a loamy Orthic Grey Luvisol (Argiboroll) in North Dakota (446 mm MAP), Biondini et al. (1998) found that soil nitrogen content to 10 cm remained unaffected by grazing at moderate (50% forage use) and heavy (90% forage use) intensities. The response of the carbon-to-nitrogen ratio of rangeland to grazing soils hinges on the differential effects on soil carbon and nitrogen as well as the amount and quality of litter incorporated into the soil. On a Solonetz at a dry (310 mm MAP) site in Alberta, Smoliak et al. (1972) found that 19 years of season-long (May 1 to November 1) sheep grazing increased the carbon-to-nitrogen ratio in the Ah horizon from 9.1 to 9.9, 10.4, and 10.3 under light (0.4 AUM ha-1), moderate (0.5 AUM ha-1), and heavy (0.6 AUM ha-1) grazing treatments relative to an ungrazed control. This was due to an increase in carbon while nitrogen remained unaffected and the authors did not relate changes in the carbon-tonitrogen ratio to any other variables. No significant differences for carbon-to-nitrogen ratios between 23  various rangelands conditions were found by Snyman and du Preez (2005) in South Africa. They were investigating grazing effects on a fine sandy loam Bloemdal Form in a semi-arid (560 mm MAP) region by clipping and removing vegetation in order that plant communities conform to species distributions characteristic of desired pasture states (poor, moderate, and good). Dormaar et al. (1977) reported no significant effect due to grazing on the carbon-to-nitrogen ratio in the Ah horizon of a Black Chernozem. This finding was attributed to higher pre-grazing soil organic matter levels. As has been alluded to in several of the studies presented so far, the level of organic matter prior to grazing treatment influences the severity of response. When seasonal or grazing-induced fluctuations are relatively small compared to total organic matter content, these fluctuations are difficult to measure (Banerjee et al., 2000). Soils low in organic matter are generally less resilient and more susceptible to treatment interactions (e.g., grazing by climate) than more fertile soils with higher organic matter content (Dormaar et al., 1977). In Alberta, Dormaar et al. (1984) found a trend for higher total carbon at two sites: on a Solonetz at a mixed prairie site (310 mm MAP) after 19 years of sheep grazing at 0.6 AUM ha-1 relative to an ungrazed control and on a Black Chernozem at a fescue grassland site (500 mm MAP) after 29 years of continuous cattle grazing at 5 AUM ha-1. This result was more consistent for the Solonetz, which had lower total carbon prior to the inception of the study. The link between soil organic matter levels and factors such as climate and precipitation may be less predictable than supposed. Henderson et al. (2004) found no major influence on soil carbon from soil water content or site productivity. They examined soil carbon content at nine sites with Chernozemic soil types (Brown to Black) along an environmental gradient (930-1370 m, 350-480 mm MAP) in southern Alberta and concluded that there was no consistent grazing effect and that the “response of northern rangeland soils may be unique to each location”. Soil organic matter can be influenced directly by grazing animals. Although litterfall naturally replenishes the soil organic matter pool, trampling of this material can accelerate its decomposition and limit its influence on soil structure (Mapfumo et al., 2002), or it can increase the amount of organic matter incorporated into the soil (Ganjegunte et al., 2005). After 11 years of cattle grazing at 0.7 and 2.2 AUM ha-1 on a mesic (338 mm MAP) Brown Chernozem (Aridic Argiustoll) in Wyoming, Manley et al. (1995) found higher carbon and nitrogen in the top 30 cm of the grazed sites relative to the exclosure. A reduction in standing litter in the grazed areas compared to the exclosures and the animalmediated incorporation of this material into the soil through trampling were identified as causes of the higher levels. Schuman et al. (1999) attributed findings of higher soil carbon and nitrogen in the top 30 24  cm to changes in species composition and altered rates of carbon cycling (i.e., reduced immobilization in above-ground biomass in grazed areas), but they also cited animal influences as well: trampling increased shoot turnover, litter decomposition, and litter incorporation. Evidence for the hypothesis that semi-arid grassland ecosystem quality is maintained by low-level disturbance (i.e., light grazing) came from a study conducted on a Brown Chernozem (Aridic Argiustoll) in Wyoming (384 mm MAP) by Ganjegunte et al. (2005), where total nitrogen was higher after four years under continuous light (0.67 AUM ha-1) cattle grazing than under either heavy (2 AUM ha-1) grazing or no grazing. As previously mentioned, Naeth et al. (1991) found that total organic matter was highest under a light treatment as compared to heavy or ungrazed treatments, suggesting that light grazing and trampling pressure may stimulate the growth of vegetation and contribute to soil organic matter accumulation. Light Organic Matter Fraction The light organic matter fraction (LOMF) is composed of partially decomposed fragments of roots and above-ground litter which are not associated with mineral particles. It is generally of low molecular weight (<1.7 g cm-3) and has a higher carbon-to-nitrogen ratio than the heavier organic matter fraction (HOMF) (Compton and Boone, 2002). This definition for LOMF is liable to include a range of organic materials. It has been equated with “partially humified organic matter” (Janzen, 1987), particulate organic matter (Compton and Boone, 2002), or inter-aggregate particulate organic matter (Six et al., 1998). Despite its higher carbon-to-nitrogen ratio, LOMF tends to decompose quickly (Whalen et al., 2000) than HOMF because the mineral-associated heavy fraction is often occluded within aggregates and protected from microbial decay, whereas the LOMF is free and unprotected in inter-aggregate pores (Six et al., 1998). LOMF is sensitive to management (Compton and Boone, 2002) and has been observed to be in decline in managed soils (Campbell and Souster, 1982; Six et al., 1998). In semi-arid grasslands, litter breaks down more slowly in the soil than under moister conditions (Gregorich and Janzen, 1996a) and the LOMF content of grassland soils ranges from 0.1-9.7% (Molloy et al., 1977; Sollins et al., 1983; Strickland and Sollins, 1987; Gregorich et al., 1994). Due to its high carbon-tonitrogen ratio and high concentration of carbon, LOMF can contain a substantial proportion of the total carbon. Based on previous studies, Carter (2002b) estimated that 3-45% of organic carbon could be contained in LOMF, while Gregorich and Janzen (1996a) cite studies with values ranging from 2239%. The LOMF serves as a nitrogen repository and source (Boone, 1994; Compton and Boone, 2002), making it a source of nutrients as well as energy. (Gregorich and Janzen, 1996b) present ranges of total 25  nitrogen of 8-29%. Changes in soil organic matter may be difficult to detect when background levels are high, so measuring a major attribute may reveal differences not visible against the backdrop of the whole (Carter, 2002b); hence the inclusion of LOMF in the data set. As a component of soil organic matter, LOMF can be expected to show the same diversity of grazing effects. On a loamy Black Chernozem in northern Alberta at a sub-humid (450 mm MAP) site, Dormaar et al. (1989) hypothesized that any beneficial grazing effects on organic matter through trampling of litter into the soil could be indicated by increases of LOMF, but no evidence for this was found. Instead, when sample in May, the proportion of carbon present as LOMF was 20% lower under 4 years of a 2.6 AUM ha-1 grazing treatment relative to an ungrazed control. LOMF is also influenced by soil water and temperature (Six et al., 1998), so the significantly lower soil water content (and thus lower productivity, litterfall, and incorporation) in grazed pastures in Dormaar et al.'s (1989) study may go some way towards explaining this trend. In addition, Dormaar et al. suggested that root development over the growing season would be indicated by an increase in LOMF between spring and fall sampling dates; however, opposite trends in both sampling years resulted from a significant season-by-year interaction. Polysaccharides Polysaccharides are long-chained carbohydrates which comprise part of soil organic matter. They are relatively more labile than recalcitrant organic compounds such as chitin or lignin, and are a product of, and a material useful for, soil microbial populations. Polysaccharides contribute to the formation of soil structure by stabilizing aggregates (Gregorich et al., 1994), especially microaggregates (Dexter, 1988). Manure additions to soil adds organic residue and can stimulate microbial populations and the production of polysaccharides, and thereby foster aggregation. Increases in levels of soil polysaccharides in the Ah horizon of 9, 14, and 30% under grazing treatments of 0.4, 0.5, and 0.6 AUM ha-1 relative to ungrazed pasture have been reported by Smoliak et al. (1972). The authors noted that they “paralleled the amounts of manure deposited by sheep during the study”. Monosaccharides were found by Dormaar et al. (1997) to have increased by 8% in the top 15 cm of a clay loam Solonetz after 5 years of cattle grazing (0.39 AUM ha-1) in southern Alberta (335 mm MAP). They recorded no significant vegetation change between treatments and it is not likely that this change was due to manure deposition because of the low stocking rate. Dormaar and Willms (1998) found that a light (1.2 AUM ha-1) grazing treatment decreased did not result in significantly different levels of monosaccharides 26  relative to an ungrazed control, but heavy (2.4 AUM ha-1) and very heavy (4.8 AUM ha-1) led to a decrease of 23 and 47%, respectively. The authors did not attribute this effect exclusively to grazing, writing that that “the effects of grazing on monosaccharide content per se may not be the single cause”. Polysaccharide content was included in the data set of my study due to its relationship to aggregate stability and because, similar to the light organic matter fraction, polysaccharide analysis may reveal trends not detectable in total carbon and nitrogen. Cation Exchange Capacity Cation exchange capacity (CEC) represents the soil's ability to hold and release cations used as nutrients by plants and soil fauna. CEC influences the soil's buffering capacity, which enables soil to better withstand disturbances to soil pH such as urine and feces deposition. CEC is influenced by the amount and type of clay, the amount and type of organic matter, and soil pH. Clay minerals of the 2:1 type such as vermiculite, illite, and smectite have a greater capacity to adsorb cations than 1:1 type clays such as kaolinite due to their larger surface area and greater negative charge (Brady and Weil, 2007). The CEC of soil organic matter is high and ranges from 150 to 250 cmol c kg-1; however, this capacity is variable unlike that of clay particles because of the tendency for exchange sites on organic matter to saturate with hydrogen ions at low pH, thereby reducing the total number of sites (Brady and Weil, 2007). Soil texture is a stable soil property generally not affected by management practices, and so unless there is a major change in the amount of organic matter and the pH, CEC is unlikely to change. Exchangeable cations such as calcium, potassium, and magnesium are essential nutrients for higher plants and therefore biomass removal may decrease the amounts of these cations in the soil. Sodium is a biologically important element which affects water balance in organisms as well as soil physical properties like flocculation or dispersal. Andrews and Carroll (2001) suggested CEC as a component of an MDS based on a study by Olson et al. (1996) where CEC was presented as a good candidate for regional comparisons due to relatively low regional variability. Larson and Pierce (1991) recommended nutrient supply, including measures of supply capacity, as a candidate for an MDS. Smoliak et al. (1972) found that CEC was unchanged after 19 years of season-long (May 1 to November 1) light (0.4-0.6 AUM ha-1) sheep grazing on a Solonetz at a dry (310 mm MAP) site in Alberta. They also measured additional properties that influence CEC. Clay content was similar across grazing treatments, but total soil carbon and pH increased and decreased, respectively, with grazing intensity. Smoliak et al. (1972) found that grazing significantly decreased the amount of exchangeable 27  calcium and sodium, but that potassium was unaffected. In their long-term study in southwestern Alberta, Dormaar and Willms (1998) found that calcium and magnesium did not change with grazing, but sodium increased significantly at every level of grazing. Results for potassium were similar in the exclosure and under light grazing, but significantly higher under both heavy and very heavy grazing. This was attributed in part to the higher urine load, but it is also possible that reduced plant growth and nutrient uptake may result in higher cation concentrations in grazed pastures. Soil Reaction Soil pH's reputation as a “master variable” (Brady and Weil, 2007) makes a strong case for its inclusion in any data set: it affects nutrient availability through its influence on the solubility of compounds and the behaviour of ions on exchange sites (McLean, 1982). Soil reaction is affected by a range of influences such as season, acid rain, soil organic matter, and soil biological activity (Smith and Doran, 1996). Effects of livestock grazing on soil pH are varied. In their review, Milchunas and Lauenroth (1993) found no consistent trends for soil pH under grazing treatments. In soils of semi-arid climates formed on calcareous parent material where carbonate content increases with depth, an increase in soil reaction over time can indicate surface erosion by wind and water (Donkor et al., 2002). Dormaar and Willms (1998) found a significant increase in pH in the Ah horizon with increasing grazing intensity, attributing the change to surface soil loss. On a Black Chernozem in Alberta, Johnston et al. (1971) found that pH (in water) was significantly higher under very heavy (5 AUM ha -1) grazing at 6.2 pH units relative to under light (1.25 AUM ha-1) at 5.7 pH units, but not significantly different under moderate (1.7 AUM ha-1) and heavy (2.5 AUM ha-1) grazing pressure. Mapfumo et al. (2000) found that three years of light (0.5 AUM ha-1) and medium (0.67 AUM ha-1) grazing increased the pH significantly in the top 15 cm relative to pre-grazing. This study reported only changes in pH rather than measured values, and the significant increases were 0.11 and 0.07 for the 0-5 and 5-15-cm depths for light grazing, and 0.02 and 0.05 for the 0-5 and 5-15-cm depths for medium grazing. Heavy (1.25 AUM ha-1) grazing led to significantly increased (by 0.08) pH in the 0-5-cm depth and significantly decreased (by 0.17) pH in the 5-15-cm depth. The increase in pH in the 0-5-cm depth were all statistically similar, while heavy grazing resulted in a significantly reduction min pH relative to medium and light grazing. In this study significant increases in soil nitrogen of between 0.043 and 0.068 g 100 g-1 were found in the top 15 cm of the soil under all grazing treatments (except under medium grazing in the 0-5-cm 28  depth), and the authors cited nitrification of NH4+-N from manure and urine deposition as the cause of pH change. Soil reaction did not vary by more than 0.2 units, a range over which the authors felt that plant growth and soil chemical and microbial processes are not affected. Smoliak et al. (1972) reported decreased pH in the surface horizon under heavy grazing at a mixed prairie site. On Chernozems in Saskatchewan, Lodge (1954) found that pH in the Ah horizon did not vary due to a grazing treatment. Available Phosphorus In general, the phosphorus content of soils is low and much of it is unavailable to plants (approximately 0.01% of the total phosphorus pool is available) (Brady and Weil, 2007). In productive natural ecosystems, phosphorus is returned to the soil through decomposition of plant litterfall. In agricultural systems, soil phosphorus is removed by harvest and deficiencies are possible where no effort is made to replenish levels through fertilizer or manure additions. In rangeland environments, phosphorus is removed from the soil-plant system by surface wind and water erosion on degraded sites (Dormaar and Willms, 1998). The amount of phosphorus lost through grazing of vegetation and the removal of livestock is related to the increase in animal weight while on the range, and is likely small. Johnston et al. (1971) suggested that herbivory should increase available phosphorus in the soil through plant biomass removal, and consequently lower nutrient uptake. Some phosphorus is added to the soil through urination and defecation: the average cow produces 25.4 kg of feces (0.18% P 2O5) and 9.1 kg of urine (0.01% P2O5) daily (Petersen et al., 1956). At high stocking rates and over longer grazing times, the distribution of animal manure is more equal, but at lower stocking rates and over shorter time spans, manure effects on soil fertility are more variable and not dependable as a fertilizer (Petersen et al., 1956). Phosphorus availability is further reduced by its fixation under slightly acidic (pH of ~6.5) conditions by mainly oxides or hydroxides of iron, aluminium, or manganese, with smaller amounts fixed by silicate minerals and calcium (Brady and Weil, 2007). This indicator falls under the “nutrient supply” category identified by Larson and Pierce (1991) for inclusion in an MDS. Milchunas and Lauenroth (1993) found no consistent trends for soil phosphorus under grazing treatments in their review. On a Chernozem in Alberta, available phosphorus in the Ah horizon was found to have increased by 25% under a heavy (2.4 AUM ha -1) grazing treatment relative to a ungrazed control by Dormaar and Willms (1998). On a Solonetz in Alberta, Dormaar et al. (1997) reported an increase of available phosphorus in the Ah horizon after six years of grazing at close to the recommended stocking rate (0.39 AUM ha-1). The change in this nutrient, along with other soil fertility 29  variables in their study, was not paralleled by a change in vegetation. On loam to silt loam Chernozems in southwestern Saskatchewan (400 mm MAP) mixed prairie, Lodge (1954) found that under “light” and “accepted” stocking rates, available phosphorus was higher under grazed relative to ungrazed treatments in two of four sites, while it was not significantly different in one and significantly lower in another. Lodge suggested that available phosphorus could be reduced under grazing pressure, but that this effect may also be modified by soil water content: sites with lower water content exhibited increased available phosphorus, perhaps due to decreased uptake by plants. Despite their suggestion that herbivory can increase available soil phosphorus, Johnston et al. (1971) found no significant differences between grazing treatments for available phosphorus. On an Alberta Solonetz, Smoliak et al. (1972) also found no difference between grazing treatments for available phosphorus.  1.3.2 Grazing Effects on Soil Biological Quality Root Biomass Roots, as the repository of the majority (~90%) of vegetative carbon and nitrogen in grasses (Schuman et al., 1999), are an important source of material contributing to the formation of soil organic matter. Decomposed dead root material, root exudates, and also roots themselves aid in the formation of stable soil aggregates. Theoretically, root biomass is expected to decrease in areas affected by herbivory. In response to above-ground biomass removal, plants allocate more resources to regenerating lost photosynthetic production potential in their leaves and thus root production declines (Holland and Detling, 1990). Biondini et al. (1998) reported a significant decrease in root biomass under heavy grazing and a significant increase in root decomposition under moderate grazing. Rooting depth rather than root biomass has been suggested as a component of an MDS (Doran and Parkin, 1994), but root biomass was included in the data set of my study due to its strong relationship with plant productivity as well as other relationships with soil compaction and aggregate stability. A change in plant species composition due to grazing can also result in differences in root biomass between grazed and ungrazed treatments. On undisturbed semi-arid grassland, native vegetation roots deeply, enabling the plant community to withstand periodic droughts. Native species may be replaced by shallow-rooted pioneer species (annuals and short-lived perennials) due to grazing (Naeth et al., 1990; Greenwood and Hutchinson, 1998; Snyman, 2005). In Dormaar et al.'s (1977) study on a Black Chernozem in Alberta fescue grasslands, heavy grazing resulted in an increase in root mass in the top  30  15 cm from 17.2 dry matter tons ha -1 on ungrazed pasture to 25.3 dry matter tons ha -1 (a 60% increase) on heavily (5 AUM ha-1) grazed pasture. This increase correlated with a change in dominant vegetation from rough fescue to forbs and shrubs. At a mixed-grass prairie site in North Dakota, Lorenz and Rogler (1967) reported that 45 years of heavy grazing (1.1 AUM ha-1) resulted in a greater percentage of root mass occurring above 15 cm relative to moderate grazing (0.4 AUM ha -1) on a fine silt Dark Brown Chernozem (Haploboroll). They attributed this increase to increased blue grama cover. In southeast Australia, Greenwood and Hutchinson (1998) found that root biomass decreased deep in the soil and increased at the surface relative to ungrazed pasture after 32 years of heavy sheep grazing (20 sheep ha-1). They suggested this could have been due to a change in botanical composition. On a Solonetz in Alberta, Smoliak et al. (1972) reported that root biomass (i.e., “weight of below-ground plant parts”) increased significantly by 27% under a moderate (0.5 AUM ha -1) grazing treatment and by 61% under a heavy (0.6 AUM ha-1) grazing treatment in the top 15 cm. However, from 15-30 cm they also reported non-significant decreases under these two grazing rates of 10 and 6%. They attributed this to changes in vegetative cover (e.g., increases in blue grama and little club moss cover) as well as to manure additions to the soil surface. Milchunas and Lauenroth (1993), in their global review, suggested that there is more of a positive response of root biomass to grazing than a negative one, although they did not specify at what depth. They also found no relationship between root biomass and species dissimilarity between grazed and ungrazed plots. Although differences in above-ground net primary productivity (ANPP) have been related to differences in species composition, differences in root biomass did not always relate to differences in ANPP. In fact, they found that in 61% of the sites where grazing had negative effects on ANPP, there were positive effects on root biomass. In some individual studies, the findings are no clearer. On a Dark Brown Chernozem in Wyoming, Schuman et al. (1999) found that a light (0.6 AUM ha-1) grazing treatment resulted in a significant (α=0.10) reduction in the 0-15-cm depth in root biomass of 31% relative to an exclosure, but a heavy (2 AUM ha -1) grazing treatment resulted in a nonsignificant decline of 13%. Their finding of increased blue grama in heavily grazed pastures was supported by higher root:shoot biomass ratios in these pastures, but not by the root biomass data for the 0-60-cm depth, which were not significantly different between all grazing treatments. Root growth is restricted in compacted soils, but in none of the studies cited here were root biomass data related to data for bulk density or mechanical resistance.  31  1.3.3 Grazing Effects on Soil Physical Quality Bulk Density Bulk density is a measure of the mass of soil solids relative to total soil volume. Bulk density is an important addition to a data set (Larson and Pierce, 1991) because measurements of adsorbed nutrients can vary with the density of the soil. To produce results which are comparable between different locations, bulk density should be used to express the content of nutrients on a volumetric (i.e., in units of g m-3 or g m-2 over a specified depth) rather than gravimetric basis (Doran and Parkin, 1994). It can also be used as a proxy for the most common type of soil degradation, compaction, as well as for soil properties such as aeration and pore size distribution. Studies have investigated the effects of time of grazing and type of grazing system on soil properties, but there is generally more of the latter. In arid grassland areas, early grazing has a more severe effect on soil bulk density than grazing later in the season because soils are generally wetter in the spring from snowmelt and immature vegetation is less resilient. Naeth et al. (1990) found that bulk density in the top 7.5 cm was 8% higher under heavy (4.4 AUM ha-1) spring (June 1-30) grazing relative to heavy fall (September 15 to October 15) grazing on a loamy Black Chernozem in northeast Alberta (380 mm MAP). Light (1.5 AUM ha-1) spring and light fall grazing treatments were not significantly different. With respect of type of grazing systems, Donkor et al. (2002) found that two years of short-duration moderate (2.08 AUM ha-1) wapiti grazing resulted in bulk density increases of 15% relative to a continuous grazing treatment on a Luvisol in Alberta (~380 mm MAP). Bannerjee et al. (2000), on a Black Chernozem in Manitoba (373 mm MAP), found that bulk density values varied by stocking rate (1.1 and 2.2 AUM ha-1) but did not differ by grazing system (continuous vs. rotational). High stocking rate has been commonly identified as the main factor influencing soil compaction in rangelands. Small hoof size relative to body mass and additional force during movement enable cattle to exert up to 200 kPa on the soil surface when walking, which is substantially higher than the pressure under tractor tires (30-150 kPa) (Donkor et al., 2002). Application of stresses above a soil's supporting capacity can deform the soil, reducing porosity, destroying structure, and increasing bulk density. Soil compaction due to livestock grazing is often most visible in the surface horizons (up to a 30-cm depth) where bulk density is naturally lowest and the effects of animal trampling are strongest (Naeth et al., 1990). On a loamy Black Chernozem in northeast Alberta (380 mm MAP), Naeth et al. (1990) found that very heavy (4.8 AUM ha-1) grazing resulted in bulk density values 20% higher than the ungrazed 32  control in the 0-7.5-cm layer, while heavy (2.4 AUM ha -1) grazing was 11% higher and light (1.6 AUM ha-1) was 7% higher. On another loamy Black Chernozem, Dormaar et al. (1989) found that heavy grazing (2.6 AUM ha-1, which was triple the recommended grazing rate) increased bulk density by up to 15% more than the ungrazed control. Finer-textured soils are more susceptible to compaction because destroyed aggregates in these soils decompose into smaller particles which can pack together more tightly. Van Haveren (1983) found that a series of grazing intensities (0, 0.6, 0.8, and 1.4 AUM ha-1) resulted in bulk density increases (up to 13%) for fine-textured loam soils but not for coarsetextured sandy loam soils. Orr (1960), in South Dakota, found the greatest difference in bulk density values between heavily grazed and ungrazed pasture on sites with the highest silt and clay content. Vegetation and soil organic matter may contribute a cushioning effect depending on their amount and distribution, and susceptibility to compaction may increase where plant and soil organic matter have been reduced by grazing (Naeth et al., 1990). In their long-term (44 years) study on a Black Chernozem, Dormaar and Willms (1998) found that bulk density was higher and total carbon was lower under heavy (2.4 AUM ha-1) and very heavy (4.8 AUM ha -1) cattle grazing compared to light (1.2 AUM ha-1) grazing and the ungrazed exclosure. Significant differences may also be found below the surface horizon. In southwestern Alberta (550 mm MAP), Naeth et al. (1990) found that 36 years of cattle grazing at 4.8 AUM ha-1 increased bulk density values (up to 20% greater than the control up to a 7.5cm depth) on a Black Chernozem relative to stocking rates of 0, 1.2, 1.6, 2.4 AUM ha -1. They attributed the increase in the top 15 cm under heavy and very heavy grazing to the influence of animal treading, but at depths of 55 and 60 cm, they considered the effects of species composition change (i.e., from deep- to short-rooted species) more significant. Mechanical Resistance Soil mechanical resistance has been found to be a more sensitive indicator of compaction than bulk density in rangeland ecosystems (Naeth and Chanasyk, 1995a; Rodd et al., 1999). Due to ease of measurement, more samples and thus a data set more representative of the natural variability of rangeland soils are possible. Mechanical resistance increases under compacted conditions and root extension decreases at values higher than 2500 kPa (Taylor et al., 1966; Greacen et al., 1969; Busscher et al., 1986). With this same upper limit in view, Carter (2002b) classified mechanical resistance ranges into limiting-low (<500 kPa), optimum (500-2000 kPa), and limiting-high (>2500 kPa). Naeth et al. (1990) investigated the effect of different grazing times and reported that the mechanical 33  resistance of a Solonetzic soil at a mixed-grass prairie site was significantly different between the control, early season, and late season grazing (with early 11% higher than late). In the same study, on a Black Chernozem at a parkland fescue site, mechanical resistance in the control plot was not significantly different from light June grazing at the 0-2.5-cm and the 5-30-cm depths. Heavy June grazing resulted in significantly higher resistance at the surface relative to light June grazing. There were, however, no significant differences between light and heavy autumn grazing, which indicated higher susceptibility to compaction in the spring. The authors attributed this elevated susceptibility to higher soil water content in the spring as well as higher soil organic matter content of the autumngrazed pastures (Naeth et al., 1990). As with bulk density, mechanical resistance is most affected by the stocking rate. Mapfumo et al. (1999) found that mechanical resistance on perennial (meadow bromegrass) pastures increased under heavy grazing (by up to 90%) in the 0-5-cm depth and under medium grazing (up to 55%) in the 5-15cm depth relative to light grazing. Donkor et al. (2002) found that mechanical resistance was higher under grazed treatments than in the ungrazed control at all sampling times and at almost all sampling depths. Naeth et al. (1990) examined a Black Chernozem at a foothills fescue site which exhibited 113, 81, 68, and 56% higher soil mechanical resistance in the 0-10-cm depth under very heavy, heavy, moderate, and light (4.8, 2.4, 1.6, and 1.2 AUM ha -1, respectively) grazing treatments relative to the ungrazed control. Aggregate Stability Primary soil particles conglomerate into formations known as peds or aggregates. A well-aggregated soil has a good proportion of large, medium, and small pores, which ensures good gas exchange and water drainage. Aggregates form under the influence of freeze/thaw action, faunal and root activity, the adhesive properties of clay particles, and various types of organic compounds. Tisdall and Oades (1982) identified three types of organic binding agents: persistent, temporary, and transient. Persistent binding agents consist of humic acids, fulvic acids, and humin, the the main components of humus. These compounds enter into microaggregate organo-mineral complexes with amorphous aluminium and iron oxides in the 20-250-μm size range, which may involve 52-98% of soil organic matter. Temporary binding agents consist of roots and fungal hyphae, which may persist for a few years and contribute to the formation of macroaggregates (>250 μm). Plant and microbial polysaccharides and gums comprise the transient binding agents. These are effective for only a very short time (weeks to 34  months) and are associated with both micro- (Dexter, 1988) and macroaggregates (Tisdall and Oades, 1982). Macroaggregates are composed of microaggregates and represent a higher and more complex hierarchical order. They generally have higher porosity, weaker internal bonds, and are therefore less stable (Zhang, 1994). When disturbed, larger aggregates are destroyed more easily than smaller ones, although Dexter (1988) suggests that it is more accurate to speak of the transformation of one hierarchical order into another rather than the “destruction” of aggregates. The arrangement of the components of aggregates (e.g., microaggregates, or organic matter and soil particles) and the nature and number of contact points (i.e., bonds) between them determines aggregate strength. Colloidal water films on component surfaces thin and retreat towards contact points when aggregates dry. As water evaporates, components are drawn closer together, and aggregate strength increases as the number of contact points and solute (e.g., silica, carbonates, or particulate organic matter) deposition increase (Amézketa, 1999). The retreating menisci also draw out water from inter-aggregate pores, strengthening aggregates further (Dexter, 1988). This effect is more pronounced where distances between components are small (e.g., in the smallest aggregate size fraction). The determination of aggregate stability by the wet-sieving method subjects aggregates to the physical stresses of slaking as well as mechanical action (Haynes and Swift, 1990). Slaking refers to the tendency of soil aggregates to disintegrate when immersed in water and is due to a number of factors. Intrusion of water into shear fractures created by differential shrinkage during air-drying (e.g., in the field) and the subsequent expansion of these surface water films forces aggregate components apart. Water is also drawn into smaller capillaries and entrapped air is pressurized, which results in "explosions" which help to break down the aggregate (Yoder, 1936). Decreasing aggregate stability with increasing water content at the time of sampling has been reported in a number of studies (Haynes and Swift, 1990; Churchman and Tate, 1987; Hermawan and Bomke, 1996; Krzic et al., 2000; Wallace et al., 2009), presumably due to wetting rates of dry and moist aggregates during the wet-sieving process. Relatively moister aggregates already have some water intrusion and are slightly hydrophilic, so rehydration occurs rapidly. Rapid re-wetting concentrates the energy of rehydration and leads to increased aggregate disruption. Components in dry aggregates are less prone to fractures caused by wetting-drying cycles and are packed more closely together. They may exhibit more hydrophobicity and re-wet more slowly. Thus they are relatively resistant to the destructive effects of slaking and the mechanical action of the wet-sieving process, and maintain more stability than moist aggregates. 35  Aggregate chemical characteristics may also play a role in stability measurements. Aggregates of pastures with a higher organic matter content have been shown to strengthen more following air-drying than aggregates of arable land. In addition to its bond strength, organic matter becomes hydrophobic when dried and therefore decreases the rate of re-wetting, dissipating rehydration energy over a longer time (Haynes and Swift, 1990). Aggregate stability is a useful indicator due to its ability to integrate physical, chemical, and biological processes. Aggregates are destroyed when pressure is applied to soil, and so can serve as an indicator of compaction. Both temporary and transient organic binding agents are strongly affected by soil management (Karlen and Stott, 1994). Soil organic matter quality can thus serve as a proxy for soil aggregate quality. Conversely, since so much of soil organic matter can be contained within aggregates, the quality and stability of aggregates can influence the quality of soil organic matter. Bird et al. (2002) found that the carbon-to-nitrogen ratio correlated well (r 2=0.55) with sand-free <250 μm aggregate stability on semi-arid rangeland in New Mexico. Aggregate stability has been shown to decrease under grazing pressure, especially at higher stocking rates. Knoll and Hopkins (1959) found that aggregate stability on a silty clay loam was reduced by grazing: on ungrazed, moderately grazed (2.5 AUM ha -1), and heavily grazed (9.9 AUM ha-1) sites the proportion of aggregates larger than 0.5 mm was 89, 64, and 56 percent, respectively (a 40% decrease between ungrazed and heavily grazed treatments). In their long-term study, Dormaar and Willms (1998) found that heavy (2.4 AUM ha-1) and very heavy (4.8 AUM ha-1) grazing significantly reduced the mean weight diameter (MWD) of aggregates by 15 and 20% compared to the ungrazed control, but light (1.2 AUM ha-1) grazing did not. Total carbon decreased as grazing intensity increased and results followed the same pattern as the MWD results. The carbon-to-nitrogen ratios for all grazing levels were 13.4 (exclosure), 12.6 (light), 8.9 (heavy), and 7.1 (very heavy), but whether or not these differences were significant was not presented. The non-significant difference in aggregate stability in Dormaar and Willm's study between the exclosures and the light grazing treatment supports the notion mentioned previously that rangeland ecosystems may have evolved with some level of grazing disturbance from native herbivores and thus may not show negative consequences at stocking rates similar to ancient grazing pressures (most likely light grazing). Previous work (Lamagna, 2008) on soil aggregates in the Southern Interior of BC that also included some sites on the Lac du Bois rangelands revealed particular behaviours of aggregate size fractions.  36  Polysaccharides correlated most strongly with the 0.25-1-mm size fraction, followed by the 2-6-mm size fraction and the MWD. Although the 1-2-mm aggregate size fraction correlated well with litter cover, Sandberg's bluegrass cover, exposed mineral soil, and total carbon and nitrogen, it did not correlate well with polysaccharides. Lamagna, along with other studies on soil quality in the Southern Interior of BC (Broersma et al., 2000; Krzic et al., 2000; Wallace et al., 2009), found that the 1-2-mm aggregate size class comprised the smallest proportion of the total sample at 15%. Based on these observations, Lamagna suggested that in loamy soils this class was likely formed from the destruction of larger aggregates and was transient in nature.  1.3.4 Grazing Effects on Vegetation Determining the status of rangeland vegetation is critical to ranch operation because the main product of rangelands is animal protein, which depends almost exclusively on forage. Ranch productivity is based on net primary productivity of rangeland vegetation and stocking rates are often calculated to make use of a certain proportion of this (e.g., 50% forage utilization). Cattle can consume up to 100 kg of fresh plant matter per day (Bilotta et al., 2007), so in marginally productive environments such as semi-arid grasslands, knowing the condition of rangeland vegetation is central to understanding range health. Dormaar et al. (1997) suggested that an assessment of range quality is not complete without an assessment of treatment effects on plants and plant productivity. Vegetation has a strong influence on soil quality and vice-versa. Vegetation protects the soil surface from the effects of treading, reduces soil erosion, and improves surface water quality (Bilotta et al., 2007). Plant litter that accumulates as a mulch on the surface can conserve soil water by shading and cooling the soil and can also delay plant growth in the spring due to lower soil temperature (Willms et al., 1986). Plants can affect soil chemical quality through the decomposition of their litter and roots. The loss of deeply-rooted vegetation can influence the distribution of soil organic matter within the soil profile. Vegetation can respond in various ways to the direct effects of herbivory (e.g., plant tissue removal) by grazing animals. McNaughton (1983) identified three hypotheses about plant responses to herbivory. The first takes the view that herbivory is detrimental to the plant due to removal of plant tissue and reduction of photosynthetic capacity. The second allows that plants can compensate for low levels of herbivory (Holland and Detling, 1990), at least up to a certain point beyond which the disturbance becomes detrimental. The third hypothesis suggests that plants can overcompensate for losses due to 37  moderate herbivory, so that at low or moderate grazing levels there may actually be an increase in plant productivity (McNaughton, 1983). Plants can also be affected indirectly by grazing through modification of local conditions. For example, compaction of the soil through hoof treading can restrict root growth in dense soils and lead to lowered biomass production (Donkor et al., 2002). The removal of a protective canopy of vegetation may increase evaporation of water from the soil surface, and it may reduce soil water losses due to reduced transpiration by plants (Proffitt et al., 1993). Any treatment effects on vegetation should be situated within their proper climatic context: in xeric environments where the most limiting resource is water, plant productivity may be affected more by the amount of rainfall than by grazing (Biondini et al., 1998). Responses to herbivory vary between species and types of plant, depending on evolutionary history and competitive ability. Prairie grasses evolved under grazing pressure from bison and are highly resilient to herbivory. However, the bunchgrasses of the southern BC interior experienced no such intense selection pressure. Caldwell et al. (1981) compared the responses to herbivory of two semi-arid bunchgrasses, bluebunch wheatgrass and desert wheatgrass (Agropyron cristatum ssp. desertorum (Fisch. ex Link) A. Löve). Desert wheatgrass was able to re-establish a canopy three to five times larger than that of bluebunch wheatgrass, invest smaller amounts of nitrogen in new tissue, produce shorterlived stems, and allocate more resources to shoots and less to roots following defoliation, while bluebunch wheatgrass continued root growth even after defoliation. These features in desert wheatgrass contribute to a higher resistance to herbivory. Bluebunch wheatgrass is much more sensitive: if grazed often during its critical growth period in the spring, it can take years to recover (van Ryswyk and McLean, 1989). Milchunas and Lauenroth (1993) found that the biggest decreases in dominant species cover under grazing pressure were observed when that species was a bunchgrass. In southwestern Montana (350 mm MAP), Evanko and Peterson (1955) found that bluebunch wheatgrass cover was significantly reduced under a heavy grazing treatment, but not under moderate or light. They also found increased cover of Sandberg's bluegrass and, in some pastures, Junegrass, true to these species' reputation as “increasers” (Newman et al., 2008). Rough fescue is reduced rapidly under grazing pressure, but also re-establishes rapidly once protected (van Ryswyk and McLean, 1989). Native species composition can also be affected by the introduction of new forage species. Bluebunch wheatgrass, Sandberg's bluegrass, needle-and-thread grass, and Junegrass cover were found to be lower on pastures which had been seeded with the exotic crested wheatgrass (Krzic et al., 2000). 38  Herbivory can affect the evolutionary development of species as well as their composition. Grazing pressure sustained over a long term selects for plant traits such as lateral spread and rapid growth (McNaughton, 1983). Species composition can be affected by the selectivity of the grazing animal (Bilotta et al., 2007). Although moderate grazing may promote biodiversity in ecosystems which have evolved with moderate herbivory, in sensitive ecosystems such as semi-arid grasslands, preferred species tend to decrease in number while less palatable species increase under grazing pressure. At high stocking rates, preferred plants may be so strongly affected that they essentially disappear and force a change in animal grazing behaviour (Bilotta et al., 2007). The notions that plant communities progress through seral stages and that disturbances such as grazing send them “backwards” to more primitive ecological states have helped formed the basis for range assessments (Fraser, 2007; Delesalle et al., 2009). Above-ground Biomass Plant productivity responses to grazing have been mixed. In a survey of 20 grazing studies and eight clipping studies in western American rangelands, Lacey and Poollen (1981) found that exclosures had higher herbage production (68±46%) than moderately grazed (40-60% forage utilization) pastures. Detling et al. (1979) reported that biomass production of clipped blue grama plants was lower and photosynthetic rates were higher than in unclipped plants. New growth in clipped plants was preferentially allocated to leaves (50%) rather than roots (18%) while in unclipped plants it was 33% and 29%, respectively. Gill (2007) found that long-term grazing significantly reduced biomass production relative to protected exclosures, although this study was done on a subalpine pasture in Utah (932 mm MAP). Milchunas and Lauenroth (1993) found that over a variety of sites and studies grazing most often had a negative effect on above-ground plant productivity, but that in some cases, it had a positive effect. Dormaar et al. (1997), at a mixed prairie site in Alberta, reported that neither productivity nor species composition showed any significant differences under grazing pressure. Species Cover Shifts in species composition can occur without significant changes in biomass production. Biondini and Manske (1996) found that although plant basal cover did not respond to grazing treatments, species composition did. Under grazing pressure, cool-season grass cover was reduced, while warm-season grass and forb cover was increased. Blue grama cover, for example, was higher in grazed areas relative 39  to exclosures. Herbivore preference for rough fescue has been shown by a number of studies where grazed fields showed lower cover of rough fescue while exclosures were dominated by it (Johnston et al., 1971; Dormaar et al., 1977, 1989; Willms et al., 1985). On fescue grasslands (Dormaar et al., 1989) and mixed-grass prairie (Dormaar et al., 1997) the ecosystem was considered to have “regressed” due to the loss of species characteristic of late seral stages. At a semi-arid (446 mm MAP) mixed-grass prairie site in North Dakota, Biondini et al. (1998) found no significant effects of grazing at moderate (50% forage use) and heavy (90% forage use) levels on relative cover of dominant grasses. Rather, they found that ANPP was correlated with rainfall and not grazing intensity and they identified precipitation as a major influence on vegetation responses. In grassland ecosystems, Milchunas and Lauenroth (1993) found that three variables, changes in ANPP, length of evolutionary history of grazing (i.e., plant adaptation to grazing), and level of consumption, were able to explain 54% of the variance of species dissimilarity between grazed and ungrazed sites. ANPP by itself was able to account for 40% of the variance, while in another model, precipitation accounted for 24%, underlining the importance of climatic factors in structuring vegetation communities. Vegetation responses to grazing in the Southern Interior grasslands of BC are influenced by their evolutionary history and the time and intensity of grazing. Low intensity historical grazing and summer drought conditions have resulted in grasses with low resilience to grazing. In the Middle grasslands, cattle prefer to graze bluebunch wheatgrass, Kentucky bluegrass, Junegrass, and needle-and-thread grass. Preferential grazing can increase the occurrence of unpalatable vegetation and lead to differences in species composition between grazed and ungrazed pastures (D. Thompson, personal communication, 2010). In this region range health has been linked to the amount of exposed mineral soil, the percent cover of litter, and the percent cover of three native grasses (Sandberg's bluegrass, Junegrass, and rough fescue) (Newman et al., 2008). The time of grazing relative to plant development can also influence the structure of plant communities (Plate F.1, Appendix E), especially when plants are grazed early in their life-cycles.  1.4 Summary of General Introduction Rangelands comprise fifty percent of the world's land area, and are often located in drier regions that are otherwise unsuitable for rain-fed agriculture. In BC, 20 percent of the 10 million hectares of rangeland are located on the province's sparsely distributed semi-arid grasslands. Although ranching is most obviously focused on forage production and thus on the status of rangeland 40  vegetation, these plants ultimately depend on the roles soil plays as a growth matrix, nutrient and water provider, and biological habitat. Historically, these ecosystems have evolved with some degree of disturbance from herbivory by endemic ungulates, and some studies support the notion that a light grazing intensity is similar to this level of historical disturbance (Naeth et al., 1991; Dormaar and Willms, 1998; Willms et al., 2002; Ganjegunte et al., 2005; Snyman and du Preez, 2005). However, overgrazing by cattle and other animals has caused, and continues to cause range degradation. The removal of preferred vegetation leads to a shift in the makeup of plant communities, but plant communities may re-establish in a relatively short time following disturbance. When the quality of the soil is compromised, amelioration may take longer and negative effects on plant communities may persist over a longer term. Semi-arid rangeland ecosystems may be resilient to drought and low soil water content but not to other disturbances such as trampling, and under these conditions a low stocking rate is preferable. Due to these concerns, the inclusion of soil quality in rangeland health assessments is desirable. Generally, grazing has been found to have negative effects on soil physical quality, increasing bulk density and mechanical resistance while decreasing hydraulic conductivity, infiltration, and aggregate stability. The effects on soil chemical and biological quality are more varied, with increases, decreases, and lack of differences reported for polysaccharides, total carbon and nitrogen content, available phosphorus, pH, cation exchange capacity, and root biomass density. Grazing has mixed effects on biomass productivity and species composition. The current project is unique in three respects. First, it was conducted on a large (~270-hectare) longterm experiment established and maintained by Agriculture and Agri-food Canada, Kamloops, BC, where the objective is to determine the effects of grazing on plant species composition and for which there is a complete record of management since 1978. Second, soil data sets following 20 (1998) and 30 (2008) years of grazing are utilized to illustrate long-term effects on soil quality and enable a comparative approach to assessing any changes over time. Long-term experiments face the challenges of persisting through changing landowners, managers, or, in the case of federal research facilities, governments and their research priorities. Most university research projects themselves are limited by the extent of funding and the length of the graduate program, and rarely go beyond five years in length, while management effects can take years or decades to manifest themselves. Third, although numerous studies have been conducted on rangeland soil quality in Alberta (e.g., Dormaar et al., 1977; Naeth et al., 1991; Naeth and Chanasyk, 1995; Dormaar and Willms, 1998; Mapfumo et al., 2002), Manitoba 41  (e.g., Banerjee et al. 2000), Saskatchewan (e.g., Lodge 1954), as well as in the Great Plains states (e.g., Gamougoun et al., 1984; Abdel-Magid et al., 1987; Frank et al., 1995; Manley et al., 1995; Biondini et al., 1998), there has been very little research published on rangeland soil quality in BC (e.g., Krzic et al. 1999) or rangeland soil quality for the Southern Interior grasslands (e.g., Krzic et al., 2000). The unique nature of southern BC grasslands is expressed, for example, as the low resilience to disturbance exhibited by bunchgrasses. This is due to their adaptation to the low-intensity grazing behaviour of small populations of elk, deer, and bighorn sheep, as opposed to the vigorous Prairie grasses which evolved under the influence of large bison herds and periodic severe trampling. The relationship between vegetation quality and soil quality begs the question of whether southern BC grassland soils are resilient enough to function in a well-managed ranching agroecosystem. This study aims to quantify the long-term effects of time of grazing and stocking rate on these soils.  1.5 Study Objectives and Hypotheses This study examines the effect of cattle grazing on selected soil properties in grasslands in southern BC after 20 and 30 years. This project was guided by the following objectives and associated hypotheses. Objective 1: Compare the effects of spring and fall grazing treatments on selected grassland soil properties after periods of 20 and 30 years. Hypothesis under Objective 1: Spring grazing will more strongly affect selected soil properties than fall grazing, and these differences will be more pronounced 30 years after the establishment of the treatment relative to 20 years. Objective 2: Compare the effects of 0 and 2 AUM ha-1 grazing rates on selected grassland soil properties after periods of 20 and 30 years. Hypothesis under Objective 2: Grazing at a rate of 2 AUM ha -1 will more strongly affect selected soil properties than a rate of 0 AUM ha-1, and these differences will be more pronounced 30 years after the establishment of the treatment relative to 20 years. 42  The long-term goals of this study are (1) to add to the body of knowledge in rangeland ecology on grazing impacts on the soil quality of BC's Southern Interior grasslands, and (2) to contribute to the establishment of a sustainable rangeland management system for BC's Southern Interior grasslands. As demand for local range-fed beef increases, these rangelands will experience increased production pressures, and as urban development around and perhaps on these rangelands continues, the pressures of development and recreational use will also grow. Studies like this one can help develop policies and management strategies that will reconcile the short-term demands of urbanization and food production with the long-term goal of sustaining the environments of the Southern Interior grasslands and the communities, both human and animal, that depend on them.  43  Chapter 2- Grazing Effects on Selected Grassland Soil Properties in Southern British Columbia after a Period of 20 and 30 Years 2.1 Introduction Grasslands are present in a large part of central North America, but those found in British Columbia (BC) are unique in a number of respects. Historically, these bunchgrass-dominated grasslands have developed under the low-intensity grazing of small populations of elk, deer, and bighorn sheep, as opposed to the vigorous Prairie grasslands, which evolved under the influence of large bison herds and periodic severe trampling. While grasslands in BC cover only 1.4% of the provincial land area, they contain almost 20% of wildlife species of concern (BC Ministry of Environment, Lands and Parks, 1998). British Columbia's grasslands have faced a number of challenges to their long-term sustainability, including inappropriate land management, urban development pressure, as well as inappropriate grazing times and stocking rates. Stocking rate recommendations in BC have historically relied on indicators of plant health, such as percent cover or stubble height, while soil properties have been overlooked. Similarly, the effects of different intensities of grazing and different grazing times on soil properties of grasslands in BC has not been well documented. Studies elsewhere have investigated grazing effects (both time of grazing and stocking rate treatments) on soil properties. Time of grazing was the focus of several studies carried out on grasslands in southern Alberta. On a Black Chernozem, soil organic matter was significantly lower (26%) under a heavy (4.4 AUM ha-1) autumn treatment compared to a light (1.5 AUM ha -1) autumn grazing treatment (Naeth et al., 1991). At the same site in a different study, bulk density in the 0-7.5-cm depth was found to be 20% higher relative to an ungrazed control under a heavy (4.4 AUM ha -1) June grazing treatment, but only 7, 8, and 11% higher under light (1.5 AUM ha -1) autumn, light June, and heavy autumn grazing treatments, respectively (Naeth et al., 1990). In the same study by Naeth et al. (1990), but on a Solonetz at a different site, soil mechanical resistance was found to be higher under grazing treatments relative to an ungrazed control, with early-season grazing (0.9 AUM ha -1) being 13% higher than late-season (0.9 AUM ha-1). There is a substantial amount of research into the response of soil properties under stocking rate treatments at semi-arid to sub-humid grassland sites, but it is largely from outside of BC. Naeth et al.  44  (1991) found that soil organic matter in the Ah horizon of a Black Chernozem decreased by 13 and 19% due to 36 years of continuous cattle grazing at stocking rates of 2.4 and 4.8 AUM ha -1 compared to ungrazed exclosures. On a Dark Brown Chernozem in Wyoming, Schuman et al. (1999) found that in the top 30 cm total soil carbon was 22% higher after 12 years of season-long cattle grazing at 2 AUM ha-1 relative to ungrazed pasture. On a Dark Brown Chernozem in North Dakota, Frank et al. (1995) reported a 25% decrease in total soil nitrogen over 75 years under a heavy (1.1 AUM ha -1) cattle grazing treatment relative to an ungrazed control. In a clipping study from a semi-arid (560 mm MAP) region of South Africa, Snyman and du Preez (2005) reported no significant differences for carbon-tonitrogen ratios between various rangeland conditions (poor, moderate, and good). On a Solonetz in Alberta, Smoliak et al. (1972) reported that 19 years of season-long (May 1 to November 1) sheep grazing at a moderate (0.5 AUM ha -1) grazing rate increased soil polysaccharides by 14% relative to an ungrazed control, but that CEC was unaffected. On a Black Chernozem in Alberta, Johnston et al. (1971) found that pH (in water) was not significantly different under moderate (1.7 AUM ha -1) and heavy (2.5 AUM ha-1) grazing pressure. Available phosphorus was 25% higher in the Ah horizon under a heavy (2.4 AUM ha-1) grazing treatment relative to an ungrazed control on a Chernozem in Alberta (Dormaar and Willms, 1998). On a fine silt Dark Brown Chernozem (Haploboroll) in North Dakota, Lorenz and Rogler (1967) reported that 45 years of grazing at moderate (0.4 AUM ha -1) and heavy (1.1 AUM ha-1) rates did not result in significant differences in total root weight. On a Black Chernozem in Alberta, Naeth et al. (1990) found bulk density was 11% higher while mechanical resistance was 81% higher under heavy (2.4 AUM ha-1) grazing. Dormaar and Willms (1998) found that heavy (2.4 AUM ha-1) grazing significantly reduced the mean weight diameter (MWD) of aggregates by 15% compared to the ungrazed control, but light (1.2 AUM ha-1) grazing did not. In the southern interior of BC, there are concerns that cattle grazing is negatively impacting soil and water quality, ultimately leading to declines in forage production for wildlife and domestic grazers (Wikeem and Wikeem, 2004). However, no study has evaluated the impacts of either time of grazing or stocking rate on a range of soil properties within these unique ecosystems. This study, the first of its kind in BC, examined the long-term (20 and 30 years) effects of cattle grazing on selected soil properties in the grasslands of the southern interior of the province. The specific objectives of this study were to compare (1) the effects of spring and fall grazing treatments on selected soil properties after periods of 20 and 30 years and (2) the effects of 0 and 2 AUM ha -1 grazing rates on selected soil properties after periods of 20 and 30 years. 45  2.2 Materials and Methods 2.2.1 Site Description The study site was located on the Lac du Bois Range (50° 45’ N and 120° 25’ W) north of Kamloops, BC. This study was carried out within a broader long-term experiment established in 1978 and maintained by Agriculture and Agri-Food Canada, which investigates the impacts of cattle grazing on plant composition. My study focused on four pastures (approximately 65 ha in size), each containing two exclosures (25 × 50 m2, fenced with 1.5 m-high barbed wire) for protection from livestock grazing. Pastures were grazed for one month either in the spring (mid-April to mid-May) or the fall (midSeptember to mid-October) at a rate of 2 AUM ha-1. The four pastures were located along an elevation gradient from 600 to 850 m above sea level on Brown and Dark Brown Chernozems of four different soil series: McKnight, Glimpse, McQueen, and Aylmer (for detailed soil descriptions, see Appendix B). These soils developed from ablation till deposits associated with basic volcanic and limestone bedrock (Young et al., 1992). These soils are characterized by a loam texture (45% sand, 45% silt, and 10% clay) and average coarse fragments (diameter >2 mm) content of 7, 14, and 18% (volume) of the 0-7.5-, 7.5-15-, and 15-30-cm depths, respectively. The region experiences hot, dry summers and moderately cold winters, often with little snowfall. Average annual air temperature in this region is 8.9°C, with an average July temperature of 21°C (Figure D.1, Appendix D) (Government of Canada and Meteorological Service of Canada, 2010). The region is characterized by a semi-arid climate with an average annual precipitation of 270 mm, 32% of which falls between May and July (Figure D.1, Appendix D). The study site is located within the Bunchgrass biogeoclimatic zone, where the dominant vegetation is the association of bluebunch wheatgrass (Pseudoroegneria spicata (Pursh) A. Löve syn. Agropyron spicatum) and basin big sagebrush (Artemisia tridentata Nutt. ssp. Tridentata) (Plate F.2, Appendix E).  2.2.2 Sampling and Analyses Soil samples were taken in early June 1998 and then again in early June 2008. In both years, samples were taken along two 10-m long parallel transects laid out within each study plot. Transects were always laid out at least 3 m away from the exclosure edge to avoid fenceline effects. In both years, soil samples were obtained at two random points along each transect. Identical laboratory procedures were 46  used in both years except where noted. Soil Chemical and Biological Properties Soil samples for chemical analysis were taken from the 0-7.5, 7.5-15, and 15-30-cm depths, air-dried and passed through a 2-mm sieve before performing laboratory analysis. Total carbon (Nelson and Sommers, 1982) and nitrogen (McGill and Figueiredo, 1993) were determined by dry combustion using an automated elemental analyzer (LECO CNS-2000, Leco Corp., St. Joseph, MI). The light organic matter fraction (LOMF) of samples collected in 1998 was isolated by flotation on sodium iodide (1.7 g cm-3) (Gregorich and Ellert, 1993). Soil polysaccharides were determined using the phenol-sulfuric acid method as described by Lowe (1993). Sodium bicarbonate extraction (NaHCO3) (Olsen et al., 1954) was used to determine available phosphorus. Soil cation exchange capacity (CEC) and exchangeable cations (calcium, magnesium, potassium, and sodium) were determined only on samples collected in 1998 by the ammonium acetate (pH 7) method (Soil Survey Laboratory Staff, 2004). Soil pH was determined on a 1:1 (v/v) soil to distilled water slurry and a 1:2 (v/v) soil to 0.01 M CaCl2 slurry (Hendershot and Lalande, 1993). Root biomass was determined on intact soil cores taken from the 0-7.5, 7.5-15, and 15-30-cm depths using a double-cylinder, drop-hammer sampler with a 7.5cm-diameter by 7.5-cm-deep core. In 1998, samples were dispersed using calcium chloride, the soil was wet-sieved over a 1-mm sieve, and the roots remaining on the sieve were dried at 65°C for 48 hours. Ash content was determined by ignition at 550°C for four hours and root biomass is reported on an ash-free, oven-dry basis. In 2008, root biomass was determined using a method for rapid sample processing developed by Metcalfe et al. (2007). The mass of roots extracted over four 10-minute periods was graphed and a logarithmic regression equation was fitted to plot extracted root mass to the point where the predicted extracted mass over one 10-minute period was equal to 1% of the predicted total extracted mass. The accuracy of the logarithmic equation based on four intervals was visually checked using a series of extractions consisting of 12 intervals (120 minutes) (see Figure D.2, Appendix D). The ash content of the roots was determined by ignition at 380°C for twelve hours. This temperature was chosen to not confound the dehydroxylation of phyllosilicates or decarboxylation of carbonates with oxidation of organic matter (Ben-Dor and Banin, 1989). Root biomass is reported on an ash-free, oven-dry basis.  47  Soil Physical Properties Bulk density was determined on intact soil cores taken from 0-7.5, 7.5-15, and 15-30-cm depths using a double-cylinder, drop-hammer sampler with a 7.5-cm-diameter by 7.5-cm-deep core. Cores were ovendried at 105°C for 24 hours. Coarse fragments (diameter >2 mm) within the sample were removed by sieve and weighed. The volume of coarse mineral fragments was determined from their dry mass based on an particle density of 2.65 g cm -3. Bulk density was calculated on a coarse fragment-free basis as the mass of dry soil per volume of field-moist soil (Blake and Hartge, 1986). Soil mechanical resistance (Bradford, 1986) was recorded at 1.5-cm intervals down to a depth of 30 cm using a hand-pushed 13-mm diameter 30° cone penetrometer with a data logger (Agridy Rimik PTY Ltd. Toowoomba, QLD, Australia). Measurements were made at five random locations along each transect for a total of 10 measurements per treatment. As soil mechanical resistance is affected by the soil water content at the time of measurement, soil water content was determined gravimetrically (Gardner, 1965) on separate samples taken from the 0-7.5, 7.5-15, and 15-30-cm depths. Aggregate stability was assessed using a variation of the wet sieving method (Nimmo and Perkins, 2002). Two composite samples, each consisting of five individual subsamples, were collected along each transect from the 0-7.5-cm depth. Samples were stored at 4°C until they were passed through a 6mm sieve and collected on a 2-mm sieve. Immediately before wet sieving, a pre-sieved 2-6-mm sample (of about 15 g) was placed on top of three nested sieves with openings of 2, 1, and 0.25 mm and moistened in a humidifier for 30 minutes to minimize slaking. Wet sieving was performed by a motordriven mechanical device with a vertical stroke of 2.5 cm and an oscillating motion through an angle of 30° for 10 minutes at a rate of 30 strokes per minute. After the sieves were removed from the water, the material retained on each sieve was oven-dried at 105°C for 24 hours and weighed. A correction was made to account for non-aggregate particles: after crushing all material retained on each sieve, nonaggregate particles were weighed and their mass was subtracted from the total size fraction mass to determine the mass of true aggregates. The mass for each size fraction was expressed as a percentage of the total non-aggregate particle-free sample mass. Aggregate stability was expressed as the MWD, which represents the summation across size fractions of the product of the mean diameter of each size fraction (Di) and the proportion of the sample weight occurring in that size fraction (Wi). 4  MWD=∑ W i D i  (Equation 4)  i=l  48  2.2.3 Statistical Analyses Bulk density, aggregate stability, total carbon and nitrogen, polysaccharides, available phosphorus, pH, and root biomass data were analyzed as a split-plot completely randomized design with four replicates and four subsamples (i.e., two locations along two transects). Mechanical resistance data were analyzed as a split-plot completely randomized design with four replicates and ten subsamples (i.e., five locations along two transects). All soil data were analyzed separately for each depth of sampling. Grazing time was the main treatment at the level of the experimental unit (whole plot), while the stocking rate was the treatment at the split-plot level. Mean values were compared using the general linear model and the ANOVA procedure in the statistical software package SAS (SAS Institute Inc., 2003). Results were considered significant at p<0.10. The ANOVA table is presented in Table C.1 in Appendix C.  2.3 Results and Discussion 2.3.1 Grazing Effects on Soil Chemical and Biological Properties Cation Exchange Capacity Twenty years after the inception of the long-term grazing experiment, there were no significant differences in CEC under either the time of grazing or stocking rate treatments in the 0-7.5-cm depth (Table 2.1). Table 2.1: Grazing effects on cation exchange capacity (CEC) in the 0-7.5-cm depth as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC. Year  Depth (cm)  Spring Ungrazed Grazed  Fall Ungrazed  Grazed  Mg m-3 2008 0-7.5 0.86 (0.146) 1.03 (0.110) 0.87 (0.104) 0.93 (0.103) † The bracketed number represents the standard deviation of the mean (n=16).  S-F(UG)  S-F(G)  0.7502  0.0386  (S)UG-G (F)UG-G p  0.0048  0.2136  A significant interaction between the time of grazing and stocking rate treatments was detected in the 7.5-15 and 15-30-cm depths (Table 2.2). Spring-grazed areas had 27 and 35% significantly lower CEC than fall-grazed areas in the 7.5-15 and 15-30-cm depths, respectively. Additionally, spring exclosures had 10 and 11% significantly higher CEC than spring-grazed pastures, while fall grazed pastures were 3 and 8% higher (though not significantly) than fall exclosures in the 7.5-15 and 15-30-cm depths, respectively. 49  Table 2.2: Time of grazing and stocking rate treatment interaction effects on cation exchange capacity (CEC) in the 7.5-15 and 15-30-cm depths as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC. Spring (S) Fall (F) S-F(UG) S-F(G) (S)UG-G (F)UG-G Ungrazed (UG) Grazed (G) Ungrazed (UG) Grazed (G) cmolc kg-1 p 7.5-15 257.1 (29.14)† 234.1 (25.86) 290.8 (45.80) 298.4 (35.30) 0.0036 0.0001 0.0197 0.3391 15-30 241.1 (36.62) 216.5 (36.95) 271.9 (49.17) 293.3 (53.17) 0.0067 0.0001 0.0177 0.0300‡ † The bracketed number represents the standard deviation of the mean (n=16). ‡ Based on four meaningful treatment mean comparisons, a Bonferonni correction of 0.10÷4=0.025 was used for α Depth (cm)  The CEC ranged from 216.5 to 318.0 cmol c kg-1, and was comparable to the CEC reported in a descriptive study carried out on the Lac du Bois Range by van Ryswyk et al. (1966) for Black Chernozems (264 cmolc kg-1), Dark Brown Chernozems (185 cmolc kg-1), and Brown Chernozems (130 cmolc kg-1). My results are different from those of Smoliak et al. (1972), who found that CEC was unchanged after 19 years of season-long (May 1 to November 1) sheep grazing at rates of 0.4, 0.5, and 0.6 AUM ha-1 on a Solonetz at a dry (310 mm mean annual precipitation- MAP) site in Alberta. In their study, clay content, which influences CEC was similar across grazing treatments. However, under the 0.6 AUM ha-1 grazing treatment, total soil carbon increased significantly from 1.10 to 1.38% while soil pH decreased significantly to 5.8 from 6.4 in the ungrazed control. The changes in soil carbon and pH may have resulted in lack of differences in CEC in this study. Precipitation and length of grazing trial were similar to those in the study by Smoliak et al. (1972), and so it appears that the difference in stocking rate and type of livestock between their study and mine are responsible for the different responses to grazing pressure. The occurrence of a significant difference between ungrazed and grazed areas in spring pastures but not in fall pastures suggests that spring-grazed pastures are more susceptible to reductions in CEC due to grazing pressure, which was hypothesized. The significant trend for spring-grazed pastures to exhibit lower CEC than fall-grazed pastures lends additional support to this hypothesis. Additional sampling of above-ground biomass in 1998 revealed a significant interaction: productivity was lower in spring-grazed areas compared to fall-grazed areas and spring and fall exclosures (Table E.1, Appendix E). This pattern has also been observed in an unpublished longterm grazing study on the Lac Du Bois range (D. Thompson, 2010, personal communication). The trend of CEC interaction means is similar to the pattern for above-ground biomass, in that springgrazed areas show the largest reduction. Two of the four exchangeable cations analyzed after 20 years of grazing showed significant differences under the time of grazing treatment: significantly lower exchangeable magnesium (by 32-43%) and 50  sodium (by 82-92%) were found in spring-grazed relative to fall-grazed pastures at all depths while exchangeable calcium and potassium were not significantly different (Table 2.3). Twenty years of grazing at a stocking rate of 2 AUM ha -1 did not result in significant differences in levels of exchangeable cations between grazed and ungrazed areas (Table 2.3). Table 2.3: Grazing effects on exchangeable cations as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC. Depth (cm)  Time of grazing Spring  Fall  Stocking Rate p cmolc kg  0 AUM ha-1  2 AUM ha-1  p  -1  Ca 0-7.5  142.8 (24.92)†  121.0 (29.90)  0.185  131.8 (28.63)  132.0 (30.67)  0.990  7.5-15 15-30  152.3 (40.01)  118.5 (34.66)  0.176  135.7 (37.37)  135.2 (44.65)  0.948  154.5 (68.02)  111.3 (43.73)  0.168  132.2 (55.61)  133.6 (66.38)  0.869  Mg 0-7.5  49.5 (13.55)  72.7 (13.54)  0.010  62.7 (20.73)  59.5 (14.44)  0.613  7.5-15 15-30  51.9 (18.04)  81.0 (16.58)  0.010  69.2 (25.28)  63.7 (19.54)  0.489  55.4 (19.97)  96.6 (32.38)  0.003  77.2 (30.83)  74.7 (37.00)  0.773  K 0-7.5  20.8 (4.55)  24.2 (7.87)  0.304  21.0 (5.39)  24.0 (7.39)  0.330  7.5-15 15-30  19.3 (3.44)  24.1 (8.28)  0.167  20.0 (5.58)  23.4 (7.41)  0.279  18.5 (6.13)  22.2 (7.97)  0.213  19.5 (6.71)  21.2 (7.84)  0.590  0.3 (0.35)  0.9 (2.48)  0.166  7.5-15 0.2 (0.08) 1.8 (4.34) 0.008 0.6 (0.79) 15-30 0.3 (0.10) 3.8 (9.67) 0.016 1.4 (2.39) † The bracketed number represents the standard deviation of the mean (n=32).  1.5 (4.37) 2.6 (9.66)  0.270 0.278  Na 0-7.5  0.2 (0.08)  1.1 (2.46)  0.006  Smoliak et al. (1972) reported significant differences in levels of exchangeable cations resulting from a stocking rate treatment (rather than a time of grazing treatment as seen in my study). Exchangeable calcium and sodium were both higher under light (0.4-0.6 AUM ha -1) grazing pressure, while exchangeable potassium was unchanged. Dormaar and Willms (1998), on a clay loam to loam Black Chernozem at a sub-humid (550 mm MAP) site in Alberta, found that 44 years of season-long (May 15 to November 15) cattle grazing at light, heavy, and very heavy stocking rates (1.2, 2.4, and 4.8 AUM ha-1, respectively) led to significantly higher levels of exchangeable sodium and potassium. Exchangeable sodium was 142, 195, and 347% higher under light, heavy, and very heavy grazing treatments, respectively, while potassium was 43 and 67% higher under heavy and very heavy grazing treatments, respectively. There were, however, no significant differences for exchangeable calcium and 51  magnesium. This pattern of higher exchangeable cations on grazed land was not observed in my study. Soil Reaction Soil pH in both H2O and CaCl2 was significantly higher (by 0.3 and 0.4 pH units) in spring-grazed pastures relative to fall-grazed pastures only in the 7.5-15-cm depth in 2008 (Table 2.4). Table 2.4: Grazing effects on soil reaction as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Year  Depth (cm)  Time of grazing Spring  Fall  Stocking Rate p  0 AUM ha-1  2 AUM ha-1  p  pH in H2O 1998  2008  1998  2008  0-7.5  7.2 (0.34)†  6.9 (0.33)  0.191  7.0 (0.32)  7.2 (0.39)  0.068  7.5-15  7.2 (0.30)  7.1 (0.27)  0.637  7.1 (0.23)  7.3 (0.30)  0.006  15-30  7.4 (0.31)  7.3 (0.40)  0.702  7.3 (0.31)  7.4 (0.39)  0.237  0-7.5  7.0 (0.42)  6.7 (0.28)  0.153  6.8 (0.42)  6.8 (0.35)  0.479  7.5-15  7.0 (0.42)  6.7 (0.28)  0.077  6.8 (0.40)  6.8 (039)  0.710  15-30  7.0 (0.50)  6.7 (0.36)  0.865 6.9 (0.54) pH in CaCl2  6.8 (0.38)  0.181  0-7.5  6.3 (0.26)  6.1 (0.30)  0.187  6.1 (0.25)  6.3 (0.32)  0.016  7.5-15  6.3 (0.33)  6.2 (0.24)  0.579  6.1 (0.30)  6.3 (0.24)  0.031  15-30  6.4 (0.26)  6.3 (0.36)  0.563  6.3 (0.26)  6.4 (0.36)  0.273  0-7.5  6.6 (0.43)  6.3 (0.30)  0.116  6.4 (0.45)  6.4 (0.35)  0.915  7.5-15  6.7 (0.44)  6.3 (0.30)  0.059  6.5 (0.45)  6.5 (0.39)  0.812  6.5 (0.55)  6.5 (0.39)  0.817  15-30 6.7 (0.50) 6.3 (0.36) 0.117 † The bracketed number represents the standard deviation of the mean (n=32).  After 20 years of grazing in 1998, pasture under the 2 AUM ha -1 treatment had significantly higher pH in H2O and CaCl2 in the top 15 cm relative to ungrazed pasture by a margin of 0.2 pH units (Table 2.4). Ten years after that in 2008, no significant differences between grazed and ungrazed pasture were detected. Soil pH in H2O was generally close to 7 while pH in CaCl2 was close to 6, which is typical for soils in this area (K. Broersma, personal communication, 2010). On a Dark Brown Chernozem in central Alberta (400 mm MAP), Mapfumo et al. (2000) found increased pH under three years of light (0.5 AUM ha-1), medium (0.67 AUM ha-1), and heavy (1.25 AUM ha-1) stocking rates. This study reported only changes in pH rather than measured values. They found that pH was increased by 0.11, 0.02, and 0.08 pH units under light, medium, and heavy grazing,  52  respectively. The authors considered the small increases in pH (<0.20 pH units) to be inconsequential to soil and plant processes. In my study, the differences in pH observed between treatments (ranging from 0.2 to 0.4 pH units) were somewhat larger than those reported by Mapfumo et al. (2000), but are still too low to cause any major changes in soil and/or plant processes. Another study carried out on a Black Chernozem in Alberta (Johnston et al., 1971) that focused on long-term effects of several grazing rates on soil properties found that after 17 years of continuous season-long (May 15-November 15) grazing, pH in H2O was significantly higher under very heavy (5 AUM ha -1) grazing (pH 6.2) relative to light (1.25 AUM ha-1) grazing (pH 5.7), but not significantly different under moderate (1.7 AUM ha 1  ) and heavy (2.5 AUM ha-1) grazing pressure relative to light. The resilience of the soil in this study  under the moderate and heavy stocking rates (which are similar to the moderate rate of 2 AUM ha -1 used in my study) could have been due to relatively high levels of soil organic matter (9.7-11.7%). In calcareous semi-arid soils, an increase in pH due to grazing, such as that observed in my study on grazed pastures in 1998, can be a result of grazing enhanced erosion. Increases in soil pH in surface horizons on grazed plots due to removal of topsoil and bringing of the alkaline subsoil to the surface has been reported in other grazing studies (Dormaar and Willms, 1998; Donkor et al., 2002). Available Phosphorus Available phosphorus levels were similar between spring-grazed and fall-grazed pasture after both 20 and 30 years of grazing (Table 2.5). Similarly, no significant differences between grazed and ungrazed areas were detected after either 20 or 30 years (Table 2.5). Table 2.5: Grazing effects on available phosphorus as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Spring  Time of grazing Fall  0-7.5 7.5-15 15-30  13.3 (6.71)† 7.4 (3.93) 5.8 (4.43)  14.1 (6.26) 8.9 (5.34) 7.0 (3.49)  0 AUM ha-1 mg kg soil-1 0.792 14.4 (6.13) 0.618 8.0 (4.41) 0.517 6.7 (4.08)  0-7.5 7.5-15 15-30  9.1 (4.39) 6.2 (3.37) 4.9 (3.95)  11.0 (3.17) 6.5 (2.82) 6.4 (3.05)  0.408 0.817 0.267  Year  Depth (cm)  1998  2008  p  10.0 (4.09) 6.4 (3.40) 5.3 (3.39)  Stocking Rate 2 AUM ha-1  p  13.0 (6.77) 8.4 (5.06) 6.0 (3.95)  0.471 0.967 0.556  10.1 (3.81) 6.3 (2.80) 5.9 (3.80)  0.733 0.933 0.659  † The bracketed number represents the standard deviation of the mean (n=32).  My results are similar to results from several studies in Alberta that did not detect significant differences in available phosphorus between grazing treatments. For example, on a Black Chernozem at 53  a fescue grassland site in Alberta, Johnston et al. (1971) found no significant differences after 17 years between grazed (at a range of stocking rates from 1.25-5 AUM ha -1) and ungrazed treatments. Smoliak et al. (1972) reported that 19 years of continuous (May 1-November 1) light (0.4-0.6 AUM ha -1) sheep grazing resulted in no differences for available phosphorus relative to ungrazed pasture on a Solonetz. My results are consistent with the general lack of trends for available phosphorus found by Milchunas and Lauenroth (1993) in their global review of grazing effects on vegetation and soil. Lower levels of available phosphorus in 2008 relative to 1998, especially in the 0-7.5-cm depth, may be cause for concern as phosphorus can be lost through wind and water erosion (Dormaar and Willms, 1998). Total Carbon and Nitrogen Neither the time of grazing nor the stocking rate treatments led to significant differences in total soil carbon and nitrogen either 20 or 30 years after the establishment of the grazing experiment (Table 2.6). Several other studies carried out in Alberta also reported that total soil carbon and total soil nitrogen were unaffected by grazing treatments, especially where soil organic matter levels have been relatively high. Dormaar et al. (1977), in a two-site study, reported that levels of total soil carbon or nitrogen did not significantly diverge between grazed and ungrazed areas on a Black Chernozem at a fescue grassland site (500 mm MAP) after 22 years of continuous season-long (May to November) cattle grazing (5 AUM ha-1), but total soil carbon and nitrogen were higher after 19 years of continuous season-long (May to November) sheep grazing (0.6 AUM ha -1) relative to an ungrazed control on a Solonetz at a mixed-grass prairie site (310 mm MAP). The authors attributed this difference of grazing response to the much higher level (about 10 times) of soil organic matter of the Black Chernozem compared to the Solonetz. At a fescue grassland site (550 mm MAP), Johnston et al. (1971) found that levels of soil organic matter were not significantly different among four cattle grazing rates (1.25, 1.7, 2.5, and 5 AUM ha-1) after 17 years of continuous season-long (May 15-November 15). This study was done on a Black Chernozem where soil organic matter ranged from 9.7 to 11.7% across treatments, which may explain the resilience of this soil to grazing pressure. Even though levels of soil organic matter in my study (3.4 to 3.8%) were lower than in the studies mentioned above, significant differences between grazing treatments were not detected.  54  Table 2.6: Grazing effects on total soil carbon, total soil nitrogen, and carbon-to-nitrogen ratio as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Time of grazing Stocking Rate Year Depth (cm) Spring Fall p 0 AUM ha-1 2 AUM ha-1 p -2 Total carbon (kg m ) 1998 0-7.5 1.90 (0.497)† 2.40 (0.676) 0.151 2.09 (0.760) 2.21 (0.496) 0.657 7.5-15 1.67 (0.438) 1.56 (0.469) 0.679 1.56 (0.488) 1.67 (0.416) 0.240 15-30 2.68 (0.773) 2.27 (0.795) 0.279 2.42 (0.693) 2.54 (0.911) 0.706 2008  0-7.5 7.5-15 15-30  1.86 (0.663) 1.46 (0.339) 2.66 (0.758)  2.14 (0.568) 1.57 (0.354) 2.44 (0.827)  1.92 (0.572) 1.44 (0.312) 2.39 (0.776)  0.314 0.238 0.389  0.20 (0.056) 0.14 (0.043) 0.20 (0.061)  0.473 2.07 (0.680) 0.578 1.59 (0.371) 0.466 2.71 (0.793) Total nitrogen (kg m-2) 0.289 0.18 (0.061) 0.687 0.14 (0.045) 0.298 0.22 (0.062)  1998  0-7.5 7.5-15 15-30  0.17 (0.042) 0.15 (0.037) 0.24 (0.047)  0.19 (0.040) 0.15 (0.034) 0.22 (0.052)  0.611 0.300 0.986  2008  0-7.5 7.5-15 15-30  0.17 (0.058) 0.14 (0.033) 0.24 (0.058)  0.19 (0.049) 0.15 (0.034) 0.24 (0.080)  0.481 0.515 0.867  0.19 (0.058) 0.15 (0.037) 0.25 (0.071)  0.17 (0.050) 0.14 (0.030) 0.23 (0.068)  0.259 0.259 0.651  1998  0-7.5 7.5-15 15-30  11.1 (0.52) 10.9 (1.30) 11.5 (3.36)  11.9 (0.55) 10.8 (0.44) 11.4 (3.14)  0.007 0.554 0.916  11.6 (0.66) 10.8 (0.52) 11.2 (2.32)  11.5 (0.69) 10.9 (1.27) 11.7 (3.95)  0.829 0.996 0.772  2008  0-7.5 7.5-15 15-30  11.2 (0.59) 10.7 (0.67) 11.5 (3.06)  11.2 (0.47) 10.4 (0.45) 10.1 (0.91)  0.990 0.378 0.081  11.2 (0.48) 10.5 (0.67) 11.3 (3.16)  11.3 (0.57) 10.5 (0.49) 10.3 (0.82)  0.430 0.686 0.206  C:N  † The bracketed number represents the standard deviation of the mean (n=32).  Significant grazing effects on soil organic matter have been reported in other studies at sites with precipitation regimes similar to that of my study. For example, on a Solonetz at a mixed-prairie site in Alberta (355 mm MAP), Naeth et al. (1991) found that 21 years of a late (August-October) light (0.9 AUM ha-1) grazing treatment increased soil organic matter by 58% in the Ah horizon (over the 0-8 cm depth) relative to an ungrazed exclosure while an early (May-July) light grazing treatment did not. On a Brown Chernozem at a semi-arid (384 mm MAP) site in Wyoming, Schuman et al. (1999) found that total soil carbon to 30 cm was 21 and 22% higher after 12 years of season-long cattle grazing at 0.67 and 2 AUM ha-1, respectively, relative to ungrazed pasture. At the same site, Ganjegunte et al. (2005) found that relative to soil carbon levels in an ungrazed control (2.09%), four years of season-long grazing at a rate of 0.67 AUM ha-1 resulted in significantly higher soil carbon (2.60%), while grazing at a rate of 2 AUM ha-1 resulted in significantly lower soil carbon (1.98%). The carbon-to-nitrogen ratio was significantly lower in spring-grazed relative to fall-grazed pastures in  55  the 0-7.5-cm depth after 20 years of grazing, while it was significantly higher in spring-grazed relative to fall-grazed pastures in the 15-30-cm depth after 30 years (Table 2.6). Grazed relative to ungrazed pastures exhibited no significant differences in carbon-to-nitrogen ratios after 20 or 30 years of grazing (Table 2.6). In a clipping study in South Africa, Snyman and du Preez (2005) observed a higher carbonto-nitrogen ratio in pastures of moderate condition relative to pastures in good and poor conditions. Increased vegetative growth as a response to moderate clipping disturbance (i.e., overcompensation response) (McNaughton, 1983) may have increased litter and root matter contributions to the soil, thereby raising the carbon-to-nitrogen ratio. In my experiment, a higher carbon-to-nitrogen ratio in the top 7.5 cm of fall-grazed pastures in 1998 may have resulted from the incorporation by trampling of relatively more abundant standing dead material and plant litter into the soil surface. Light Organic Matter Fraction After 20 years of grazing, levels of the light organic matter fraction (LOMF) were not significantly different under the time of grazing treatment. The stocking rate treatment resulted in 26% lower LOMF in the 0-7.5-cm depth relative to the ungrazed pasture, while at the two lower depths there were no significant differences between grazed and ungrazed treatments (Table 2.7). Table 2.7: Grazing effects on light organic matter fraction (LOMF) as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC. Depth (cm) 0-7.5 7.5-15 15-30  Spring  Time of grazing Fall  0.021 (0.0107)† 0.025 (0.0114) 0.018 (0.0226) 0.012 (0.0040) 0.009 (0.0048) 0.007 (0.0029)  p 0.362 0.952 0.285  0 AUM ha-1 kg kg-1 0.027 (0.0112) 0.018 (0.0226) 0.009 (0.0052)  Stocking Rate 2 AUM ha-1  p  0.020 (0.0102) 0.012 (0.0042) 0.007 (0.0022)  0.048 0.605 0.252  † The bracketed number represents the standard deviation of the mean (n=32).  In a study by Dormaar et al. (1989) carried out on a Black Chernozem in northern Alberta , despite the authors' hypothesis that any beneficial effects on organic matter from trampling of litter into the soil could be detected in increases of LOMF, they found instead that the proportion of carbon present as LOMF was 20% lower under a short-duration 2.6 AUM ha-1 grazing treatment relative to the ungrazed exclosures. In my study, the trend for lower LOMF under the 2 AUM ha -1 stocking rate may have resulted from decreased plant growth (e.g., Table E.1, Appendix E) and thus reduced contributions of litter fall and root decomposition.  56  Polysaccharides In 1998, after 20 years of grazing, no significant differences in soil polysaccharides between springgrazed and fall-grazed treatments were detected (Table 2.8). Ten more years of grazing pressure led to significantly lower polysaccharide levels (by 33%) in spring-grazed pastures relative to fall-grazed in the 7.5-15-cm depth (Table 2.8). Levels of polysaccharides were similar in grazed and ungrazed pastures after 20 years of grazing at all three sampling depths, while 30 years of grazing at 2 AUM ha -1 led to significantly lower (by 20%) soil polysaccharides in the 0-7.5-cm depth relative to an ungrazed control (Table 2.8). Table 2.8: Grazing effects on soil polysaccharides as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Year  Depth (cm)  1998  Time of grazing  Stocking Rate  Spring  Fall  0-7.5  0.46 (0.118)†  0.54 (0.109)  0.137  0.50 (0.118)  0.50 (0.124)  0.978  7.5-15  0.39 (0.094)  0.39 (0.074)  0.946  0.39 (0.093)  0.39 (0.076)  0.898  15-30  0.66 (0.147)  0.75 (0.355)  0.772  0.69 (0.222)  0.73 (0.319)  0.774  0-7.5  0.59 (0.377)  0.83 (0.259)  0.117  0.79 (0.323)  0.63 (0.348)  0.051  7.5-15  0.47 (0.253)  0.70 (0.275)  0.043  0.62 (0.289)  0.54 (0.284)  0.348  15-30 1.09 (0.509) 1.07 (0.474) 0.903 1.19 (0.348) † The bracketed number represents the standard deviation of the mean (n=32).  0.97 (0.583)  0.148  2008  p kg m-2  0 AUM ha  -1  2 AUM ha-1  p  Soil polysaccharides are comprised of a wide range of monosaccharides with simple and complex molecular structure and are formed in the soil either as a result of plant root activity or microbial activity during the decomposition processes (Tisdall and Oades, 1982; Haynes and Swift, 1990). Lower polysaccharides levels under the 2 AUM ha-1 treatment relative to the ungrazed control as observed in 2008 in my study were in agreement with those made by Dormaar and Willms (1998) who carried out a 44-year grazing study on a Black Chernozem in Alberta. In their study, continuous (May 15-November 15) grazing at rates of 1.2, 2.4, 4.8 AUM ha-1 resulted in significantly lower (by 7, 27, and 32%, respectively) levels of monosaccharides relative to exclosures. Reduced plant growth on grazed relative to ungrazed pastures could be one of the reasons for the observed differences in soil polysaccharides. Some other studies have suggested that addition of relatively large quantities of organic material as manure in grazed pastures may have fostered the soil microbial populations which produce polysaccharides. Smoliak et al. (1972) found that levels of polysaccharides were greater inside  57  sheep holding pens and also in pastures where grazing intensity and manure deposition were higher, relative to areas protected from grazing. Due to the moderate stocking rate used in my study, manure deposition is not suspected as a factor in polysaccharide levels. The trend for decreased polysaccharides in spring-grazed relative to fall-grazed pastures and in grazed areas relative to exclosures may result from decreased plant growth in these areas (e.g., Table E.1, Appendix E) due to grazing. Root biomass Root biomass did not significantly differ between the spring and fall grazing treatments after either 20 or 30 years (Figure 2.1).  Figure 2.1: Root biomass under spring and fall grazing treatments at 0-7.5 cm, 7.5-15 cm, and 15-30 cm as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32).  Twenty and 30 years of grazing pressure also did not result in any significant differences between grazed and ungrazed pasture (Figure 2.2).  Figure 2.2: Root biomass under 0 and 2 AUM ha-1 treatments at 0-7.5 cm, 7.5-15 cm, and 15-30 cm as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32).  58  The lack of differences for root biomass data is supported by the findings for total soil carbon and nitrogen (Table 2.6), where no significant differences between grazing treatments were also found. This is not surprising since fibrous grass roots are the main source of soil carbon in grasslands. These findings suggest that any effort to bolster soil carbon storage through elimination of moderate grazing on semi-arid grasslands of the southern interior of BC may not result in changes in soil organic matter. My results are similar to those reported by Schuman et al. (1999) for a Dark Brown Chernozem at a semi-arid site (384 mm MAP) in Wyoming. They found no significant effect on root biomass under 12 years of season-long heavy (2 AUM ha-1) cattle grazing relative to an ungrazed control. There are also studies that have reported changes in root biomass due to grazing, especially at heavy stocking rates, and those changes were generally accompanied by changes in plant species composition from deep-rooted perennial species to shallow-rooted weedy species (Naeth et al., 1990; Greenwood and Hutchinson, 1998; Snyman, 2005). For example, heavy (5 AUM ha-1) grazing resulted in an increase in root mass in the top 15 cm from 17.2 dry matter tons ha -1 on ungrazed pasture to 25.3 dry matter tons ha-1 on heavily grazed pasture on a Black Chernozem (500 mm MAP) in Alberta (Dormaar et al., 1977). This increase correlated with a change in dominant vegetation from rough fescue to forbs and shrubs. In my study, 40% of root biomass was found in the top 15 cm under both the 0 and 2 AUM ha-1 grazing treatments in both sampling years. This indicates that grazing did not lead to a substantial shift in plant species composition from the deep-rooted bunch grasses typical for these grasslands to shallow-rooted species. This is corroborated by limited species cover sampling conducted on the study site in May, 2010 which found that the only species to change with grazing relative to the ungrazed control was Sandberg's bluegrass, which accounted for only 5% of cover on grazed sites (Figure D.8, Appendix D). However, these results are not supported by findings for above-ground biomass determined in 1998 (Table E.1, Appendix E) and in another unpublished study (D. Thompson, personal communication, 2010). This warrants further investigation.  2.3.2 Grazing Effects on Soil Physical Properties Bulk Density Soil bulk density was generally not affected by grazing during this study, with two exceptions. In 2008 at the 0-7.5 cm depth a significant interaction between time and rate of grazing was observed (Table 2.9). The significant interaction observed in 2008 showed that soil bulk density on the spring-grazed  59  areas was 20% greater than on spring exclosures, while fall-grazed areas had bulk density that was 7% greater (though not significant) than fall exclosures. Table 2.9: Time of grazing and stocking rate treatment interaction effects on soil bulk density as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Year  Depth (cm)  Spring Ungrazed  Fall Grazed  Ungrazed  S-F(UG)  Grazed  S-F(G)  Mg m-3 2008  0-7.5  0.86 (0.146)  1.03 (0.110)  (S)UG-G  (F)UG-G  0.0048  0.2136  p  0.87 (0.104)  0.93 (0.103)  0.7502  0.0386  † The bracketed number represents the standard deviation of the mean (n=32).  In 1998 in the 15-30-cm depth, spring grazing resulted in 11% greater bulk density relative to the fall grazed treatment (Table 2.10) Table 2.10: Grazing effects on soil bulk density as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Year  Depth (cm)  Time of grazing Spring  Fall  Stocking Rate p  -1  0 AUM ha  2 AUM ha-1  p  Mg m  -3  1998  0-7.5  1.09 (0.107)†  1.03 (0.080)  0.202  1.03 (0.099)  1.08 (0.092)  0.180  7.5-15  1.06 (0.080)  1.02 (0.121)  0.452  1.03 (0.119)  1.06 (0.085)  0.404  15-30  1.09 (0.085)  0.98 (0.148)  0.073  1.02 (0.104)  1.05 (0.154)  0.244  7.5-15  0.91 (0.093)  0.90 (0.107)  0.843  0.89 (0.097)  0.92 (0.101)  0.244  15-30 0.92 (0.072) 0.87 (0.117) 0.257 0.89 (0.105) † The bracketed number represents the standard deviation of the mean (n=32).  0.90 (0.098)  0.765  2008  In arid grassland areas, early grazing generally has a more severe effect on soil bulk density than grazing later in the season because soils are generally wetter in the spring from snowmelt and immature vegetation is less resilient to withstand grazing (Laycock and Conrad, 1967; Naeth et al., 1991). Time of grazing at light and heavy stocking rates was also the focus of a study carried on a loam Black Chernozem in northeast Alberta (380 mm MAP) by Naeth et al. (1990). They found that bulk density in the top 7.5 cm was 8% greater after 12 years of heavy (4.4 AUM ha -1) spring (June 1-30) grazing relative to heavy fall (September 15-October 15) grazing. However, bulk density on light (1.5 AUM ha 1  ) spring and fall grazing treatments was not significantly different at 0-7.5 cm depth, suggesting that at  lower stocking rate those soils were not susceptible to compaction due to spring grazing. In my study, after 30 years of grazing, soil bulk density was 20% greater in spring-grazed areas relative to spring exclosures, while bulk density on fall-grazed areas was 7% greater than on fall exclosures. The pattern of the interaction means supports the hypothesis that spring-grazed pastures are more susceptible to 60  compaction due to grazing than fall-grazed pastures. Effects of grazing rates on soil compaction have received more attention in rangeland studies than effects of time of grazing. For example, Dormaar et al. (1989) found that heavy grazing (2.6 AUM ha -1, which was triple the recommended grazing rate) increased bulk density of a loam Black Chernozem by up to 15% relative to the ungrazed control. In their long-term (44 years) study on a Black Chernozem, Dormaar and Willms (1998) found that bulk density was 53% and 57% higher under heavy (2.4 AUM ha-1) and very heavy (4.8 AUM ha-1) stocking rates, respectively, compared to light (1.2 AUM ha -1) grazing and the ungrazed exclosure. Naeth et al. (1990) found significant increases in bulk density due to long-term grazing in the surface horizon (the top 7.5 cm) at two different sites. Seventeen years of month-long cattle grazing in the spring (June 1-30) and fall (September 15-October 15) at a rate of 1.5 AUM ha-1 resulted in increases of 7 and 8%, respectively, with respect to ungrazed exclosures, while 41 years of season-long (May-September) cattle grazing at a rate of 1.6 and 2.4 AUM ha -1 resulted in increases of 7 and 11%, respectively, with respect to ungrazed exclosures. The fact that after 30 years of grazing the soil bulk density was significantly greater at the 0-7.5 cm depth on 2 AUM ha -1 treatment relative to the ungrazed exclosures in spring-grazed but not fall-grazed pastures indicates that these rangeland soils are starting to show signs of compaction. This warrants caution when using this stocking rate, especially at the spring grazing. Mechanical Resistance After 20 years of grazing, soil mechanical resistance was significantly higher (11-32%) at depths of 1.5-4.5, 10.5, and 13.5 cm under spring grazing relative to fall grazing (Figure 2.3). However, ten years later there were no significant differences in soil mechanical resistance between spring and fall grazing treatments (Figure 2.3). The 1998 data were consistent with the hypothesis that spring grazing on these semi-arid grasslands can lead to higher compaction relative to fall grazing. My results are similar to those reported by Naeth et al. (1990) on a Solonetz at a mixed-prairie site in Alberta (355 mm MAP), where early (May-July) heavy (0.9 AUM ha-1) grazing resulted in greater soil mechanical resistance at 0 and 2.5 cm (154% and 39%, respectively) relative to an ungrazed control. Early heavy grazing also resulted in greater soil mechanical resistance at 0 and 2.5 cm (13 and 11%, respectively) relative to late (August-October) heavy grazing.  61  Figure 2.3: Soil mechanical resistance under spring and fall grazing treatments as determined 20 years (i.e., 1998) (a) and 30 years (i.e., 2008) (b) after establishment of an experiment on rangeland north of Kamloops, BC. The dotted line represents the critical upper limit for root growth. Error bars represent the standard deviation of the mean (n=32). Significant differences (p<0.10) were marked by an asterisk.  Soil mechanical resistance was significantly affected by the stocking rate after both 20 and 30 years of grazing. In 1998, mechanical resistance in grazed pastures was higher relative to the ungrazed exclosures at depths of 1.5-7.5, 10.5, 13.5, and 16.5 cm, and in 2008 significant differences were observed over the 3-12 cm depth (Figure 2.4). Soil mechanical resistance measurements in the top 7.5 cm were 21-83% higher after 20 years of grazing (1998) and 28-70% higher after 30 years of grazing (2008) relative to ungrazed exclosures. This is similar to increases under stocking rate treatments reported in other rangeland studies. Naeth et al. (1990) found significantly higher (46-80%) mechanical resistance up to 15 cm after 41 years of moderate (1.6 AUM ha-1) cattle grazing relative to the ungrazed control on a Black Chernozem at a foothills fescue site in Alberta (550 mm MAP). Those grazed plots were also characterized by lower soil organic matter relative to ungrazed exclosures, as determined in another study by Naeth et al. (1991). Lower soil organic matter content generally increases a soil’s susceptibility to compaction (Soane, 1990; Aragón et al., 2000; Zhao et al., 2008). Donkor et al. (2002) reported a 4-63% increase in soil mechanical resistance in the top 15 cm after two years of continuous (May-September) moderate (2.08 AUM ha-1) wapiti grazing. Their study was done on a Dark Gray Luvisol with 2.1% soil organic matter and a clay content of 25%; the state of both of these properties would have rendered this soil susceptible to compaction. Despite more favourable organic matter and clay contents in the soils in my 62  Figure 2.4: Soil mechanical resistance under ungrazed and grazed treatments as determined 20 years (i.e., 1998) (a) and 30 years (i.e., 2008) (b) after establishment of an experiment on rangeland north of Kamloops, BC. The dotted line represents the critical upper limit for root growth. Error bars represent the standard deviation of the mean (n=32). Significant differences (p<0.10) were marked by an asterisk.  study (5.4% organic matter and 10% clay) as well as a shorter grazing period, increases of 28-70% were observed due to the difference in the length of the grazing trial (i.e., 2 years versus 30 years). The greater soil mechanical resistance observed under the 2 AUM ha -1 relative to the 0 AUM ha-1 stocking rate treatment were in agreement with greater soil bulk density in the 0-7.5 cm depth after 30 years of grazing, especially in spring-grazed areas (Table 2.9). This further supports the hypothesis that the 2 AUM ha-1 stocking rate is causing soil compaction relative to the ungrazed control. Mechanical resistance has been found in several studies carried out in either grassland or forest ecosystems to be a more sensitive indicator of soil compaction than bulk density (Naeth and Chanasyk, 1995a; Rodd et al., 1999; Voorhees et al., 1978). It is relatively easier to take a larger number of mechanical resistance measurements than bulk density samples for the same amount of time and effort, which allows for better spatial coverage of an area and a much more representative sample (Courtin et al., 1983). Under all treatments in my study, the commonly cited critical limit for restricted root growth of 2500 kPa (Taylor et al., 1966; Greacen et al., 1969; Busscher, 1990) was reached within ~8 cm. Reflecting this, a significant negative correlation between mechanical resistance and root biomass was also found (Figure D.7, Appendix D). Restricted root growth below about 8 cm on grazed pastures is further supported by the occurrence of 56% of root biomass in the top 7.5 cm of grazed areas (observed after 30 years of grazing) relative to 44% in ungrazed areas (Figure 2.1 and Figure 2.2). 63  In 1998, soil mechanical resistance below 10 cm was around the root-restricting limit of 2500 kPa, while ten years later it was higher, ranging from 3000 to 3700 kPa (Figure 2.3 and Figure 2.4). This trend observed for soil mechanical resistance was not supported by soil bulk density data, which were generally lower in 2008 than in 1998 (Table 2.9 and Table 2.10). Additionally, the mechanical resistance of ungrazed areas has increased along with that of grazed areas (Figure 2.3 and Figure 2.4). This warrants further investigation and more frequent measurements both during the growing season and between growing seasons. Aggregate Stability Twenty years of spring and fall grazing did not result in any significant differences in MWD (Figure 2.5a). After 30 years of grazing, in 2008, MWD was found to be 50% higher on spring-grazed relative to fall-grazed pastures (Figure 2.5a). Significant differences between grazed and ungrazed pastures were not evident after either 20 or 30 years of grazing (Figure 2.5b). In general, MWD values appeared lower in 2008 than in 1998.  Figure 2.5: Aggregate mean weight diameter (MWD) under time of grazing (a) and stocking rate (b) treatments as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32). Significant differences (p<0.10) were marked by an asterisk.  Twenty years of moderate grazing pressure did not lead to any differences for any individual aggregate size fraction between spring- and fall-grazed treatments (Figure 2.6). Ten years after this initial determination, three of the individual aggregate size fractions (viz. the 2-6, 1-2, and <0.25-mm size fractions) were significantly different between spring- and fall-grazed sites (Figure 2.6). In 2008, the 64  proportions of the two largest size fractions (2-6 and 1-2 mm) were greater under the spring grazing treatment relative to the fall. The proportion of the total aggregate sample present as the smallest aggregate size (<0.25 mm), on the other hand, was lower in spring-grazed pastures relative to fallgrazed. These data seem to suggest that in these semi-arid ecosystems, grazing carried out in the spring actually favours creation of larger aggregate size fractions.  (c)  (d)  Figure 2.6: 2-6 mm, 1-2 mm, 0.25-1 mm and <0.25 mm aggregate size fractions under spring and fall grazing treatments as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32). Differences were considered significant at p<0.10 and marked by an asterisk.  MWD has been found to be inversely related to soil water content in a number of studies (Haynes and Swift, 1990; Hermawan and Bomke, 1996; Churchman and Tate, 1987), including some carried out on semi-arid grassland rangelands in BC (Krzic et al., 2000; Wallace et al., 2009). At low water content, aggregate strength and thus MWD are increased by enhanced soil cohesion and strength due to capillary forces and increased contact points among primary particles. The water content of individual aggregate samples was determined as part of the wet-sieving process, and water content of aggregates was significantly higher on fall-grazed pastures relative to spring-grazed pastures after 30 years of grazing but not after 20 years (Figure D.4, Appendix D).  65  Unlike bulk density and soil mechanical resistance data, individual aggregate size fractions showed no significant differences under the stocking rate treatment after 20 or 30 years (Figure 2.7). Since the proportions of these size fractions were used to calculate MWD, these data are in agreement with the lack of significant differences in MWD between grazed and ungrazed areas (Figure 2.7).  Figure 2.7: 2-6 mm, 1-2 mm, 0.25-1 mm and <0.25 mm aggregate size fractions under ungrazed and grazed treatments as determined 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32).  Other studies have shown that grazing led to lower aggregate stability relative to ungrazed pastures. On a silty clay loam in Kansas, Knoll and Hopkins (1959) reported that after four years, aggregate samples from ungrazed, moderately grazed (2.5 AUM ha -1), and heavily grazed (9.9 AUM ha -1) pastures contained 89, 64, and 56% water-stable aggregates, respectively. The authors defined as water-stable aggregates with a diameter of 0.5, 1, and 2 mm. Although the length of the grazing period in this study was not specified, both stocking rates were higher than the moderate rate of 2 AUM ha -1 used in my study, potentially leading to observed differences that were lacking in my study. Dormaar and Willms (1998) found that 44 years of season-long (May 15-November 15) cattle grazing at a moderate rate (2.4 66  AUM ha-1) reduced MWD to 1.43 mm from 1.69 mm (a 15% decrease) on a Black Chernozem in southwest Alberta. Total soil carbon also decreased by 34% under this grazing treatment, which may have led to decreased aggregate resilience to deformation due to trampling. In my study there were no significant differences in soil carbon between the 0 and 2 AUM ha -1 grazing rates (Table 2.6), which was in agreement with lack of difference in MWD (Figure 2.5b). Furthermore, both of these studies occurred at wetter (~550 mm MAP) sites without annual drought conditions, where wetting-drying cycles may not have played as important a role in influencing aggregate stability. It has been generally inferred that soil polysaccharides enhance formation and stabilization of soil aggregates (Dexter, 1988; Gregorich et al., 1994). Even though trends for polysaccharides and aggregate stability parameters observed in my study do not quite match, the correlation coefficients obtained between individual aggregate size fractions and polysaccharides (pooled over the two years of sampling) were all statistically significant (Figure 2.8). The only exception was 0.25-1 mm size fraction. Larger size fractions (2-6 mm and 1-2 mm) were negatively correlated and the smallest (<0.25 mm) fraction was positively correlated to polysaccharides. Some studies have found good correlation between carbohydrate content and soil aggregate stability (Haynes and Swift, 1990; Angers and Mehuys, 1993), while others have not (Carter and Kunelius, 1993). The most likely reason for lack of agreement among the studies is that other types of soil organic matter compounds such as the hydrophobic aliphatic fraction (Capriel et al., 1990) and fungal hyphae (Tisdall and Oades, 1982) are involved in aggregate stability, especially of macroaggregates. Another reason for opposing findings reported in the literature is related to the lack of standard extraction and hydrolysis procedures used to determine soil carbohydrate and polysaccharides (Gregorich et al., 1994). In a study by Lamagna (2008) carried out on rough fescue grasslands in the southern interior of BC, 12-mm aggregates did not correlate with polysaccharides while all other size fractions did. The lack of correlation between 1-2-mm aggregates and polysaccharides as well as the low proportion of this size class (15%) of the total sample mass were related to the transient nature of this size fraction, which is most likely formed as a result of destruction of larger aggregates rather than by aggregation of soil particles by binding agents. In my study, I have similar evidence to support this hypothesis for the 0.251-mm size class instead. Similar to findings of the study mentioned above, the 0.25-1-mm aggregate size class also comprised ~15% of total stable aggregates.  67  Figure 2.8: Relationship between polysaccharides and 2-6 (a), 1-2 (b), 0.25-1 (c) and <0.25 (d) mm aggregate size fractions as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC.  Mean weight diameter and individual aggregate size classes other than the 0.25-1-mm class correlated well with soil water content and bulk density, but not with mechanical resistance (Table A.1, Appendix A). Additionally, total carbon, total nitrogen, and the carbon-to-nitrogen ratio were not good predictors of any aggregate size class according to their correlation coefficients (Table A.1, Appendix A). Correlations with root biomass were significant and positive for the MWD and 2-6-mm size fraction, and significant and negative for the <0.25-mm size fraction. This is in accordance with the hypothesis that roots act as temporary binding agents and aid in stabilizing larger aggregates (Tisdall and Oades, 1982). Unlike soil bulk density and mechanical resistance data, aggregate stability parameters showed no significant differences between the two stocking rates either after 20 or 30 years (Figure 2.5b and Figure 2.7). Although aggregate stability has been shown to be a useful indicator of management impacts, since it integrates physical, chemical, and biological properties, this may not be the case on these loam soils grazed at a stocking rate of 2 AUM ha-1.  68  2.4 Summary of Grazing Effects on Soil Physical, Chemical, and Biological Properties Cation exchange capacity, exchangeable Mg and Na, polysaccharides and bulk density were all negatively affected by the spring grazing treatment. While grazing effects were detected for components of soil organic matter such as LOMF (which was lower under the 2 AUM ha -1 grazing treatment) and polysaccharides (which were lower under both the spring grazing and 2 AUM ha -1 grazing treatments), grazing effects were not detected for the total soil organic matter pool (represented by total soil carbon and nitrogen). This was in agreement with observations made by Carter (2002b) that it is generally difficult to measure small changes in soil organic matter against a relatively large background mass. Hence, attributes of soil organic matter (e.g., LOFM, polysaccharides, microbial biomass) are more sensitive to change in organic matter inputs than the total mass and can better indicate the direction of change of the total organic matter mass. Accordingly, significant effects on root biomass, a major contributor to soil organic matter in grassland ecosystems, were not detected either. In these semi-arid bunchgrass rangelands, long-term grazing carried out either in the spring or fall at a moderate rate of 2 AUM ha-1 did not affect root biomass and, in turn, soil organic matter and carbon storage. Consequently, reducing grazing pressure on these semi-arid grasslands may not increase the amount of carbon sequestered below ground. For those soil properties which were measured after both 20 and 30 years of grazing, various trends between 1998 and 2008 were observed. For example, soil mechanical resistance, bulk density, and the carbon-to-nitrogen ratio were significantly higher under the spring grazing treatment, and soil reaction was significantly higher under the 2 AUM ha-1 stocking rate treatment after 20 years of grazing but not after 30 years. The lack of significant differences ten years after the initial sampling in 1998 suggests that the effects of the time of grazing and 2 AUM ha-1 stocking rate treatments are becoming less evident for these soil properties, contradicting the hypotheses that grazing at the recommended moderate stocking rate of 2 AUM ha-1 and spring grazing are detrimental to soil quality in the semi-arid grasslands of the southern interior of BC. On the other hand, MWD and soil pH were significantly higher under the spring grazing treatment, soil bulk density was significantly higher under the 2 AUM ha -1 stocking rate treatment, and polysaccharides were significantly lower under both the spring grazing and the 2 AUM ha -1 stocking rate treatments after 30 years of grazing but not after 20 years. Although there appears to be a general 69  increase in mechanical resistance between 1998 and 2008, grazing effects may not be the primary cause in this case, as mentioned previously. This lends support to the hypotheses that grazing at the recommended moderate stocking rate of 2 AUM ha -1 and spring grazing are detrimental to soil quality in these semi-arid grasslands. Some soil properties of the semi-arid grasslands of the southern interior of BC showed grazing effects as hypothesized and were similar to those documented in other grassland studies. Specifically, lower CEC and polysaccharides as well as higher soil pH, bulk density, and mechanical resistance under both spring grazing and 2 AUM ha-1 grazing treatments conformed to my hypotheses that soil properties would be more strongly affected under both spring grazing relative to fall grazing and under a 2 AUM ha-1 grazing rate relative to 0 AUM ha-1. Except for the CEC results, the trend for these effects adhered to patterns reported in grazing literature on Black Chernozems in Alberta (Naeth et al., 1990; Dormaar and Willms, 1998; Mapfumo et al., 2000). Some soil properties did not exhibit any significant differences, similar to other studies. Specifically, available P and root biomass exhibited no significant differences, as found in studies on Dark Brown Chernozems in Wyoming (Schuman et al., 1999) and Back Chernozems in Alberta (Johnston et al., 1971). Finally, some soil properties (unchanged total soil carbon and nitrogen and higher aggregate MWD under the spring grazing treatment) conformed neither to my hypotheses nor to research elsewhere, underlining the fact that the semi-arid grasslands of the southern interior of BC are a unique ecosystem with characteristic responses to grazing pressure. A lack of significant grazing treatment effects on total soil carbon and nitrogen such as those reported in my study has been associated with soils with relatively high levels of soil organic matter, such as Black Chernozems (Johnston et al., 1971; Dormaar et al., 1977) rather than with soils with relatively less organic matter such as the Brown and Dark Brown Chernozems in my study. Aggregate MWD has been generally shown to decrease under grazing pressure (Dormaar and Willms, 1998), but in my study the lack of difference between grazed and ungrazed areas indicates that a moderate stocking rate of 2 AUM ha-1 sustained over a 30-year period did not reduce the stability of aggregates. Furthermore, the higher MWD and proportion of 2-6 and 1-2 mm size fractions on spring-grazed relative to fall-grazed pastures, along with findings of unchanged root biomass and lower polysaccharides under the spring relative to the fall grazing treatment, suggest that aggregation on these grasslands was most likely enhanced by fungal hyphae and/or enzymes (e.g., glomalin) produced by some species of arbuscular mycorrhizae fungi. Reduced vegetation cover and productivity on spring-grazed pastures (Table E.1, Appendix E) could have resulted in lower 70  interception by vegetation, higher infiltration of rainfall and snowmelt water, and thus a tendency for higher soil water content over the 30-year period between 1978 and 2008. This could have resulted in greater fungal population in spring-grazed relative to fall-grazed pastures and in turn enhanced macroaggregate formation. This warrants further investigation.  2.5 Conclusions Greater soil bulk density, mechanical resistance, pH, as well as lower polysaccharides and CEC were determined under the spring grazing treatment and showed that long-term grazing early in the growing season had negative impacts on soil productivity relative to fall grazing in these semi-arid grasslands of the southern interior of BC. The only soil property that exhibited a positive effect as a result of spring grazing was aggregate stability (i.e., MWD, 2-6, and 1-2 mm size fractions). It is possible that aggregate stability of these soils, which have a relatively low clay content (7-13%), was enhanced by greater activity of soil organisms (particularly fungi) in spring-grazed pastures relative to fall-grazed pastures. The moderate grazing rate of 2 AUM ha -1, recommended by the provincial rangeland planning unit, led to greater soil mechanical resistance and pH, as well as lower soil polysaccharides and LOMF relative to the ungrazed control. After 30 years of grazing, soil bulk density was greater in the 0-7.5-cm depth under the 2 AUM ha-1 treatment relative to the ungrazed exclosure in spring-grazed but not in fallgrazed areas, indicating that this stocking rate, when used for spring grazing, has led to soil compaction. The long-term impacts of spring grazing and the 2 AUM ha -1 stocking rate observed in my study showed that rangeland managers in the southern interior of BC should consider adjustments of the current recommendations regarding the time of grazing and stocking rate, which were both based solely on vegetation responses, and should consider including soil properties in rangeland health assessments.  71  Chapter 3- General Conclusions and Recommendations for Future Research Range health assessment in BC has traditionally focused on plants (Fraser, 2007; Delesalle et al., 2009). Stocking rates are recommended based on an expected level of forage utilization by livestock. Although the determination of plant productivity and quality is crucial to effective ranch operation, the recognition of the critical role soil plays makes a strong case for the use of soil properties as indicators of ecosystem sustainability (Carter, 1996; Carter et al., 1997). Due to the low productivity of the semiarid grasslands of the Lac du Bois Range in Kamloops, BC, the lower-elevation pastures in the Lower and Middle Grasslands are used for early- (spring) and late-season (fall) grazing, both before and after summer grazing of the more productive higher-elevation Upper Grasslands. The current stocking rate of 2 AUM ha-1 is considered moderate and is based on a desired forage utilization of 50%. In this study, grazing effects (time of grazing and stocking rate treatments) on soil quality were assessed by evaluating the levels of selected soil chemical, biological, and physical properties. My study represents the first attempt to determined the long-term effects of time of grazing and stocking rate on selected soil chemical, biological, and physical properties on these unique semi-arid grasslands of the southern interior of BC. The findings of my study will help ranchers and government range ecologists evaluate the sustainability of current (on-going) ranching practices that were originally devised based only on vegetation.  3.1 General Conclusions Increased susceptibility of soil in spring-grazed pastures to degradation due to grazing pressure was shown by significant treatment interactions for soil bulk density and CEC, where the pattern of treatment interaction means shows greater differences between grazed and ungrazed areas in spring pastures than grazed and ungrazed areas in fall pastures. Soil bulk density on the spring-grazed areas was 20% greater than on spring exclosures, while bulk density in fall-grazed areas was only 7% greater (though not significant) than fall exclosures. Soil under the spring grazing treatment had 27 and 35% lower CEC than under the fall-grazing treatment in the 7.5-15 and 15-30-cm depths, respectively. Additionally, spring exclosures had 10 and 11% higher CEC than spring-grazed pastures, while fall grazed pastures were 3 and 8% higher (though not significant) than fall exclosures in the 7.5-15 and 1530-cm depths, respectively.  72  Significant effects of the time of grazing treatment were observed for soil bulk density, mechanical resistance, MWD, the carbon-to-nitrogen ratio, polysaccharides, and soil reaction. Specifically, under the spring grazing treatment, soil bulk density was 11% higher in the 15-30-cm depth in 1998, mechanical resistance was 11-23% higher at depths of 1.5-4.5, 10.5, and 13.5 cm in 1998, MWD was 50% higher in 2008, the carbon-to-nitrogen ratio decreased from 11.9 to 11.1 in the 0-7.5-cm depth in 1998, polysaccharide levels were 33% lower in the 7.5-15-cm depth in 2008, and soil pH in H 2O and CaCl2 was 0.3-0.4 pH units higher in 2008. These changes can be considered detrimental effects in the cases of bulk density, mechanical resistance, polysaccharides, and soil reaction. Increased bulk density and mechanical resistance can lead to reduced porosity, gas diffusion, water infiltration, and root growth (Greacen and Sands, 1980) while a decrease in polysaccharide levels can affect soil microbial populations and aggregate formation (Tisdall and Oades, 1982; Dexter, 1988). Higher soil reaction can be detrimental because many soil micronutrients are more available at relatively lower pH. For MWD and the carbon-to-nitrogen ratio, these changes can be considered beneficial effects. The formation of larger aggregates promotes higher porosity and greater gas diffusion and water infiltration in the soil, while organic matter with a relatively lower carbon-to-nitrogen ratio (i.e., more labile) promotes aggregate formation (Kay et al., 1998). Twenty and 30 years of spring grazing had a relatively more detrimental effect on these soil properties than fall grazing. The 2 AUM ha-1 grazing rate significantly affected mechanical resistance, polysaccharides, the LOMF, and soil reaction. Specifically, under the 2 AUM ha -1 treatment, mechanical resistance in the top 7.5 cm was 21-83% higher in 1998 and 28-70% higher in 2008, polysaccharides were 20% lower in the 0-7.5cm depth in 2008, the light organic matter fraction was 26% lower in the 0-7.5-cm depth in 1998, and soil pH in both H2O and CaCl2 was 0.2 pH units higher in the 0-7.5 and 7.5-15-cm depths in 1998. These changes can be considered detrimental effects. As previously mentioned, higher bulk density and mechanical resistance can lead to reduced porosity, gas diffusion, water infiltration, and root growth. Decreased polysaccharides can affect soil microbial populations and aggregate formation, while higher soil reaction can result in lower micronutrient availability. The LOMF represents a transient organic matter pool between fresh plant tissue (e.g., litterfall and material sloughed from roots) and more stable soil organic matter. Decreased levels of LOMF can indicate fewer contributions of organic residues to the soil, which may also decrease soil organic matter levels. Twenty and 30 years of grazing at a moderate rate of 2 AUM ha-1 had a relatively more detrimental effect on these soil properties than a rate of 0 AUM ha-1. 73  Grazing effects on vegetation are more obvious and have long been recognized: for example, range health assessments in BC focus mostly on plants. The results of this study make a case for including soil properties which show significant grazing effects in assessments of range health on semi-arid rangelands.  3.2 Recommendations for Future Research The results for MWD under the spring grazing treatment, which contradict the hypothesis under the research objectives, are surprising and warrant further investigation into the variables influencing aggregation in these semi-arid grasslands. In particular, it would be interesting to characterize influences derived from the presence of fungal populations during the spring such as fungal hyphae and/or extracellular glycoprotein produced by arbuscular mycorrhizae fungi (e.g., glomalin). The lack of significant relationships between the 0.25-1-mm aggregate size fraction and other soil properties such as polysaccharides, total carbon, total nitrogen, bulk density, mechanical resistance, root biomass, or soil water (Table A.1, Appendix A) is surprising given the theoretical relationships with these soil properties. Further investigation into the nature of this aggregate size class could include the determination of properties shown in other studies to influence aggregation such as microbial biomass, fungal hyphae, enzymes (i.e., glomalin), or soil water content. These analyses could be performed on individual aggregate size fractions in order to characterize individual relationships. At a stocking rate of 2 AUM ha -1, which is considered moderate, there is some question with regard to the distribution of grazing effects on pastures of large size (in the case of my study, individual pastures were ~65 ha). Currently, some work is being done with tagging of cattle with GPS-enabled collars on the Lac du Bois Range (D. Thompson, personal communication, 2010), which will reveal cattle preferences for certain areas (due to habit or occurrence of favourite forage) or dislike of other areas (due to slope or exposure). Comparing sampling sites with estimates of actual cattle traffic would illuminate the extent of the direct effects of grazing (i.e., trampling) on soil and vegetation quality. Some interaction means (e.g., CEC) point to the possibility that spring grazing is having a detrimental effect on soil properties while fall grazing is having a beneficial effect or no effect. This may be due to decreased plant growth on spring-grazed pastures as well as animal-related actions such as the trampling of vegetation into the soil surface on fall-grazed pastures. A three-way comparison of springgrazed areas, fall-grazed areas, and exclosures would need to be done to determine whether there is a differential effect. Because of the negative effects of the stocking rate treatment demonstrated in this 74  study, it would be also beneficial to investigate the effects of stocking rates lower, and possibly higher, than the current moderate rate of 2 AUM ha -1 in order to establish a stocking rate which is more suitable for the grasslands in this region. Although above-ground biomass sampling in 1998 as well as another unpublished study (D. Thompson, personal communication, 2010) have reported decreased productivity in spring-grazed areas relative to fall-grazed areas and spring and fall exclosures (Table E.1, Appendix E), this pattern was not reflected by total soil carbon and nitrogen or by root biomass. The relationship between aboveground and below-ground plant biomass and its effects on soil organic matter in these semi-arid grasslands warrants further investigation. Increased soil mechanical resistance between 1998 and 2008 would seem to suggest that compaction due to grazing is increasing. However, mechanical resistance of ungrazed areas has increased as well as that of grazed areas, suggesting that the factors responsible are affecting all areas rather than just grazed areas. This warrants further investigation and more frequent measurements both during the growing season and within the growing seasons. The establishment of more long-term grazing experiments on the grasslands of the southern interior of BC and elsewhere, as well as the continuation of this particular long-term experiment, are recommended because such long-term experiments permit the evaluation of the effects of management practices in real-life settings. The number of potential variables in such settings also enables many different kinds of studies to be run in the same area, and allows for the inclusion of new or different parameters.  75  References Abdel-Magid, A. H., Schuman, G. E., and Hart, R. H. 1987. Soil bulk density and water infiltration as affected by grazing systems. Journal of Range Management, 40: 307–309. Alexandrovskiy, A. L. 2007. Rates of soil-forming processes in three main models of pedogenesis. Revista Mexicana de Ciencias Geológicas, 24: 283-292. Amézketa, E. 1999. Soil aggregate stability: a review. Journal of Sustainable Agriculture, 14: 83-151. Andrews, S. S., and Carroll, C. R. 2001. Designing a soil quality assessment tool for sustainable agroecosystem management. Ecological Applications, 11: 1573-1585. Angers, D. A., and Mehuys, G. R. 1993. Aggregate stability to water. Pages 651–657 in M.R. Carter (ed.) Soil sampling and methods of analysis. Lewis Publishers, Boca Raton, FL. Aragón, A., Garcia, M. G., Filgueira, R. R., and Pachepsky, Y. A. 2000. Maximum compactibility of Argentine soils from the Proctor test: The relationship with organic carbon and water content. Soil and Tillage Research, 56: 197–204. Arnold, R. W., Szabolcs, I., and Targulian, V. O. 1990. Global soil change. Report of an IIASAISSS-UNEP task force on the role of soil in global change. International Institute for Applied Systems Analysis, Laxenburg, Austria. Arshad, M. A., and Coen, G. M. 1992. Characterization of soil quality: physical and chemical criteria. American Journal of Alternative Agriculture, 7: 25-31. Arshad, M. A., and Martin, S. 2002. Identifying critical limits for soil quality indicators in agroecosystems. Agriculture, Ecosystems and Environment, 88: 153-160. Banerjee, M. R., Burton, D. L., McCaughey, W. P., and Grant, C. A. 2000. Influence of pasture management on soil biological quality. Journal of Range Management, 53: 127–133. Bauer, A., Cole, C. V., and Black, A. L. 1987. Soil property comparisons in virgin grasslands between grazed and nongrazed management systems. Soil Science Society of America Journal, 51: 176182. Bawtree, A. H. 2005. A history of range use in British Columbia. Rangelands, 27: 36-39. BC Grasslands Conservation Council. 2009. GCC-Ranching Today in BC. [Online] Available: 76  http://www.bcgrasslands.org/ranchinginbc.htm (24 Jul. 2009) BC Ministry of Environment, Lands and Parks. 1998. Habitat Atlas for Wildlife at Risk: South Okanagan and Lower Similkameen. B.C. Ministry of Environment, Lands and Parks, Penticton, BC. 124 pp. Ben-Dor, E., and Banin, A. 1989. Determination of organic matter content in arid-zone soils using a simple “loss-on-ignition” method. Communications in Soil Science and Plant Analysis, 20: 1675–1695. Bilotta, G. S., Brazier, R. E., and Haygarth, P. M. 2007. The impacts of grazing animals on the quality of soils, vegetation, and surface waters in intensively managed grasslands. Advances in Agronomy, 94: 237–280. Biondini, M. E., and Manske, L. 1996. Grazing frequency and ecosystem processes in a northern mixed prairie, USA. Ecological Applications, 6: 239-256. Biondini, M. E., Patton, B. D., and Nyren, P. E. 1998. Grazing intensity and ecosystem processes in a northern mixed-grass prairie, USA. Ecological Applications, 8: 469-479. Bird, S. B., Herrick, J. E., Wander, M. M., and Wright, S. F. 2002. Spatial heterogeneity of aggregate stability and soil carbon in semi-arid rangeland. Environmental Pollution, 116: 445– 455. Blake, G. R., and Hartge, K. H. 1986. Bulk density. Pages 363–375 in C. A. Black (ed.) Methods of soil analysis, Part 1: Physical and mineralogical methods. ASA-SSSA, Madison, WI. Boone, R. D. 1994. Light-fraction soil organic matter: origin and contribution to net nitrogen mineralization. Soil Biology and Biochemistry, 26: 1459-1468. Bradford, J. M. 1986. Penetrability. Pages 463-478 in A. Klute and A. L. Page (ed.) Methods of soil analysis, vol. 1. ASA-SSSA, Madison, WI. Brady, N. C., and Weil, R. R. 2007. The Nature and Properties of Soils. 13th ed. Prentice Hall, Upper Saddle River, New Jersey. 965 pp. Brejda, J. J., Moorman, T. B., Karlen, D. L., and Dao, T. H. 2000. Identification of regional soil quality factors and indicators: I. central and southern high plains. Soil Science Society of America Journal, 64: 2115-2124.  77  British Columbia Cattlemen's Association. 2009. BC Beef Industry Statistics. [Online] Available: http://www.cattlemen.bc.ca/statistics.htm (24 Jul. 2009) Broersma, K., Krzic, M., Thompson, D., and Bomke, A. 2000. Soil and vegetation of ungrazed crested wheatgrass and native rangelands. Canadian Journal of Soil Science, 80: 411–417. Busscher, W. 1990. Adjustment of flat-tipped penetrometer resistance data to a common water content. Transactions of the American Society of Agricultural Engineers, 33: 519–524. Busscher, W., Sojka, R., Doty, C., and USDA, A. 1986. Residual effects of tillage on coastal plain soil strength. Soil Science, 141: 144–148. Caldwell, M., Richards, J., Johnson, D., Nowak, R., and Dzurec, R. 1981. Coping with herbivory: Photosynthetic capacity and resource allocation in two semiarid Agropyron bunchgrasses. Oecologia, 50: 14-24. Campbell, C., and Souster, W. 1982. Loss of organic matter and potentially mineralizable nitrogen from Saskatchewan soils due to cropping. Canadian Journal of Soil Science, 62: 651–656. Campbell, C., and Bawtree, A. 1998. Rangeland handbook for B.C. British Columbia Cattlemen's Association, Kamloops, BC. 203 pp. Capriel, P., Beck, T., Borchert, H., and Harter, P. 1990. Relationship between soil aliphatic fraction extracted with supercritical hexane, soil microbial biomass, and soil aggregate stability. Soil Science Society of America journal (USA). Carder, A. 1970. Climate and the rangelands of Canada. Journal of Range Management, 23: 263-267. Carter, M. R. 1996. Concepts of soil quality. Pages 5-9 in Soil Quality is in the Hands of the Land Manager. Centre for Environmental Management, University of Ballarat, Ballarat, Victoria, Australia. Carter, M. R. 2002a. Quality, critical limits and standardization. Pages 1062–1065 in R. Lal (ed.) Encyclopedia of soil science. Marcel Dekker, New York, NY. Carter, M. R. 2002b. Soil quality for sustainable land management: organic matter and aggregation interactions that maintain soil functions. Agronomy Journal, 94: 38-47. Carter, M. R., Gregorich, E. G., Anderson, D. W., Doran, J. W., Janzen, H. H., and Pierce, F. J. 1997. Concepts of soil quality and their significance. Pages 1-19 in E.G. Gregorich, M.R. Carter 78  (ed.) Soil quality for crop production and ecosystem health. Elsevier, Amsterdam. Carter, M. R., and Kunelius, H. T. 1993. Effect of undersowing barley with annual ryegrasses or red clover on soil structure in a barley-soybean rotation. Agriculture, Ecosystems & Environment, 43: 245–254. Churchman, G., and Tate, K. 1987. Stability of aggregates of different size grades in allophanic soils from volcanic ash in New Zealand. European Journal of Soil Science, 38: 19–27. Compton, J., and Boone, R. 2002. Soil nitrogen transformations and the role of light fraction organic matter in forest soils. Soil Biology and Biochemistry, 34: 933-943. Courtin, P., Feller, M., and Klinka, K. 1983. Lateral Variability in Some Properties of Disturbed Forest Soils in Southwestern British Columbia. Canadian Journal of Soil Science, 63: 529–539. Delesalle, B. P., Coupe, B. J., Wikeem, S. J., and Wikeem, B. M. 2009. Grassland monitoring manual for British Columbia: a tool for ranchers. Grasslands Conservation Council of British Columbia, Kelowna, BC. 72 pp. Detling, J. K., Dyer, M. I., and Winn, D. T. 1979. Net photosynthesis, root respiration, and regrowth of Bouteloua gracilis following simulated grazing. Oecologia, 41: 127-134. Dexter, A. 1988. Advances in characterization of soil structure. Soil and Tillage Research, 11: 199-238. Donkor, N., Gedir, J., Hudson, R., Bork, E., Chanasyk, D., and Naeth, M. 2002. Impacts of grazing systems on soil compaction and pasture production in Alberta. Canadian Journal of Soil Science, 82: 1-8. Doran, J. W., and Parkin, T. B. 1994. Defining and assessing soil quality. Pages 3-21 in Doran, J.W., Coleman, D.C., Bezdicek, D.F., and Stewart, B.A. (ed.) Defining Soil Quality for a Sustainable Environment. Soil Science Society of America, Madison, WI. Dormaar, J. F., Adams, B. W., and Willms, W. D. 1997. Impacts of rotational grazing on mixed prairie soils and vegetation. Journal of Range Management, 50: 647-651. Dormaar, J. F., Johnston, A., and Smoliak, S. 1977. Seasonal variation in chemical characteristics of soil organic matter of grazed and ungrazed mixed prairie and fescue grassland. Journal of Range Management, 30: 195-198. Dormaar, J. F., Johnston, A., and Smoliak, S. 1984. Seasonal changes in carbon content, and 79  dehydrogenase, phosphatase, and urease activities in mixed prairie and fescue grassland ah horizons. Journal of Range Management, 37: 31-35. Dormaar, J. F., Smoliak, S., and Willms, W. D. 1989. Vegetation and soil responses to short-duration grazing on fescue grasslands. Journal of Range Management, 42: 252–256. Dormaar, J. F., and Willms, W. D. 1998. Effect of forty-four years of grazing on fescue grassland soils. Journal of Range Management, 51: 122-126. Evanko, A., and Peterson, R. 1955. Comparisons of protected and grazed mountain rangelands in southwestern Montana. Ecology, 36: 71-82. Evans, R. 1998. The erosional impacts of grazing animals. Progress in Physical Geography, 22: 251268. Frank, A., Tanaka, D., Hofmann, L., and Follett, R. 1995. Soil carbon and nitrogen of northern Great Plains grasslands as influenced by long-term grazing. Journal of Range Management, 48: 470-474. Fraser, D. 2007. Rangeland Health Field Guide. B.C. Ministry of Forests, Research Branch, Kelowna, BC. 144 pp. [Online] Available: http://www.for.gov.bc.ca/hfd/pubs/Docs/Mr/Mr117.htm Gamougoun, N., Smith, R., Wood, M., and Pieper, R. 1984. Soil, vegetation, and hydrologic responses to grazing management at Fort Stanton, New Mexico. Journal of Range Management, 37: 538-541. Ganjegunte, G., Vance, G., Preston, C., Schuman, G., Ingram, L., Stahl, P., and Welker, J. 2005. Soil organic carbon composition in a northern mixed-grass prairie: effects of grazing. Soil Science Society of America Journal, 69: 1746-1756. Gardner, W. H. 1965. Water content. Pages 82-127 in C. A. Black (ed.) Methods of soil analysis, Part 1: Physical and mineralogical methods. ASA, Madison, WI. Gayton, D. 2003. British Columbia grasslands: monitoring vegetation change. FORREX, Kamloops, BC. 58 pp. Gifford, G., and Hawkins, R. 1978. Hydrologic impact of grazing on infiltration: a critical review. Water Resour. Res., 14: 305–313. Gill, R. 2007. Influence of 90 years of protection from grazing on plant and soil processes in the 80  subalpine of the Wasatch Plateau, USA. Rangeland Ecology and Management, 60: 88-98. Government of Canada, E. C., and Meteorological Service of Canada, E. C. 2010. Canadian Climate. [Online] Available: http://climate.weatheroffice.gc.ca/climate_normals/results_e.html? Province=ALL&StationName=kamloops&SearchType=BeginsWith&LocateBy=Province&Pro ximity=25&ProximityFrom=City&StationNumber=&IDType=MSC&CityName=&ParkName= &LatitudeDegrees=&LatitudeMinutes=&LongitudeDegrees=&LongitudeMinutes=&NormalsCl ass=A&SelNormals=&StnId=1275& (25 Mar. 2010) Grasslands Conservation Council of British Columbia. 2010. Species At Risk in BC's Grasslands. [Online] Available: http://www.bcgrasslands.org/sar_list.htm (17 Jul. 2010) Greacen, E., Barley, K., and Farrell, D. 1969. The mechanics of root growth in soils with particular reference to the implications for root distribution. Pages 256-269 in W.J. Whittington (ed.), Root growth. Butterworths, London, U.K. Greacen, E., and Sands, R. 1980. Compaction of forest soils. A review. Australian Journal of Soil Research, 18: 163-189. Greenwood, K., and Hutchinson, K. 1998. Root characteristics of temperate pasture in New South Wales after grazing at three stocking rates for 30 years. Grass and Forage Science, 53: 120-128. Gregorich, E. G., Carter, M. R., Angers, D. A., Monreal, C. M., and Ellert, B. H. 1994. Towards a minimum data set to assess soil organic matter quality in agricultural soils. Canadian Journal of Soil Science, 74: 367-385. Gregorich, E. G., and Ellert, B. H. 1993. Light fraction and macroorganic matter in mineral soils. Pages 397–407 in M.R. Carter (ed.) Soil sampling and methods of analysis. Canadian Soil Science Society, Lewis Publishers, Boca Raton, FL. Gregorich, E. G., and Janzen, H. H. 1996a. Storage of soil carbon in the light fraction and macroorganic matter. Structure and organic matter storage in agricultural soils, 167–190. Gregorich, E. G., and Janzen, H. H. 1996b. Storage of soil carbon in the light fraction and macroorganic matter. Pages 167–190 in M.R. Carter and B.A. Stewart (ed.) Structure and organic matter storage in agricultural soils. Lewis Publishers, Boca Raton, FL. Grime, J. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. The American Naturalist, 111: 1169-1194. 81  Harris, R., and Bezdicek, D. 1994. Descriptive aspects of soil quality/health. Pages 23-35 in J.W. Doran, D.C.Coleman, D.F. Bezdicek, and B.A. Stewart (ed.) Defining soil quality for a sustainable environment. Soil Science Society of America, Madison, WI. van Haveren, B. P. 1983. Soil bulk density as influenced by grazing intensity and soil type on a shortgrass prairie site. Journal of Range Management, 36: 586-588. Haynes, R., and Swift, R. 1990. Stability of soil aggregates in relation to organic constituents and soil water content. Eur J Soil Science, 41: 73-83. Hendershot, W., and Lalande, H. 1993. Soil reaction and exchangeable acidity. Pages 141–145 in M.R. Carter (ed.) Soil sampling and methods of analysis. Canadian Soil Science Society, Lewis Publishers, Boca Raton, FL. Henderson, D., Ellert, B., and Naeth, M. 2004. Grazing and soil carbon along a gradient of Alberta rangelands. Journal of Range Management, 57: 402-410. Hermawan, B., and Bomke, A. 1996. Aggregation of a degraded lowland soil during restoration with different cropping and drainage regimes. Soil Technology, 9: 239-250. Herrick, J., Brown, J., Tugel, A., Shaver, P., and Havstad, K. 2002. Application of soil quality to monitoring and management: paradigms from rangeland ecology. Agronomy Journal, 94: 3-11. Herrick, J., and Whitford, W. 1995. Assessing the quality of rangeland soils: challenges and opportunities. Journal of Soil and Water Conservation, 50: 237-242. Holechek, J., Pieper, R., and Herbel, C. 2004. Range management: principles and practices. Prentice Hall, Upper Saddle River, NJ. 444 pp. Holland, E., and Detling, J. 1990. Plant response to herbivory and belowground nitrogen cycling. Ecology, 71: 1040-1049. Janzen, H. 1987. Soil organic matter characteristics after long-term cropping to various spring wheat rotations. Canadian Journal of Soil Science, 67: 845–856. Jenny, H. 1980. The soil resource: origin and behavior. Springer, New York. 377 pp. Johnston, A., Dormaar, J. F., and Smoliak, S. 1971. Long-term grazing effects on fescue grassland soils. Journal of Range Management, 24: 185–188. Karlen, D. L., Mausbach, M. J., Doran, J. W., Cline, R. G., Harris, R. F., and Schuman, G. E. 82  1997. Soil quality: a concept, definition, and framework for evaluation. Soil Science Society of America Journal, 61: 4-10. Karlen, D., and Stott, D. 1994. A framework for evaluating physical and chemical indicators of soil quality. Pages 53-72 in Doran, J.W., Coleman, D.C., Bezdicek, D.F., and Stewart, B.A. (ed.) Defining soil quality for a sustainable environment. Soil Science Society of America, Madison, WI. Kay, B., Lal, R., Kimble, J., Follett, R., and Stewart, B. 1998. Soil structure and organic carbon: a review. Pages 169-197 in R. Lal (ed.) Soil processes and the carbon cycle. CRC Press, Boca Raton, FL. Knoll, G., and Hopkins, H. 1959. The effects of grazing and trampling upon certain soil properties. Transactions of the Kansas Academy of Science, 62: 221-231. Krzic, M., Broersma, K., Thompson, D., and Bomke, A. 2000. Soil properties and species diversity of grazed crested wheatgrass and native rangelands. Journal of Range Management, 53: 353358. Krzic, M., Newman, R., Broersma, K., and Bomke, A. 1999. Soil compaction of forest plantations in Interior British Columbia. Journal of Range Management, 52: 671-677. Lacey, J., and van Poollen, H. 1981. Comparison of herbage production on moderately grazed and ungrazed western ranges. Journal of Range Management, 34: 210-212. Lamagna, S. 2008. Development of health indicators for rough fescue grasslands in the Southern Interior of British Columbia, University of British Columbia, Vancouver, BC. Larson, W., and Pierce, F. 1991. Conservation and enhancement of soil quality. Pages 175–203 in Evaluation for sustainable land management in the developing world: proceedings of the international workshop on evaluation for sustainable land management in the developing world, vol. 2. International Board Soil Research Management, Bangkok, Thailand. Larson, W., and Pierce, F. 1994. The dynamics of soil quality as a measure of sustainable management. Pages 37-51 in Doran, J.W., Coleman, D.C., Bezdicek, D.F., and Stewart, B.A. (ed.) Defining soil quality for a sustainable environment, vol. 35. Soil Science Society of America, Madison, WI. Laycock, W., and Conrad, P. 1967. Effect of grazing on soil compaction as measured by bulk density 83  on a high elevation cattle range. Journal of Range Management, 20: 136-140. Lodge, R. W. 1954. Effects of grazing on the soils and forage of mixed prairie in southwestern Saskatchewan. Journal of Range Management, 7: 166-170. Lorenz, R. J., and Rogler, G. A. 1967. Grazing and fertilization affect root development of range grasses. Journal of Range Management, 20: 129-132. Lowe, L. 1993. Total and labile polysaccharide analysis of soils. Pages 373–376 in M.R. Carter (ed.) Soil sampling and methods of analysis. CRC Press, Boca Raton, FL. Mack, R. N., and Thompson, J. N. 1982. Evolution in steppe with few large, hooved mammals. The American Naturalist, 119: 757-773. Manley, J., Schuman, G., Reeder, J., and Hart, R. 1995. Rangeland soil carbon and nitrogen responses to grazing. Journal of Soil and Water Conservation, 50: 294-298. Mapfumo, E., Chanasyk, D. S., Baron, V. S., and Naeth, M. A. 2000. Grazing impacts on selected soil parameters under shortterm forage sequences. Journal of Range Management, 53: 466-470. Mapfumo, E., Chanasyk, D. S., Naeth, M. A., and Baron, V. 1999. Soil compaction under grazing of annual and perennial forages. Canadian Journal of Soil Science, 79: 191–199. Mapfumo, E., Naeth, M. A., Baron, V. S., Dick, A. C., and Chanasyk, D. S. 2002. Grazing impacts on litter and roots: perrenial versus annual grasses. Journal of Range Management, 55: 16-22. Martin, S., Baize, D., Bonneau, M., Chaussod, R., Gaultier, J. P., Lavelle, P., Legros, J. P., Leprêtre, A., and Sterckeman, T. 1998. The French National "Soil Quality Observatory". in 16th World Congress of Soil Science, vol. 25. Montpellier, France. McGill, W. B., and Figueiredo, C. T. 1993. Total nitrogen. Pages 201–211 in M.R. Carter (ed.) Soil sampling and methods of analysis. Canadian Soil Science Society, Lewis Publishers, Boca Raton, FL. McLean, E. O. 1982. Soil pH and lime requirement. Pages 199-224 in A. L. Page, R. H. Miller, and D. R. Keeney (ed.) Methods of Soil Analysis, Part 2: Chemical and microbiological properties. American Society of Agronomy, Madison, WI. McNaughton, S. J. 1983. Compensatory plant growth as a response to herbivory. Oikos, 40: 329-336. Metcalfe, D. B. et al. 2007. A method for extracting plant roots from soil which facilitates rapid sample 84  processing without compromising measurement accuracy. New Phytologist, 174: 697-703. Milchunas, D. G., and Lauenroth, W. K. 1993. Quantitative effects of grazing on vegetation and soils over a global range of environments. Ecological Monographs, 63: 328-366. Molloy, L. F., Bridger, B. A., and Cairns, A. 1977. Studies on a climosequence of soils in tussock grasslands. New Zealand Journal of Science, 20: 167–77. Naeth, M. A., Bailey, A. W., Pluth, D. J., Chanasyk, D. S., and Hardin, R. T. 1991. Grazing impacts on litter and soil organic matter in mixed prairie and fescue grassland ecosystems of Alberta. Journal of Range Management, 44: 7-12. Naeth, M. A., and Chanasyk, D. S. 1995a. Grazing impacts on bulk density and soil strength in the foothills fescue grasslands of Alberta, Canada. Canadian Journal of Soil Science, 75: 551-557. Naeth, M. A., and Chanasyk, D. S. 1995b. Grazing effects on soil water in Alberta foothills fescue grasslands. Journal of Range Management, 48: 528-534. Naeth, M. A., Chanasyk, D. S., Rothwell, R. L., and Bailey, A. W. 1991. Grazing impacts on soil water in mixed prairie and fescue grassland ecosystems of Alberta. Canadian Journal of Soil Science, 71: 313-325. Naeth, M. A., Pluth, D. J., Chanasyk, D. S., Bailey, A. W., and Fedkenheuer, A. W. 1990. Soil compacting impacts of grazing in mixed prairie and fescue grassland ecosystems of Alberta. Canadian Journal of Soil Science, 70: 157–167. Nelson, D. W., and Sommers, L. E. 1982. Total carbon, organic carbon, and organic matter. Pages 539-579 in A. L. Page, R. H. Miller, and D. R. Keeney (ed.) Methods of soil analysis. Part 2, vol. 2. ASA-SSSA, Madison, WI. Newman, R., Lamagna, S., Wallace, B., Tucker, R., Njenga, F., and Krzic, M. 2008. Development of cost-effective health indicators for fescue grasslands of British Columbia. Forest Science Program Final Report, Kamloops, BC. Newman, R., Tucker, R., Njenga, F., Krzic, M., Wurtz, S., and Hope, G. D. 2007. Linking Range Health Assessment Methodology with Science: Final Report. Forest Investment Account (FIA) Forest Science Program. Kamloops, BC. 46 pp. Nimmo, J., and Perkins, K. S. 2002. Aggregate stability and size distribution. Pages 317-328 in J. H.  85  Dane and G. C. Topp (ed.) Methods of soil analysis, Part 4: Physical methods. Soil Science Society of America, Madison, WI. Norfleet, M. L., Ditzler, C. A., Puckett, W. E., Grossman, R. B., and Shaw, J. N. 2003. Soil quality and its relationship to pedology. Soil Science, 168: 149-155. Olsen, S. R., Cole, C. V., Watanabe, F. S., and Dean, L. A. 1954. Estimation of available phosphorus in soils by extraction with sodium bicarbonate. USDA circ, 939: 1–19. Olson, G. L., McQuaid, B. F., Easterling, K. N., and Scheyer, J. M. 1996. Quantifying soil condition and productivity in Nebraska. Pages 357-370 in Doran, J.W., Jones, A. J. (ed.) Methods for assessing soil quality. Soil Science Society of America, Madison, WI. Orr, H. K. 1960. Soil Porosity and Bulk Density on Grazed and Protected Kentucky Bluegrass Range in the Black Hills. Journal of Range Management, 13: 80-86. Petersen, R. G., Woodhouse, W. W., and Lucas, H. L. 1956. The distribution of excreta by freely grazing cattle and its effect on pasture fertility: II. Effect of returned excreta on the residual concentration of some fertilizer elements. Agronomy Journal, 48: 444-449. Pieper, R. 2005. Grasslands of central North America. Pages 221-264 in Grasslands of the world. Food and Agriculture Organization, Rome, Italy. Proffitt, A., Bendotti, S., Howell, M., and Eastham, J. 1993. The effect of sheep trampling and grazing on soil physical properties and pasture growth for a red-brown earth. Australian Journal of Agricultural Research, 44: 317-331. Pyke, D. A., Herrick, J. E., Shaver, P., and Pellant, M. 2002. Rangeland health attributes and indicators for qualitative assessment. Journal of Range Management, 55: 584-597. Rezaei, S. A., Gilkes, R. J., and Andrews, S. S. 2006. A minimum data set for assessing soil quality in rangelands. Geoderma, 136: 229-234. Rodd, A. V., Papadopoulos, Y. A., Laflamme, L. F., McRae, K. B., Fillmore, S. A. E., and Wilson, R. W. 1999. Effect of rotational grazing on selected physical properties of a gleyed brunisolic gray Luvisol loam in Nova Scotia. Canadian Journal of Soil Science, 79: 117-125. van Ryswyk, A., Marchand, L. S., and MacLean, A. 1966. The climate, native vegetation, and soils of some grasslands at different elevations in British Columbia. Canadian Journal of Plant  86  Science, 46: 35-50. van Ryswyk, A., and McLean, A. 1989. Geology and soils of grassland ranges, Kamloops, British Columbia. Rangelands, 11: 65-67. SAS Institute Inc. 2003. SAS/STAT 9.1. 4th ed. SAS Institute Inc., Cary, N.C., USA. Schuman, G. E., Reeder, J. D., Manley, J. T., Hart, R. H., and Manley, W. A. 1999. Impact of grazing management on the carbon and nitrogen balance of a mixed-grass rangeland. Ecological Applications, 9: 65-71. Seybold, C. A., Herrick, J. E., and Brejda, J. J. 1999. Soil resilience: a fundamental component of soil quality. Soil Science, 164: 224-234. Six, J., Elliott, E. T., Paustian, K., and Doran, J. W. 1998. Aggregation and soil organic matter accumulation in cultivated and native grassland soils. Soil Science Society of America Journal, 62: 1367-1377. Smith, J. L., and Doran, J. W. 1996. Measurement and use of pH and electrical conductivity for soil quality analysis. Pages 169-186 in Doran, J.W., Jones, A. J. (ed.) Methods for assessing soil quality. Soil Science Society of America, Madison, WI. Smoliak, S., Dormaar, J. F., and Johnston, A. 1972. Long-term grazing effects on stipa-bouteloua prairie soils. Journal of Range Management, 25: 246-250. Snyman, H. A. 2005. Rangeland degradation in a semi-arid South Africa--I: influence on seasonal root distribution, root/shoot ratios and water-use efficiency. Journal of Arid Environments, 60: 457481. Snyman, H. A., and du Preez, C. C. 2005. Rangeland degradation in a semi-arid South Africa--II: influence on soil quality. Journal of Arid Environments, 60: 483-507. Soane, B. 1990. The role of organic matter in soil compactibility: A review of some practical aspects. Soil and Tillage Research, 16: 179-201. Soil Survey Laboratory Staff. 2004. Soil survey laboratory methods manual. Soil Surv. Invest. Reps. USDA-SCS, Washington, DC. 700 pp. Sollins, P., Spycher, G., and Topik, C. 1983. Processes of soil organic-matter accretion at a mudfloe chronosequence, Mt. Shasta, California. Ecology, 64: 1273-1282. 87  Statistics Canada. 2009. British Columbia- total cattle and calves: farms reporting, number of animals, and percentage distributions by size class, Canada and provinces, census years 1976 to 2006. [Online] Available: http://www.statcan.gc.ca/pub/95-632-x/2007000/t/4185583-eng.htm (25 Feb. 2010) Strickland, T. C., and Sollins, P. 1987. Improved method for separating light- and heavy-fraction organic material from soil. Soil Science Society of America Journal, 51: 1390-1393. Task Group on Unity in Concepts and Terminology Committee. 1995. New concepts for assessment of rangeland condition. Journal of Range Management, 48: 271–282. Taylor, H. M., Roberson, G. M., and Parker Jr, J. J. 1966. Soil strength-root penetration relations for medium-to coarse-textured soil materials. Soil Science, 102: 18-22. Tilman, D. 1985. The resource-ratio hypothesis of plant succession. American Naturalist, 125: 827852. Tisdale, E. W. 1947. The grasslands of the Southern Interior of British Columbia. Ecology, 28: 346382. Tisdall, J. M., and Oades, J. M. 1982. Organic matter and water-stable aggregates in soils. European Journal of Soil Science, 33: 141-163. Vallentine, J. F. 1989. Range development and improvements. 2nd ed. Bringham Young University Press, Provo, Utah. 524 pp. Voorhees, W. B., Senst, C. G., and Nelson, W. W. 1978. Compaction and soil structure modification by wheel traffic in the northern corn belt. Soil Science Society of America Journal, 42: 344-349. Wallace, B. M., Krzic, M., Forge, T. A., Broersma, K., and Newman, R. F. 2009. Biosolids increase soil aggregation and protection of soil carbon five years after application on a crested wheatgrass pasture. Journal of Environmental Quality, 38: 291-298. Whalen, J. K., Bottomley, P. J., and Myrold, D. D. 2000. Carbon and nitrogen mineralization from light- and heavy-fraction additions to soil. Soil Biology and Biochemistry, 32: 1345-1352. Wikeem, B. M., McLean, A., Bawtree, A., and Quinton, D. 1993. An overview of the forage resource and beef production on Crown land in British Columbia. Canadian Journal of Animal Science, 73: 779-779.  88  Wikeem, B. M., and Wikeem, S. J. 2004. The grasslands of British Columbia. Grasslands Conservation Council of BC, Kamloops, BC. 479 pp. Willms, W. D., Dormaar, J. F., Adams, B. W., and Douwes, H. E. 2002. Response of the mixed prairie to protection from grazing. Journal of Range Management, 55: 210-216. Willms, W. D., Smoliak, S., and Bailey, A. W. 1986. Herbage production following litter removal on Alberta native grasslands. Journal of Range Management, 39: 536–539. Willms, W. D., Smoliak, S., and Dormaar, J. F. 1985. Effects of stocking rate on a rough fescue grassland vegetation. Journal of Range Management, 38: 220-225. Yoder, R. E. 1936. A direct method of aggregate analysis of soils and a study of the physical nature of erosion losses. Agronomy Journal, 28: 337-351. Young, G., Fenger, A. M., and Luttmerding, H. A. 1992. Soils of the Ashcroft map area. BC Soil Survey Rep. 26. BC Environment, Integrated Management Branch, Victoria, BC. 233 pp. Zhang, H. 1994. Organic matter incorporation affects mechanical properties of soil aggregates. Soil and Tillage Research, 31: 263-275. Zhao, Y., Krzic, M., Bulmer, C. E., and Schmidt, M. G. 2008. Maximum Bulk Density of British Columbia Forest Soils from the Proctor Test: Relationships with Selected Physical and Chemical Properties. Soil Science Society of America Journal, 72: 442-452.  89  Appendix A- Correlations Aggregate MWD and Size Fraction Correlation Matrix Table A.1: Aggregate size fraction correlation coefficients. All data from 0-7.5-cm depth. n=16 for all correlations. MWD r  p  2-6-mm aggregates r p  1-2-mm aggregates r p  0.25-1-mm aggregates r p  <0.25-mm aggregates r p  pH (CaCl2)  -0.38 0.1482  -0.39 0.1353  0.13 0.6270  0.23 0.3945  0.36 0.1767  pH (H2O)  0.58 0.0190  0.57 0.0213  0.43 0.0951  -0.40 0.1278  -0.59 0.0167  Available phosphorus  0.44 0.0908  0.44 0.0881  0.09 0.7528  -0.27 0.3205  -0.43 0.0981  Polysaccharides  -0.59 0.0168  -0.56 0.0227  -0.68 0.0040  0.04 0.8854  0.64 0.0074  % polysaccharides  -0.62 0.0097  -0.60 0.0131  -0.63 0.0083  0.06 0.8205  0.67 0.0041  % carbon as polysaccharides  -0.71 0.0020  -0.70 0.0027  -0.59 0.0160  0.31 0.2437  0.74 0.0010  Total carbon  0.07 0.7992  0.08 0.7598  -0.22 0.4202  -0.38 0.1431  -0.02 0.9271  Total nitrogen  -0.01 0.9771  0.01 0.9831  -0.23 0.3976  -0.37 0.1554  0.05 0.8459  % total carbon  0.05 0.8414  0.07 0.7924  -0.29 0.2791  -0.41 0.1104  0.00 0.9936  % total nitrogen  -0.01 0.9614  0.00 0.9895  -0.30 0.2655  -0.40 0.1199  0.07 0.8083  Carbon-tonitrogen ratio  0.38 0.1493  0.38 0.1458  0.03 0.9185  -0.07 0.7923  -0.38 0.1485  Root biomass  0.54 0.0317  0.54 0.0296  0.09 0.7281  -0.42 0.1046  -0.52 0.0393  Bulk density  0.56 0.0246  0.54 0.0296  0.50 0.0506  -0.09 0.7465  -0.60 0.0150  Mechanical resistance  0.02 0.9342  0.00 0.9900  0.41 0.1158  0.19 0.4795  -0.07 0.8022  Soil water  -0.80 0.0002  -0.78 0.0003  -0.54 0.0295  0.19 0.4779  0.83 0.0001  90  Soil Quality Indicator Correlation Matrix Table A.2: Correlation matrix for selected soil quality indicators. Data were averaged by exclosure location. n=48 for all correlations.  pH (CaCl2)  Available phopshorus % Total carbon % Total nitrogen Carbon-to-nitrogen ratio % polysaccharides Root biomass Bulk density Mechanical resistance Coarse fragments  r 0.37 -0.55 -0.26 -0.23 -0.01 0.15 -0.31 -0.31 0.27 0.34  % polysaccharides Root biomass Bulk density Mechanical resistance Coarse fragments  Carbon-tonitrogen ratio r p -0.19 0.1932 0.31 0.0307 0.25 0.0804 -0.25 0.0832 -0.04 0.7919  pH (H2O)  p 0.0103 0.0001 0.0789 0.1034 0.9545 0.3053 0.0329 0.0337 0.0641 0.0169  pH (H2O) r  p  -0.20 -0.21 -0.26 0.23 -0.56 -0.06 0.52 0.09 0.49  0.1657 0.1598 0.0712 0.1176 0.0001 0.7071 0.0002 0.5350 0.0004  % polysaccharides  Available phosphorus r p  0.52 0.43 0.24 -0.01 0.46 0.32 -0.69 -0.59  0.0002 0.0020 0.0987 0.9304 0.0010 0.0273 0.0001 0.0001  Root biomass  % Total carbon  % Total nitrogen  r  p  r  p  0.97 0.12 0.49 0.47 -0.10 -0.70 -0.72  0.0001 0.4083 0.0004 0.0006 0.5126 0.0001 0.0001  -0.11 0.55 0.38 -0.17 -0.63 -0.70  0.4597 0.0001 0.0073 0.2543 0.0001 0.0001  Bulk density  r  p  r  p  r  p  0.04 -0.69 -0.26 -0.52  0.7774 0.0001 0.0708 0.0002  0.18 -0.57 -0.40  0.2164 0.0001 0.0045  -0.13 0.14  0.3630 0.3384  Mechanical resistance r p  0.73  0.0001  91  Appendix B- Pasture and Soil Series Descriptions At the lower end of the elevation gradient, exclosures (a) and (b) were located in the lower spring pasture on the McKnight series, characterized as a moderately alkaline and weakly saline Orthic Brown Chernozem (with up to 40% inclusion of Lithic Brown Chernozemic soil) developed from mediumtextured morainal deposits, with a thin, loamy eolian capping and slight to moderate stoniness (Figure B.1). In the lower fall pasture, exclosure (c) also occurred on this soil series, and all three exclosures were found within the BG zone. Exclosure (d) was located on the Glimpse series, which occurs on the border of the IDF zone. It is characterized as a mildly to moderately alkaline Orthic Dark Brown Chernozem developed from fluvio-glacial deposits derived from a variety of bedrock, with moderate to high stoniness. The vegetation is dominated by grasses and shrubs due to disturbances such as grazing and fire. In the middle fall and middle spring pastures, exclosures (e), (f), (g), and (h) were on the McQueen series, which is a moderately alkaline Orthic Dark Brown Chernozem developed from morainal deposits associated with basic volcanic and limestone bedrock, with slight stoniness. Vegetation on the McQueen series is similar to that of the Glimpse series, also due to grazing and fire (Young et al., 1992). Soil texture was similar across the entire study site (an average of 10% clay, 44% silt, and 46% sand). The clay fraction in this area is dominated by vermiculite and montmorillonite, with smaller amounts of quartz, feldspar, chlorite, illite, and amphibole (L. Lavkulich, personal communication, 2010).  92  Figure B.1: The layout of the long-term grazing study in the Lac du Bois Rangeland north of Kamloops, BC  93  Appendix C- ANOVA Table Table C.1: ANOVA Table for completely randomized split-plot design with subsampling for assessment of grazing effects on grassland soil quality after 20 years (i.e., 1998) and 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC  Source  Degrees of freedom  Expected Mean Squares  Time of grazing  2-1=1  TR 8  2R 4  2R  2 A  2 R A    A  2 R B A   B A   Error 1  4-1(2)=6  8  Stocking rate  2-1=1  TR 4  2R  2  Interaction  (2-1)(2-1)=1  TR 4  2R  2  4  B  B A  AB  Error 2  (4-1)(2-1)2=6  4  Sampling Error  (4-1)*2*2*4=48    Total  2*2*4*4-1=63  2 R B A    2   B A    2   2  94  Appendix D- Additional Figures 60  precipitation (mm)  50  (a)  1997-1998 2007-2008 1971-2000 average  40  *  30 20 10 0 Jan 25  Feb  Mar  Apr  May  Jun  Jul  Aug  Sep  Oct  Nov  Dec  Jan  Feb  Mar  Apr  May  (b)  *  20  temperature (° C)  Jun  15 10 5 0 -5 Jan  Feb  Mar  Apr  May  Jun  Jul  Aug  Sep  Oct  Nov  Dec  Jan  Feb  Mar  Apr  May  Jun  Figure D.1: Monthly precipitation (a) and monthly average temperature (b) at Kamloops airport, BC in 1997-1998 and 2007-2008 compared to 1971-2000 averages. The asterisks indicate sampling dates in 1998 and 2008. 0.500 0.400  0-7.5 cm  0.500 0.400  7.5-15 cm  0.500 0.400  0.300  0.300  0.300  0.200  0.200  0.200  0.100 0.000  0.100  f(x) = 0.07 ln(x) + 0.22 1 2 3 4 5 6 7 8 9 101112  0.000  f(x) = 0.04 ln(x) + 0.22 1 2 3 4 5 6 7 8 9 101112  15-30 cm  0.100 0.000  f(x) = 0.02 ln(x) + 0.12 1 2 3 4 5 6 7 8 9 101112  Figure D.2: Logarithmic regressions based on 4 extractions compared to 12 extractions on the same sample for the 0-7.5, 7.5-15, and 15-30-cm depths. The vertical axis measures root mass density in grams (g cm-3), while the horizontal axis represents the number of sequential 10-minute extractions.  95  Figure D.3: Total daily rainfall at Kamloops airport, BC, January 1997-June 2008  Figure D.4: Aggregate water content prior to wet-sieving analysis under spring and fall grazing and 0 and 2 AUM ha -1 treatments as determined 30 years (i.e., 2008) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=32). Differences were considered significant at p<0.10 and marked by an asterisk.  96  Figure D.5: Relationship between total carbon and total nitrogen as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC.  Figure D.6: Relationship between total carbon and mechanical resistance as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC. Data for soil mechanical resistance were averaged over the 07.5, 7.5-15, and 15-30-cm depths for the purposes of the comparison.  97  Figure D.7: Relationship between soil mechanical resistance and root biomass as obtained on the long-term grazing experiment 20 and 30 years after its establishment on a rangeland north of Kamloops, BC. Data for soil mechanical resistance were averaged over the 0-7.5-, 7.5-15-, and 15-30-cm depths for the purposes of the comparison.  98  Figure D.8: Time of grazing and stocking rate treatment effects on bareground, litter, rock, and species cover as determined 32 years (i.e., 2010) after establishment of an experiment on rangeland north of Kamloops, BC. Error bars represent the standard deviation of the mean (n=64). Differences were considered significant at p<0.10 and marked by an asterisk.  99  Appendix E- Additional Tables Table E.1: Grazing effects on above-ground biomass as determined 20 years (i.e., 1998) after establishment of an experiment on rangeland north of Kamloops, BC. Spring Fall S-F(UG) S-F(G) (S)UG-G (F)UG-G Ungrazed Grazed Ungrazed Grazed t ha-1 p 1.70 (0.799)† 1.01 (0.733) 0.0099‡ 1.59 (0.470) 1.53 (0.484) 0.4625 0.0026 0.6623 † The bracketed number represents the standard deviation of the mean (n=40). ‡ Based on four meaningful treatment mean comparisons, a Bonferonni correction of 0.10÷4=0.025 was used for α  100  Appendix F- Plates  Plate F.1: Vegetation differences between a spring grazing treatment (right) and a fall grazing treatment (left).  Plate F.2: Illustration of a typical sampling sites in the Lac du Bois grasslands.  101  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0071583/manifest

Comment

Related Items