Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ecology of the microbiotic crust of the antelope-brush (Purshia tridentata) shrub steppe of the South… Atwood, Lynne Beryl 1998

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

Item Metadata

Download

Media
831-ubc_1998-0347.pdf [ 6.45MB ]
Metadata
JSON: 831-1.0088593.json
JSON-LD: 831-1.0088593-ld.json
RDF/XML (Pretty): 831-1.0088593-rdf.xml
RDF/JSON: 831-1.0088593-rdf.json
Turtle: 831-1.0088593-turtle.txt
N-Triples: 831-1.0088593-rdf-ntriples.txt
Original Record: 831-1.0088593-source.json
Full Text
831-1.0088593-fulltext.txt
Citation
831-1.0088593.ris

Full Text

ECOLOGY OF THE MICROBIOTIC CRUST OF THE ANTELOPE-BRUSH {PURSHIA TRIDENTATA) SHRUB STEPPE OF THE SOUTH OKANAGAN, BRITISH COLUMBIA by LYNNE BERYL ATWOOD B.Sc, The University of Victoria, 1992  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Faculty of Forestry Department of Forest Sciences Centre for Applied Conservation Biology  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRItfSff COLUMBIA September 1998 © Lynne B. Atwood, 1998  In  presenting this  degree at the  thesis  in  University  of  partial  fulfilment  British Columbia,  of  the  requirements  I agree that the  for  an  advanced  Library shall make it  freely available for reference and study. I further agree that permission for extensive copying  of  department  this thesis for scholarly purposes may be  granted by the  or  understood  by  his  or  her  representatives.  It  is  that  publication of this thesis for financial gain shall not be allowed without permission.  Department of The University of British Columbia Vancouver, Canada  DE-6 (2/88)  head of my copying  or  my written  Abstract This thesis examines two aspects of the microbiotic crust in the antelope-brush [Purshia tridentata (Pursh) D.C.) shrub steppe of the South Okanagan, British Columbia: (1) relationships between the extent, distribution, and composition of the microbiotic crust, soil texture, and primary grazer (wildlife versus livestock), and; (2) effect of the microbiotic crust on soil moisture content and loss. The floral characteristics of the microbiotic crust of five antelope-brush sites are summarised in Chapter I. Differences in species richness and extent, lifeform composition, and quality of the microbiotic crust are described. Preliminary studies identified differences in the floral characteristics of the antelopebrush sites and discussed these differences in relation to soil texture and type of primary grazer. No conclusive relationships between microbiotic crust species numbers, lifeform composition, crust extent, or moss thickness were identified. It was concluded that microbiotic crust colonisation was influenced by soil surface stability and stability may be affected by physical disturbance, such as livestock grazing, or by the mechanical composition and degree of aggregation of the soil. Two of the sites examined were the focus of the soil moisture study presented in Chapter II. The study addressed three questions: 1.  Does the microbiotic crust influence soil water content and loss on the antelope-brush sites?  2.  Does the soil moisture effect differ with microbiotic crust species?  3.  Is the role of the microbiotic crust on soil water more important than influences from other ecosystem components? Soil moisture content was uniformly sampled after two rain events and two factors (shrub and  microbiotic crust) were tested in a 2 x 2 factorial design. Shrub cover was included in the experimental design to account for the possible effects of shading and hydraulic lift. Data were analysed with the general linear model and both analysis of variance (ANOVA) and repeated measures multivariate analysis of variance (repeated measures MANOVA) were used.  The study determined plots with microbiotic crust at the sample point (crusted plots) reduced infiltration initially, but soils beneath a crust retained the moisture for longer periods than areas sampled in bare soil (non-crusted plots). Shrub cover had little relevance to surface soil moisture conditions other than its effect on microbiotic crust characteristics. Soils beneath lichen crusts received more precipitation and soils beneath moss had better retention of soil water. The water retention ability of soils beneath moss increased with moss thickness. The effect of surrounding plot conditions varied, but generally the conditions at the sample point were paramount. If the sample point was covered by a microbiotic crust (crusted plots), the positive effect on soil moisture was enhanced by surrounding crust characteristics. However, if the sample point was bare (non-crusted plots) the ability of the surrounding microbiotic crust to moderate soil moisture varied. Vascular plant and litter cover moderated moisture conditions in bare soil, but were not as effective as the microbiotic crust.  T A B L E OF  C O N T E N T S  Abstract  ii  List of Tables  viii  List of Figures  x  Acknowledgement  xvii  CHAPTER I  1  I -1  INTRODUCTION AND BACKGROUND STUDIES  Project Scope  I- 2 The microbiotic crust 1-2.1 Microbiotic Crust & Soil Formation I - 2.2 Microbiotic Crust & Soil Surface Stabilisation I - 2.3 Microbiotic Crust & Nutrient Cycling I - 2.4 Microbiotic Crust as a Source of Food and Shelter I - 2.5 Microbiotic Crust & Vascular Plant Establishment I - 2.6 Microbiotic Crust & Soil Water Relations I- 3 Study area Research Site Summary  1-3.1  1 2 3 4 5 5 6 7 9 12  I-4  Sample design  15  I-5  Soil texture  18  Materials and Methods  1-5.1  I - 5.2 Results  I- 6 Lifeform Cover 1-6.1 Materials & Methods I - 6.2 Results I - 6.2.1 Microbiotic Crust Species Richness I - 6.2.2 Microbiotic Crust Cover I-7  18 20 20 22 22 23  Discussion of the characteristics of the microbiotic crust in relation to primary  grazer and soil texture 1-7.1 I - 7.2 I - 7.3 I - 7.4 I- 8  18  Extent of the Microbiotic Crust Lifeform and Species Diversity of the Microbiotic Crust Condition of the Microbiotic Crust Summary  Rational for site selection for the soil moisture study  28 28 29 30 31 32  iv  CHAPTER II  THE EFFECT OF THE MICROBIOTIC CRUST ON SOIL MOISTURE CONTENT AND SOIL MOISTURE LOSS 33  11-1.0  Introduction  33  II - 2.0  Materials and methods  35  11-2.1  Site Precipitation Monitoring  35  II - 2.2  Soil Moisture Sampling  36  Results  37  II-3.0  11-3.1 Kennedy Bench -3.1 .1 Soil -3.1 .2 Soil -3.1 .3 Plot  37  II- 3.1.3.1 Relationship between microbiotic crust in the 1 m plot surrounding the sample point and August soil moisture at Kennedy Bench 42 II- 3.1.3.2 Relationship between bare ground in the 1-m2 plot surrounding the sample point and August soil moisture at Kennedy Bench 48 Relationship between vascular plant cover in the 1-m surrounding the sample II- 3.1.3.3 point and August soil moisture at Kennedy Bench 49 II- 3.1.3.4 Relationship between litter in the 1-m surrounding the sample point and August soil moisture at Kennedy Bench 52 II- 3.1.3.5 Relationship between sand particle content in the 1-m surrounding the sample point and August soil moisture at Kennedy Bench 54 2  2  2  2  -3.1 .4 Soil -3.1 .5 Soil -3.1 .6 Plot II- 3.1.6.1 Relationship between microbiotic crust in the 1-m2 surrounding the sample point and October soil moisture at Kennedy Bench 59 II- 3.1.6.2 Relationship between bare ground in the 1-m2 surrounding the sample point and October soil moisture at Kennedy Bench 62 II- 3.1.6.3 Relationship between grass cover in the 1-m surrounding the sample point and October soil moisture at Kennedy Bench 64 II- 3.1.6.4 Relationship between sand particle content in the 1-m2 surrounding the sample point and October soil moisture at Kennedy Bench 66 2  v  11-3.2  Water Tower  70  II - 3.2.1 Soil Moisture Content in AUGUST at WATER TOWER II - 3.2.2 Soil Moisture Loss in AUGUST at WATER TOWER II - 3.2.3 Plot conditions associated with soil moisture in AUGUST at WATER TOWER  70 73 74  II - 3.2.3.1 Relationship between microbiotic crust in the 1-m2 plot surrounding the sample point and August soil moisture at Water Tower 74 II - 3.2.3.2 Relationship between bare ground in the 1-m surrounding the sample point on August soil moisture at Water Tower 77 II - 3.2.3.3 Relationship between vascular plant cover in the 1-m plot surrounding the sample point on August soil moisture at Water Tower 80 II - 3.2.3.4 Relationship between litter in the 1-m2 plot surrounding the sample point on August soil moisture at Water Tower 82 II - 3.2.3.5 Relationship between sand particle content in the 1-m2 plot surrounding the sample point on August soil moisture at Water Tower 84 2  2  II - 3.2.4 Soil Moisture Content in OCTOBER at WATER TOWER II - 3.2.5 Soil Moisture Loss in OCTOBER at WATER TOWER II - 3.2.6 Plot conditions associated with soil moisture in OCTOBER at WATER TOWER  86 88 90  - 3.2.6.1 Relationship between microbiotic crust in the 1-m2 surrounding the sample point on October soil moisture at Water Tower 90 - 3.2.6.2 Relationship between bare ground in the 1-m2 surrounding the sample point and October soil moisture at Water Tower 93 - 3.2.6.3 Relationship between vascular plant cover in the 1 -m plot surrounding the sample point and October soil moisture at Water Tower 94 - 3.2.6.4 Relationship between litter in the 1-m surrounding the sample point in October soil moisture at Water Tower 96 - 3.2.6.5 Relationship between sand particle content in the 1-m surrounding the sample point and October soil moisture at Water Tower 97 2  2  2  II-4.0  Discussion  98  11-4.1  Effect of microbiotic crust and shrub on soil moisture  98  II - 4.2  Effect of environmental conditions on soil moisture  100  II - 4.2.1 Effect of surrounding microbiotic crust on moisture conditions at sample points also covered with microbiotic crust  100  II - 4.2.2 Effect of surrounding microbiotic crust on moisture conditions at bare sample points II - 4.2.3 Effect of bare ground, vascular plants, litter and sand content on soil moisture  102  103  VI  11-4.3 CHAPTER III III -1.1  Summary MANAGEMENT IMPLICATIONS  ...107  Does the moisture holding capacity of microbiotic crusts have ecological value?  III -1.1.1 Vascular plant germination III - 1 . 1 . 2 Vascular plant survival III - 1 . 1 . 3 Nutrient cycling III - 1 . 2  104  .'  How do we retain the microbiotic crust?  References  107 107 108 109  109  111  Appendix A  Summary of vascular plant, bare ground, litter, and animal droppings data  118  Appendix B  ANOVA tables for non-vascular plant lifeform cover  125  Appendix C  Microbiotic crust species of the Antelope-brush shrub steppe  130  LIST OF  T A B L E S  C H A P T E R I  Table 1:  Location, soil, and land use characteristics of Kennedy Bench  12  Table 2:  Location, soil, and land use characteristics of Water Tower  13  Table 3:  Location, soils, and land use of Orchard Block 1 (OR1) and Orchard Block 2 (OR2)  Table 4:  Location, soils, and land use of East Osoyoos Lake  14  Table 5:  Sample design  15  Table 6:  Predefined rules used to assign effects to plots  16  Table 7:  Terminology used in the discussion results  17  Table 8:  ANOVA Model: Differences in soil texture between sites based on sand particle content at the soil surface 19  Table 9:  The lifeform composition and number of microbiotic crust species recorded at each of the five study areas 23  14  LIST OF  T A B L E S  C H A P T E R II  Table 1:  Precipitation received by Kennedy Bench and Water Tower between August 8th and  Table 2:  ANOVA: Effect of microbiotic crust and shrub on. soil moisture content at Kennedy Bench  Table 3:  Repeated measures MANOVA: Effect of microbiotic crust and shrub on the grams of water lost between day 1 & day 3 and between day 3 & day 5 in August at Kennedy Bench  Table 4:  ANOVA - Effect of Microbiotic crust and shrub on October Soil Moisture Content on day 1, day 3, and day 5 at Kennedy Bench 56  Table 5:  Repeated Measures MANOVA: Effect of crust and shrub on the grams of water loss  Table 6:  October 6,1995  one (day 1), three (day 3), and five days (day 5) after the August rain event  between day 1 & day 3 and between day 3 & day 5 in October at Kennedy Bench  ANOVA - Effect of microbiotic crust and shrub on August soil moisture content at Water  Tower on day 1, day 3, and day 5  35 38 40  58 71  Table 7:  Repeated Measures MANOVA: Effect of crust and shrub on the grams of water lost between day 1 & day 3 and between day 3 & day 5 in August at Water Tower 73  Table 8:  ANOVA - Effect of Microbiotic crust and shrub on October soil moisture content at Water Tower on day 1, day 3, and day 5 86  Table 9:  Repeated Measures MANOVA: Effect of crust and shrub on the grams of water lost between day 1 & day 3 and between day 3 & day 5 in October at Water Tower 89  ix  LIST O F  FIGURES  C H A P T E R I Figure 1: Distribution of antelope-brush study sites used in the microbiotic crust research study in the South Okanagan (adapted from Shatford, 1997)  10  Figure 2: Percent of soil texture components, sand, silt, and clay, at East Osoyoos Lake (EOL), Orchard Block 2 (OR2), Orchard Block 1(OR1), Water Tower (WT), and Kennedy Bench (KB). Percent sand means with the same letter are not significantly different (p < 0.05) 19 Figure 3: Example of a 1 -m  2  plot for a crust only plot  21  Figure 4: Mean percent cover of microbiotic crust recorded at Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). 24 Figure 5: Mean percent cover of microbiotic crust in the 1-m plots of shrub/crust and crust plots for 2  Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). Means with the same letter are not significantly different (p<0.05) 24 Figure 6: Mean percent microbiotic crust cover and proportion that is bryophyte (moss and liverwort) and lichen for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e. for total crust cover) 25 Figure 7: Mean moss thickness in crust, shrub/crust, shrub, and bare effect plots for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean +1 s.e.) 26 Figure 8: Mean percent cover of lichen in crusted and non-crusted plots and shrubbed and non-shrubbed plots at Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). Means with the same letter are not significantly different (p < 0.05) 27 Figure 9: Mean proportion of lichen cover comprised of fruticose, foliose, and crustose growth-forms for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.) 27  x  LIST O F  FIGURES  C H A P T E R II  Figure 1:  Mean grams of soil water measured on day 1, day 3 and day 5 for each of the four effects.  Measurements were taken in August at Kennedy Bench (mean + 1 s.e.)  38  Figure 2:  Mean grams of soil water measured on day 1, day 3, and day 5 for crusted and noncrusted and shrubbed and non-shrubbed plots in August at Kennedy Bench (mean + 1 s.e.) ; 39  Figure 3:  Correlation between soil water content on day 1, day 3 and day 5 and the percentage of microbiotic crust cover in the 1 m surrounding the sample point in August at Kennedy Bench 42 2  Figure 4:  Correlation between soil water content on day 1, day 3 and day 5 and the percentage of moss cover in the 1-m2 surrounding the sample point in August at Kennedy Bench 43  Figure 5:  Correlation between soil water content on day 1, day 3 and day 5 and average thickness of the moss in the 1-m surrounding the sample point in August at Kennedy Bench 43 2  Figure 6:  Correlation between day 1 soil water content of crusted plots and the percent of lichen cover in the 1-m surrounding the sample point in August at Kennedy Bench 2  Figure 7:  44  Correlation between the grams of soil water lost between day 1 & 3 and the percent of microbiotic crust, moss, and thickness of the moss in the 1-m surrounding the sample point in August at Kennedy Bench 44 2  Figure 8:  Correlation between the grams of soil water lost between days 3 & 5 and the percent of lichen cover in the 1-m surrounding the sample point in August at Kennedy Bench 2  Figure 9:  45  Correlation between the grams of water lost between days 3 & 5 and the percent of microbiotic crust, moss, and lichen cover in the 1-m surrounding the sample point in the non-crusted plots in August at Kennedy Bench 45 2  Figure 10:  The mean percent cover of lichen and bryophyte (moss and liverwort) in bare, shrub, crust and shrub/crust effect at Kennedy Bench 46  Figure 11:  The mean thickness of the moss in the bare, crust, shrub, and shrub/crust effect at  Figure 12:  Mean percent cover of microbiotic crust in the bare, shrub, crust, and shrub/crust effects at Kennedy Bench, (mean + 1 s.e.). Note: variables were arcsine square root transformed for ANOVA analysis; means with the same letter are not significantly different 47  Kennedy Bench (mean+ 1 s.e.)  46  xi  Figure 13:  Correlation between grams of soil water on day 1, day 3, and day 5 and the percent of  bare ground in the 1-m surrounding the sample point in August at Kennedy Bench. 2  Figure 14:  48  Correlation between the grams of soil water lost between day 1 and day 3 and the percent of bare ground in the 1-m surrounding the sample point in August at Kennedy Bench. ..49 2  Figure 15:  Mean percent cover of bare ground in the 1-m plots of bare, shrub, crust, and shrub/crust at Kennedy Bench (mean + 1 s.e.) Note: data were arcsine square root transformed for ANOVA analysis 49  Figure 16:  The distribution of vascular plant cover for bare, shrub, crust, and shrub/crust at Kennedy Bench and the correlation between vascular cover and soil moisture on day 1 in August (Grass: p = 0.0001, f = 34.28; Herb: p = 0.91, f = 0.01; df = 1,84; means with the same letter are not significantly different) 50  Figure 17:  The correlation between grams of soil water on day 1, day 3, and day 5 and percent of vascular plant cover in the 1-m surrounding the sample point of the non-shrubbed plots in August at Kennedy Bench 50  2  2  Figure 18:  The correlation between the grams of soil water on day 1, day 3, and day 5 and the percent of vascular plants in the 1-m surrounding the sample point of the areas of bare soil not protected by a shrub (bare effect) in August at Kennedy Bench 51 2  Figure 19:  Correlation between the grams of water lost between day 1 & day 3 and between day 3 & day 5 and the percent of grass cover in the 1-m surrounding the sample point in August at Kennedy Bench 51 2  Figure 20:  The mean cover of litter for bare, shrub, crust, and shrub/crust at Kennedy Bench (mean + 1 s.e.) and the correlation between the grams of soil water in August and the percent of litter cover in the 1-m surrounding the sample point. Note: data were arcsine square root transformed for analysis; means with the same letter are not significantly different 53 2  Figure 21:  The correlation between the grams of soil water lost between day 1 and day 3 and the percent cover of litter in the 1-m surrounding sample points covered by both a microbiotic crust and shrub (shrub/crust effect) in August at Kennedy Bench 54 2  Figure 22:  Correlation between the grams of soil water lost between day 3 and day 5 and the percent of sand particles in the 1-m surrounding sample points covered by a shrub (shrubbed plots) in August at Kennedy Bench 54 2  Figure 23:  Mean grams of soil water measured on day 1, day 3 and day 5 for bare, shrub, crust and shrub/crust effects. Measurements were taken in October at Kennedy Bench, (mean + 1 s.e.) 55  Figure 24:  Mean grams of soil water measured on day 1, day 3, and day 5 for crusted and noncrusted and shrubbed and non-shrubbed plots in October at Kennedy Bench (mean + 1 s.e.) 57 xii  Figure 25:  The correlation between the grams of soil water on day 5, after the 2-mm rainfall, and the percent cover of moss and lichen in the 1-m surrounding the sample point in October at Kennedy Bench 60 2  Figure 26:  The correlation between the grams of soil water on day 5 moisture and the percent cover of moss and lichen in the 1-m surrounding sample points covered with a microbiotic crust (crusted plots) in October at Kennedy Bench 60 2  Figure 27:  The correlation between the grams of soil water on day 5 and the percent cover of microbiotic crust and thickness of the moss in the 1-m surrounding sample points covered with a microbiotic crust (crusted plots). in October at Kennedy Bench 61 2  Figure 28:  The correlation between the grams of soil water lost between day 3 and day 5 and the percent of microbiotic crust cover and moss cover in the 1-m surrounding sample points covered by a shrub (shrubbed plots) in October at Kennedy Bench 61 2  Figure 29:  The correlation between the grams of soil water on day 1, day 3, and day 5 and the percent cover of bare ground in the 1-m surrounding the sample points in October at Kennedy Bench 2  Figure 30:  63  The correlation between grams of water lost between day 3 and day 5 and the percent of bare ground in the 1m surrounding the sample point in October at Kennedy Bench 63 2  Figure 31:  The correlation between grams of soil water on day 1, day 3 and day 5 and percent of grass cover in the 1m surrounding the sample point in October at Kennedy Bench 2  Figure 32:  64  The correlation between grams of water on day 1, day 3, and day 5 and the percent grass cover in the 1m surrounding the sample points not covered by a shrub (non-shrubbed plots) in October at Kennedy Bench 65 2  Figure 33:  The correlation between grams of soil water on day 1, day 3, and day 5 and the percent of grass cover in the 1m surrounding sample points of bare soil not covered by a shrub (bare effect) in October at Kennedy Bench 65 2  Figure 34:  Correlation between grams of soil water measured on day 1 and the percent of sand particles in the 1 m surrounding sample points covered with a microbiotic crust (crusted plots) in October at Kennedy Bench 67 2  Figure 35: Figure 36:  The correlation between grams of soil water and the percent of sand particle content on  day 1 in crust effect and on day 3 in shrub effect in October at Kennedy Bench  67  Correlation between grams of soil water lost between day 3 and day 5 and the percent of sand particles in the 1m surrounding sample points of bare soil that were covered by a shrub (shrub effect) in October at Kennedy Bench 68 2  xiii  Figure 37:  Mean grams of soil water measured on day 1, day 3 and day 5 for crusted and noncrusted and shrubbed and non-shrubbed plots. Measurements were taken in August at Water Tower, (mean + 1 s.e.)  72  Figure 38:  Mean grams of soil water measured on day 1, day 3, and day 5 for bare, shrub, crust, and shrub/crust effects in August at Water Tower, (mean + 1 s.e.) 72  Figure 39:  Correlation between the grams of soil water on day 1, day 3 and day 5 and the percent of microbiotic crust in the 1-m surrounding the sample point in August at Water Tower 75 2  Figure 40:  Correlation between the grams of soil water on day 1, day 3 and day 5 and the percent of moss in the 1 -m surrounding the sample point in August at Water Tower 75 2  Figure 41:  Correlation between the grams of soil water on day 1, day 3 and day 5 and the average thickness of the moss in the 1 -m surrounding the sample point in August at Water Tower. 75 2  Figure 42:  The correlation between soil moisture on day 1, day 3 and day 5 and the percent of lichen cover in the 1-m surrounding the sample point in August at Water Tower 76 2  Figure 43:  The correlation between the grams of water lost between day 1 & day 3 and the percent cover of microbiotic crust, moss, and thickness of the moss in the 1-m surrounding the sample point in August at Water Tower 77 2  Figure 44:  Correlation between grams of water lost between day 3 and day 5 and percent cover of microbiotic crust and moss and thickness of the moss in the 1-m surrounding the sample point in August at Water Tower 77 2  Figure 45:  The correlation between the grams of water lost between day 1 and day 3 and between day 3 and day 5 and the percent of bare ground in the 1-m surrounding the sample point in August at Water Tower 78 2  Figure 46:  The correlation between grams of soil water on day 1, day 3 and day 5 and the percent of bare ground in the 1-m surrounding the sample point in August at Water Tower 79 2  Figure 47: Figure 48:  Mean percent cover of bare ground in the 1 -m surrounding the sample point of bare, 2  shrub, crust, and shrub/crust effect at Water Tower, (mean + 1 s.e.)  80  The correlation between grams of soil water on day 1 and the percent cover of grass and herbs in the 1-m surrounding the sample point in August at Water Tower 81 2  Figure 49:  The correlation between the grams of water lost between day 1 and day 3 and the percent of grass cover in all plots and the percent of herbs in plots without a shrub (non-shrubbed) in August at Water Tower 81  xiv  Figure 50:  The correlations between grams of water lost between day 1 and day 3 and between day 3 and day 5 and the percent cover of litter in the 1-m surrounding the sample point in August at Water Tower 83 2  Figure 51:  The correlation between grams of soifwater on day 1, day 3, and day 5 and the percent cover of litter in the 1 -m surrounding the sample point in August at Water Tower 2  Figure 52:  The correlation between grams of soil water and percent of sand in the soil of the 1-m  83  2  plot surrounding the sample point in August at Water Tower  84  Figure 53:  The correlation between grams of soil water on day 5 and the percent of sand in the 1 -m plots of the non-crusted areas in August at Water Tower 84  Figure 54:  The correlation between the grams of water lost between day 1 and day 3 in all plots and in the crust effect and the percent of sand in the soil of the 1-m surrounding the sample point in August at Water Tower 85  2  2  Figure 55:  Mean grams of soil water measured on day 1, day 3 and day 5 for crusted and noncrusted and shrubbed and non-shrubbed plots. Measurements were taken in October at Water Tower (mean + 1 s.e.) 87  Figure 56:  Mean grams of soil water measured on day 1, day 3, and day 5 for each of the four  Figure 57:  effects (mean + 1 s.e.). October measurements at Water Tower  88  The correlation between grams of soil water on day 5 and the percent of total microbiotic crust and percent of moss in the 1-m surrounding the sample point in October at Water Tower 91 2  Figure 58:  The correlation between grams of soil water on day 5 and the percent of total microbiotic crust, moss and thickness of the moss in the 1-m surrounding sample points covered only by a microbiotic crust (crust effect) in October at Water Tower 91 2  Figure 59:  The correlation between grams of soil water on day 5 and the percent cover of lichen in the 1-m surrounding sample points covered only by microbiotic crust (crust effect) in October at Water Tower 92 2  Figure 60:  The correlation between grams of water lost between day 3 and day 5 and the percent cover of moss and microbiotic crust in the 1-m surrounding the sample point in October at Water Tower 92 2  Figure 61:  The correlation between grams of soil water on day 5 and the percent of bare soil in the 1m surrounding the sample point in October at Water Tower 93 2  Figure 62:  Average grams of soil water on day 5 in plots with an average of less than 30% (n = 54) and more than 30% (n = 34) bare ground in the 1-m surrounding the sample point in October at Water Tower, (mean + 1 s.e.) 93 2  xv  Figure 63:  The correlation between grams of soil water on day 5 and the percent cover of grass and herbs in the 1-m surrounding the sample point in October at Water Tower 94 2  Figure 64:  The correlation between the grams of water lost between day 1 and day 3 and the percent cover of grass in the 1 -m surrounding sample points covered by both a shrub and microbiotic crust (shrub/crust effect) and the correlation between the grams of water lost between day 3 and day 5 and the percent cover of vascular plants (grass and herbs) in the 1-m plots of the shrub/crust effect In October at Water Tower 95 2  2  Figure 65:  The correlation between the grams of water lost between day 3 and day 5 and the percent cover of litter in the 1-m surrounding the sample point in October at Water Tower 96 2  Figure 66:  The correlation between grams of soil water on day 5 and the percent cover of litter in all plots and in the plots where samples were taken from bare soil (bare effect) in October at Water Tower 97  Figure 67:  The correlation between grams of soil water on day 5 and the percent of sand in all plots and in the plots where samples were not protected by a shrub (non-shrubbed) in October at Water Tower 98  Figure 68:  Vascular and non-vascular plant lifeform categories used for data collection  Figure 69:  The average height of the shrubs in the shrubbed plots of East Osoyoos Lake (EOL), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and Orchard Block 2 (OR2) 120  Figure 70:  Mean percent grass cover for the shrub/crust, shrub, bare and crust effects at East Osoyoos Lake (EOL), Water Tower (WT), Orchard Block 1 (OR1), and Orchard Block 2 (OR2), and Kennedy Bench (KB), (mean + 1 s.e.) 120  Figure 71:  Mean percent cover of litter in the shrubbed and non-shrubbed plots at Orchard Block 1 (OR1), East Osoyoos Lake (EOL), Water Tower (WT), Orchard Block 2 (OR2) and Kennedy Bench (KB) (mean + 1 s.e.) 121  Figure 72:  Mean percent cover of bare ground in the shrubbed and non-shrubbed plots at Kennedy Bench (KB), Orchard Block 2 (ORBL2), Orchard Block 1 (ORBL1), Water Tower (WT), and East Osoyoos Lake (EOL)(mean + 1 s.e.) 122  119  xvi  Acknowledgment  Dr. Alastair McLean introduced me to the semiarid shrub-steppe in the South Okanagan. His love of the area and his willingness to share his knowledge was very much appreciated and will always be remembered. Dr. Pam Krannitz gave me the opportunity to learn more about this habitat by including me in the South Okanagan Grassland Conservation Project and for that I am very grateful. I would also like to thank Dr. Krannitz for arranging funding for my research from Environment Canada, the Habitat Conservation Trust Fund, and the Endangered Species Recovery Fund. I am also indebted to Dr. Fred Bunnell for recommending me for the UBC Graduate Fellowship and vanDusen scholarships, and for supporting my application for the GREAT scholarship. The scholarships provided continual encouragement and I am very thankful for their support. I would like to thank the South Okanagan property owners who allowed me free access to their land, specifically the Osoyoos First Nations Band, Blake and George Kennedy, and Sherri and Gary Kline. In addition, BC Ministry of Environment Penticton granted me the use of their lab facilities for which I am very grateful. I am also very appreciative of the help I received from field assistants, Remco Tikkemeijer and Risa Handler. Dr. Terry Mcintosh donated his time and taught me many of the mosses of the South Okanagan, and I am very thankful for that learning experience. I am also indebted to Dr. Teuvo Ahti, Dr. Roger Rosentreter, Dr. John Thomson and Trevor Goward for sharing their interest in mycology and identifying my lichen samples. Statisticians, Dr. Val LeMay, Dr. Peter Marshall, and Dr. Antal Kozak provided statistical help which was also very much appreciated. I would like to thank my committee members for their unique contribution to my years at UBC. Dr. Pam Krannitz's continual encouragement, support, and friendship will always be remembered. Dr. Fred Bunnell dispelled my mistrust of science and encouraged me to pursue applied research and for that I am grateful. I would also like to thank Dr. Sue Glenn for helping me formulate my thoughts and narrow my focus when I was beginning this process, and for always expressing an interest in my work. Dr. Michael Pitt could always put my research in context, which was very much appreciated. I am sincerely grateful for the help, encouragement, and friendship of the graduate students, faculty and staff of the Centre for Applied Conservation Biology. As the heart of the Centre, Jackie Johnson deserves special recognition. Her friendly voice and ever helpful attitude will always be remembered. A special thank you also goes to Jeff Shatford and Marilyn Fuchs whose wisdom, support, and friendship I cherish. I would also like to acknowledge my family and friends, who listened, laughed, consoled, and through it all never once doubted I would complete this journey. The support received from Wayne Biggs and Mark Walmsley of Westland Resource Group deserves special recognition. They taught me how to apply what I have learned at UBC. Although left to the end, Lyle Gawalko's contribution to this project is closest to my heart. His meticulous help in the field was very much appreciated and his eternal positive attitude and vision inspiring. I will always treasure his friendship.  CHAPTER I  INTRODUCTION A N D B A C K G R O U N D STUDIES  1-1  PROJECT SCOPE A collaborative study of the ecology of the Antelope-brush {Purshia tridentata (Pursh) DC.) shrub  steppe ecosystem in the South Okanagan of British Columbia was undertaken in 1994. Ten researchers were involved in the South Okanagan Grassland Conservation Project and they examined several components of the Antelope-brush habitat. The goal of the Grassland Conservation Project was to "provide information to land managers that will result in the enhancement of degraded lands for the well-being of  rare w//cf//fe (Krannitz 1994)". This thesis documents research focused on the ecology of the terricolous, non-vascular plant community. Chapter I has three objectives: (1) to provide a description of the study area; (2) examine the relationship between livestock and wildlife grazing and the non-vascular plant communities of the study sites, and; (3) provide a rationale for the sites selected for further study in Chapter II. Chapter II documents a soil moisture study. Water relations on two antelope-brush sites were examined to determine: (1) if the microbiotic crust plays a role in soil moisture retention and loss; (2) if soil moisture effects differ with microbiotic crust species, and; (3) if the role of the microbiotic crust on soil water is more important than influences from other ecosystem components. This chapter includes an introduction to the microbiotic crust and its reported ecological functions in Section 2, the study area and sample design are presented in Sections 3 and 4, respectively. Background studies to evaluate relationships among the microbiotic crust, soil texture, and the vascular plant community are addressed in Sections 5 and 6 and Section 7 discusses the study area characteristics in relation to primary grazer.  1  THE MICROBIOTIC CRUST  1-2  The Antelope-brush steppe habitat of the South Okanagan contains an element that is rarely noticed or studied. The element is the assemblage of non-vascular plants that cap the soil surface. It is variously called the microbiotic (Metting 1991, St.Clair and Johansen 1993), cryptobiotic (Harper and Pendleton 1993, Belnap 1993 and 1994), cryptogamic (Kleiner and Harper 1972), microfloral (Loope and Gifford 1972), or microphytic crust (West 1990). Mosses and lichens are the most visible components of the microbiotic crust, but liverworts, algae, fungi, and bacteria are also constituents. Microbiotic crusts are not limited to antelope-brush habitats. They are found in mesic, semiarid, and arid regions throughout the world, but they are limited to soils within these regions that are not frequently flooded, tilled, burned, or heavily used by animals or vehicles (West 1990). Crust species are also limited in areas with excessively stony or sandy soils (West 1990). Microbiotic crusts in the South Okanagan antelope-brush ecosystem are commonly found beneath shrubs or at the base of perennial bunchgrasses (personal observation). Crust species will also colonise open areas between vascular plants on sites that are not heavily trampled by animals, vehicles or people (West 1990). There is an abundance of scientific information on the components of the microbiotic crust. However, the synthesis and application of this information to the ecology of semiarid ecosystems is a fairly recent development. Harper and Marble (1988) reviewed studies concerning the microbiotic crust, soil moisture evaporation, and vascular plant establishment in rangelands and found non-conclusive results although previous research linked crust species to reduced soil erosion, improved water infiltration, and increased availability of soil nutrients. West (1990) provided a comprehensive summary of research conducted on crust taxa, the ecology of arid and semiarid ecosystems, and the effects of natural and human disturbance regimes on both the ecosystem and their microbiotic crust component. West emphasised the disparity in research 2  results, particularly those examining ecological functions of the microbiotic crust, and concluded that the variation was due to a general lack of research, unrealistic sampling procedures, and a lack of taxonomic knowledge, but also to the patchiness of crust species in both time and space (West 1990: 207). Metting (1991) also reviewed previous studies of the role of the microbiotic crust in soil stabilisation, nitrogen fixation, and seedling establishment. His review concluded effects of the crust on soil stabilisation and fertility were a benefit to rangelands, but he doubted that promoting the establishment and maintenance of microbiotic crusts was ecologically or economically feasible. In an accompanying paper, Knutsen and Metting (1991) discussed the importance of re-establishing microbiotic crusts to "abused rangelands"because of the substantial evidence supporting the crusts positive ecological role. Separate works on the role of microbiotic crusts in rangelands emphasise the value of the microbiotic crust to semiarid ecosystems and acknowledge our limited understanding of this component (Belnap 1994, Eldridge and Greene 1994). The authors urge land managers to minimise activities that destroy the microbiotic crust so research on its ecological functioning can continue. The microbiotic crust has been linked to six ecological functions: (1) soil formation; (2) soil surface stabilisation; (3) nutrient cycling; (4) source of food and shelter; (5) vascular plant establishment, and; (6) soil water relations.  1-2.1  MICROBIOTIC CRUST & SOIL FORMATION  Microbiotic crust species accelerate physical and chemical weathering and accumulate organic matter. Physical weathering is aided by the expansion and contraction of crustose lichens as water content in the plants vary, and by penetration of the rhizines of foliose lichen and moss species (Longton 1992). The expansion, contraction, and penetration actions introduce water that induces frost shattering of  3  parent materials. Lichens also promote chemical weathering by leaching acids and compounds that act as metal chelating agents (Lawrey 1986, Kendrick 1992). Both mosses and lichens accumulate organic matter. The taxa trap wind-blown organic particles and the dead vegetative material adds surface organic matter (Miicher et. al. 1988, Longton 1992). In addition, the primary productivity of the microscopic components, cyanobacteria and algae contribute directly to soil organic matter (Beymer and Klopatek 1991). I - 2.2  MICROBIOTIC CRUST & SOIL SURFACE STABILISATION  Both the visible and the microscopic components of the microbiotic crust play a role in reducing wind and water erosion. Larger soil aggregates, lower wind speeds, and slower water flows that result from the presence of a microbiotic crust work collectively to reduce sediment loss. Algae, bacteria, and fungi stabilise the soil surface by aggregating soil particles (Schulten 1985, Greene and Tongway 1989). The microscopic species surround soil particles with filaments that secrete a gelatinous cementing substance and create larger soil aggregates (Shields and Durrell 1964, Belnap and Gardner 1993). Visible crust components, lichens, mosses, and liverworts, have diverse growth forms, which create a surface roughness or microrelief that dissipates wind and slows water flow (Kleiner and Harper 1977, Harper and St. Clair 1985, Miicher ef. al. 1988). Eldridge (1993) concluded that in areas with at least 50% cryptogamic cover splash erosion was significantly lower. However, the splash erosion study was conducted in a laboratory.  4  I - 2.3 MICROBIOTIC CRUST & NUTRIENT CYCLING  Microscopic crust species and lichens with blue-green phycobionts fix atmospheric nitrogen and contribute many other essential elements to arid and semiarid ecosystems (Nash 1996). The fixed nitrogen is available to neighbouring vascular plants and is often the single largest source of nitrogen in rangelands and deserts (Shields 1957, Belnap and Harper 1995, Evans and Ehlringer 1993). The nitrogen fixers are primary colonisers of disturbed areas and fix nitrogen throughout the year, although the rate fixed varies depending on levels of light, water, and temperature (Nash 1996). Estimates for crust nitrogen fixation in the Great Basin (Utah) are between 10 and 100 kilograms per hectare per year (Metting 1991).  Levels  of other elements, such as carbon, phosphorus, potassium, iron, calcium, magnesium and manganese are also attributed to the presence of a good microbiotic crust cover (Kleiner and Harper 1977, Anderson et. al. 1982a). Increased nutrient levels are attributed to: the crusts ability to trap nutrient rich soil particles; the chelating action of the microscopic sheath material; leaching directly from the species vegetative structures; and the higher rate of nutrient uptake of darker coloured and thus warmer microbiotic crust species (Lawrey 1986, Kendrick 1992, Longton 1992). I - 2.4 MICROBIOTIC CRUST AS A SOURCE OF FOOD AND SHELTER  The microbiotic crust provides food and shelter for both vertebrates and invertebrates. Although it is not a major or even preferred component of many diets, vertebrates such as caribou, deer, elk, pronghorn antelope, wild and domestic sheep, marmot, squirrels and voles reportedly consume crust species (Seaward 1988, Longton 1992). Amphibians and reptiles do not eat microbiotic crust species, but many frogs, salamanders, lizards, geckos, and chameleons have developed an effective camouflage that harmonises with lichen covered habitats, thus using areas covered with a microbiotic crust for safe shelter  5  (Seaward 1988:119). In addition, birds frequently use both bryophyte and lichen species as camouflage materials for their nests and may be important microbiotic crust dispersal agents (Seaward 1988). The crust provides both food and shelter for many invertebrates, including protozoa, isopods, mites, molluscs, insects, annelids, arachnids, springtails, and nematodes (Frankland 1974, Loria and Herrnstadt 1980, Seaward 1988, Lawrey 1989, Davidson et. al. 1990, Longton 1992). There is speculation that invertebrate consumption and subsequent mineralisation of crust species is required to release microbiotic fixed nitrogen for use by higher plants (West 1990). I - 2 . 5 MICROBIOTIC CRUST & VASCULAR PLANT ESTABLISHMENT  Studies concerning the effect of microbiotic crusts on vascular plant germination, establishment, and growth have variable conclusions. Microbiotic crust species are linked to germination, growth, and establishment of vascular plants, but the crust has also been identified as competition for limited ecosystem resources (Harper and Belnap unpublished). The crust has been associated with increased (St. Clair et. al. 1984, Eldridge and Greene 1994) and decreased (Mcllvanie 1942, Fenner 1985, Schlatterer and Tisdale 1969, Keizeref. al. 1985) germination. Recent field studies conducted in Australia attributed increased vascular plant germination to the presence of a microbiotic crust (Eldridge and Greene 1994). However, it is quite likely effects of the microbiotic crust on vascular plant germination and establishment will depend on both the type of seed tested and the species composition of the crust. There is general agreement that seedling establishment is enhanced by the presence of a microbiotic crust (Schlatterer and Tisdale 1969, Harper and St. Clair 1985, Keizeref. al. 1985).  6  I - 2.6  MICROBIOTIC CRUST & SOIL WATER RELATIONS  The microbiotic crust reportedly influences surface storage of water, soil moisture evaporation rate, and percolation, infiltration, and retention of soil moisture. However, most studies linking the microbiotic crust and soil water focused on changes in infiltration rate. There was general agreement the microbiotic crust enhances water infiltration (Booth 1941, Fletcher and Martin 1948, Loope and Gifford 1972, Blackburn 1975, Brotherson and Rushforth 1983, Eldridge and Greene 1994), but there are discrepancies depending on the species found in the crust (Brotherson and Rushforth 1983) and other site conditions, such as soil texture (Eldridge and Green 1994). The microrelief resulting from the diversity of species growth forms creates additional storage sites for precipitation (Eldridge and Greene, 1994). Surface species, especially bryophytes, also act as mulch protecting the moist soil surface from desiccating winds and intense solar radiation (Booth 1941). The microscopic sheaths formed by the algae, bacteria, and fungi also have two roles: their aggregation properties lead to enhanced water infiltration and percolation; and, their ability to absorb water increases the moisture holding capacity of the soil (Shields and Durrell 1964; Belnap and Gardner 1993; Belnap 1994). Evaporation rates reportedly vary with species composition. Light coloured lichen species reflect light and reduce the evaporation rate, while dark coloured lichens absorb heat and increase soil surface temperature and the evaporation rate (Thomson 1967, Daubenmire 1968, Friedmann and Galun 1974, Grime 1979). It is the growth form rather than colour of bryophytes that influences the evaporation rate. Dense, compact colonies of bryophyte shoots reduce air movement and subsequently evaporation from their surface (Longton 1988). Chapter II of this thesis focuses on soil water relations of the microbiotic crust because water is the most limiting factor in arid region to vascular plant germination and growth. The Antelope-brush habitat of 7  the south Okanagan is endangered and its continuance may depend on preserving ecosystem components that advance the recruitment of native plant species. I was interested in studying the effect of the microbiotic crust on water conservation because of the possible benefit for native plant germination in these endangered areas.  8  1-3  STUDY AREA Four study sites were chosen. All sites were located in the Antelope-brush {Purshia tridentata  (Pursh) D.C.) steppe of the Very dry, Hot Bunchgrass biogeoclimatic zone (BGxhl) in the South Okanagan Basin Ecosection and Southern Interior Ecoprovince of British Columbia. The study sites were situated between Osoyoos and Vaseux Lakes on the east side of the Okanagan valley at an elevational range of 350 to 475 metres above sea level (Figure 1). The semi-arid South Okanagan Basin Ecosection is located in the rain shadow of the Coast Mountains and experiences hot, dry summers and cool, cloudy winters. The valley records the hottest summer temperatures in the province with mean monthly maximum temperatures averaging 28.1°C in July, and 26.7°C in August (Chilton 1988). January exhibits the coldest average temperature of - 3.6°C (Chilton 1988). The study area receives an average of 310 mm of precipitation per year. Major precipitation events occur in early summer (June, 34.8 mm) and mid-winter (December, 32.3 mm and January, 30.5 mm) (Chilton 1988). Droughts are intense during the summer months, punctuated by only the occasional thunderstorm that hovers over valley ridges. Antelope-brush ecosystems cover extensive areas of the USA, but in Canada the habitat is restricted to British Columbia, specifically the southern Okanagan valley and the southern Rocky Mountain Trench (Schluter et. al. 1994). South Okanagan Antelope-brush occurs in small pockets as far north as Kelowna, however the majority of this steppe system is found on the low elevation, sandy and gravelly soils south of Penticton. This geographic area is also the second most populated region of British Columbia and urban development, industry, and agriculture are having a devastating impact on the integrity of the Antelope-brush system. Today, Antelope-brush habitat is considered one of the four most endangered ecosystems in Canada (Schluter ef. al. 1994).  9  Figure 1:  Distribution of antelope-brush study sites used in the microbiotic crust research study in the South Okanagan (adapted from Shatford, 1997)  SOUTH OKANAGAN VALLEY  Vaseux Lake  A  49°15' N.  Mclntyre Bluff  Kennedy Bench (KB)  Water Tower (WT)  Oliver —  Study Sites Okanagan/ River  • Orchard Site (OR)  Osoyoos \^  ^East Osoyoos Lake (EOL) •  Lake Osoyoos  Canada  U.SA.  49°0'N. ^ approx.  5 km  10  Pristine antelope-brush environments contain the large, dark coloured antelope-brush shrubs intermingled with widely spaced clumps of perennial bunchgrasses and a variety of broad-leaf herbs. A well-developed and diverse microbiotic crust community links the herbs, bunchgrasses, and shrubs. The perennial bunchgrasses associated with this plant community are bluebunch wheatgrass (Elymus spicata [Pursh.] Gould.), needle and thread grass {Stipa comata Trin. & Rupr.), red-three awn (Aristida longiseta Steud.) and sand dropseed (Sporobolus crytandrus [Ion] A. Gray). Rabbitbrush (Chrysothamnus nauseosus [Pall.] Britt. in Britt. & Brown) and Great Basin sage {Artemisia tridentata Nutt.) shrubs are also common, minor components of this system, as well as the occasional ponderosa pine (Pinus ponderosa Dougl. ex P. &C. Lawson). Antelope-brush habitat provides food and cover for a diverse community of mammals, birds, reptiles, and invertebrates. 22% of the Province's red and blue listed vertebrates are associated with the antelope-brush system (Schluter ef. al. 1994). Scudder (1991) identified 152 rare and endangered invertebrates that rely on the habitats of the Southern Okanagan Basin Ecosection, many of which require antelope-brush habitat. In addition, 57% of the Province's red and blue listed vascular plant species are found in the natural habitats of the Okanagan valley, although the number endemic to antelope-brush systems is unknown (BC Conservation Data Centre 1997). Since the last half of the nineteenth century livestock grazing has occurred over extensive areas of antelope-brush steppe. Livestock use, in conjunction with habitat loss and fragmentation from urban and industrial development, has introduced new, exotic plant species into the system. Diffuse knapweed {Centaurea diffusa Lam.) and cheatgrass (Bromus tectorum L.) have become major components of antelope-brush areas.  11  I - 3.1  RESEARCH SITE SUMMARY  Kennedy Bench  Kennedy Bench was the most northern study site and is located on a level bench at the base of a rocky slope. The area is on the east side of the Okanagan Valley, adjacent to the Vaseux Creek Bighorn Sheep Reserve. The bench is the upper section of a fluvial fan deposit and contained sandy loam textured soils. The primary grazers at Kennedy Bench were Mule deer and California bighorn sheep. Livestock grazing, other than the occasional stray cow, has not been recorded at Kennedy Bench in the last century.  Table 1:  Location, soil, and land use characteristics of Kennedy Bench  LOCATION WEST LONGITUDE  .NORTH LATITUDE  SLOPE  119° 30' 25"  49° 15'41"  0.0- 2.0%  SOILS* SOIL TYPE  60% Stemwinder 40% Similkameen  CLASSIFICATION  PARENT MATERIAL  Rego Dark Brown Rego Brown  Fluvial Fan Deposit Fluvial Fan Deposit  STONINESS  1  0-3 0-1  DRAINAGE  well to rapid well  SOIL TEXTURE  Gravelly loam, very gravelly sandy loam Silt loam, loam  LAND USE  Little to no livestock use in the past century (Krannitz 1994). Mule deer and California bighorn sheep currently graze the site. Deer, sheep and Nuttall's Cottontail droppings were found. The frequency (% of plots surveyed, n = 55) of the droppings were deer/sheep 23.6%, Nuttall's Cottontail 5.5%, and horse 1.8% (Krannitz 1994). * Wittneben (1986) Stoniness class: 0= less than 0.01% of surface occupied by stones; 1 = 0.01% to 0.1% surface stones; 2 = 0.1% to 3% surface stones; 3 = 3% to 15% surface stones; 4 = 15% to 50% surface stones; and, 5 = > 50% surface stones (Wittneben 1986). 1  12  Water Tower  Water Tower is situated to the south of Kennedy Bench in the Inkameep First Nations reserve. The study area is on a level to very gently sloped bench on the east side of the Okanagan Valley. The surface soils at Water Tower were deposited by wind and are loamy sand in texture. Beneath the eolian soils are layers of older soils that were deposited by meltwaters of the last receding glacier. The site is surrounded by natural shrub-steppe habitat and is grazed year round by feral horses.  Table 2:  Location, soil, and land use characteristics of Water Tower  LOCATION WEST LONGITUDE  NORTH LATITUDE  SLOPE  119 ° 31"  49° 10' 26"  0.0-3°  SOILS* SOIL TYPE  Haynes  CLASSIFICATION  Orthic Brown  PARENT MATERIAL  Eolian veneer over fluvioglacial deposits  DRAINAGE  STONINESS  1  0  rapid  SOIL TEXTURE  Loamy sand or sand  LAND USE  Water Tower site is within the Inkaneep First Nations Reserve and is grazed year round by feral horses. There may also be some use by domestic cattle). Deer and/or sheep droppings are evident on the site, but they are very infrequent; horse droppings 32.5%, deer/sheep droppings 2.5% (n = 40) (Krannitz 1994). 'Wittneben (1986) Stoniness class: 0 = less than 0.01% of surface occupied by stones; 1 = 0.01% to 0.1% surface stones; 2 = 0.1% to 3% surface stones; 3 = 3% to 15% surface stones; 4 = 15% to 50% surface stones; and, 5 = > 50% surface stones (Wittneben 1986). 1  Orchard Site  Orchard site was also located on the Inkameep First Nations reserve to the south of Water Tower. Orchard had the most northern aspect of all sites included in the study. For analysis the area was divided into two sites, Orchard Block 1 and Orchard Block 2 based on differences in slope and soil texture. Orchard Block 1 was level to very gently sloping, while Orchard Block 2 was moderately sloped. Based on soil texture data collected during this study, Orchard Block 1 contained 55% more silts than Orchard Block 2. Horses travel across the area, but did not appear to graze either site.  13  Table 3:  Location, soils, and land use of Orchard Block 1 (OR1) and Orchard Block 2 (OR2) LOCATION  WEST LONGITUDE  NORTH LATITUDE  SLOPE  119° 32' 40"  49° 9'11'  0.0-15%  SOILS* SOIL TYPE  CLASSIFICATION  PARENT MATERIAL  STONINESS  DRAINAGE >  1  UNCLASSIFIED  SOIL TEXTURE @  OR1: sand, OR2: sand  LAND USE  Cattle do not currently graze orchard site. However, there is evidence of past grazing on the level areas within the site. In addition, the site is used as a travel corridor by horses crossing to a nearby water source. Frequency of horse droppings was 7.5% on the site. No other droppings were recorded (n=40) (Krannitz 1994). Wittneben (1986) @ the average soil texture for each block based on data collected during this study Stoniness class: 0 = less than 0.01% of surface occupied by stones; 1 = 0.01% to 0.1% surface stones; 2 = 0.1% to 3% surface stones; 3 = 3% to 15% surface stones; 4 = 15% to 50% surface stones; and, 5 = > 50% surface stones (Wittneben 1986). 1  East Osoyoos Lake East Osoyoos Lake was located on the east side of Osoyoos Lake on the Osoyoos First Nations reserve. The study area is located on a level bench at the base of a rocky slope. East Osoyoos Lake contained the sandiest soils of the study sites. The soils were deposited by meltwaters of the last receding glacier. The area was surrounded by natural shrub steppe vegetation and was grazed year round by both cattle and horses.  Table 4:  Location, soils, and land use of East Osoyoos Lake  LOCATION WEST LONGITUDE  NORTH LATITUDE  SLOPE  119° 28' 45"  49° 4' 26  0.0-2.0%  SOILS* SOIL TYPE  Osoyoos  CLASSIFICATION  Orthic Brown  PARENT MATERIAL  Fluvioglacial deposits  STONINESS  1  0  DRAINAGE  rapid  SOIL TEXTURE  Loamy sand, sandy loam  LAND USE  East Osoyoos Lake site is grazed, year round (but not every year), by both cattle and horses. Frequency of horse droppings on site were 12.5% and cattle droppings 2.5%. No wildlife droppings were recorded (n=40) (Krannitz 1994). 'Wittneben (1986) Stoniness class: 0 = less than 0.01% of surface occupied by stones; 1 = 0.01% to 0.1% surface stones; 2 = 0.1% to 3% surface stones; 3 = 3% to 15% surface stones; 4 = 15% to 50% surface stones; and, 5 = > 50% surface stones (Wittneben 1986). • 1  14  SAMPLE DESIGN  1-4  Effects of two factors, shrub and microbiotic crust, were sampled in a 2 x 2 factorial design. There were two levels for each factor, with microbiotic crust or without microbiotic crust and with shrub or without shrub. Combinations of the factors and levels resulted in four effects : crust only, shrub only, crust and shrub, and a control, no crust or shrub (bare) (Table 5). Table 5:  Sample design  Factor 1 Factor 2 Treatment  Microbiotic Crust (n=44) No Shrub (n=44) Shrub (n=44)  No Microbiotic Crust (n=44) Shrub (n=44) No Shrub (n=44)  Crust/Shrub  Crust  Shrub  n=22  n=22  n=22  Bare (no crust or shrub)  n=22  In July 1995, 88 1-m plots were established on each of the four study sites. The sites were 2  uniformly sampled based on a standard squared distance between plots. The sites were covered by a grid and one 1-m x 1-m quadrat was located within each grid-square. The preferred choice for the placement of the quadrat was in the centre of the grid (see general plot selection rule, Table 6). The effects were assigned to the quadrats systematically based on a set of predefined rules (B to E, Table 6). Effects were assigned to the plots in the following order: shrub/crust; shrub; crust, and; bare. The number of plots per effect and the layout design were determined in consultation with Dr. Peter Marshall, Statistician with the Department of Forestry, UBC. There were 44 plots for each factor or 22 plots for each effect. The terminology used in the discussion of experiment results is presented in Table 7.  15  Table 6:  Predefined rules used to assign effects to plots  A. General Plot Selection Rule i. The first choice for plot location was the centre of each grid square ii. If the centre point does not meet the effect plot rule (rules B. to E.) the first suitable location  will be chosen by moving clockwise, in 1-m concentric circles. The radius of the largest circle will not be greater than the sites computed distance between plots. iii. If no suitable plot location is found within the circle the effect will be missed. Missed plot areas will be re-visited once plot layout is complete. The interstitial areas surrounding the largest plot circle will be examined for a suitable location. If a suitable location cannot be found, the effect will be missed. iv. Neighbouring treatments will be separated by a minimum distance of 2 m; the four corners of the plot will be marked with wire pin-flags and a metal tag, marked with the plot and treatment number, will be attached to the pin-flag in the south-east corner of the plot.  B. i. ii.  Shrub/Crust Plot Rule  There must be a contiguous microbiotic crust area >_ 15 cm (2.25% of the plot area). Sampling area must be located within the dripline of a shrub. 2  C. Shrub only Plot Rule i. There must be a contiguous non-crusted (no microbiotic crust) area > 15 cm (2.25% of the 2  ii. iii. D. i.  plot area). The area without a microbiotic crust cover must not be covered with litter. The sampling area must be located within the dripline of a shrub.  Crust only Plot Rule  There must be a contiguous microbiotic crust area > 15 cm (2.25% of the plot area). No woody shrubs rooted within the 1-m plot. Sample area must not be shaded by a bunchgrass. 2  ii. iii. E. No Shrub and No Crust (Bare) Plot Rule i. There must be a contiguous non-crusted (no microbiotic crust) area > 15 cm (2.25% of the 2  2  ii. iii.  plot area). No woody shrubs rooted within the 1m plot. Sample area must not be shaded by a bunchgrass. 2  16  Table 7:  Terminology used in the discussion results  Factor Terminology  Associated Effects  CRUSTED PLOTS  Crust Shrub/Crust Shrub Bare  NON-CRUSTED PLOTS SHRUBBED PLOTS  Shrub Shrub/Crust Crust Bare  NON-SHRUBBED PLOTS  The plots on Water Tower and Orchard were separated into two and three blocks, respectively because the topography of the sites prevented a contiguous layout. Water Tower, Block 1 contained 68 plots and Block 2, 20 plots. Orchard Block 1 contained 36 plots, Block 2, 20 plots, and Block 3, 32 plots. As a result of soil texture data collected during this study, the blocks at Water Tower were combined for analysis because there was no difference in sand particle content (T = 0.582, df = 1, 86, p = 0.6) (general linear model, SAS 1988). There was a significant difference in soil texture between blocks at Orchard site (F = 6.54, df = 2, p = 0.0023), however two of the three blocks did not differ significantly (critical t = 3.374, df = 2, 85, p < 0.05) (Tukey multiple range test, SAS 1988). Therefore, I separated Orchard site into two study areas for analysis, Orchard Block 1 and Orchard Block 2. In 1994 a vascular plant floristic study conducted by Dr. Pam Krannitz established forty 0.1m  2  plots at Water Tower, Osoyoos Lake, and Orchard; Kennedy Bench held 55 plots. I collected microbiotic crust species data from each of the 0.1-m plots. This information, in addition to species data from the 12  m plots, was used to determine microbiotic crust species richness. 2  17  SOIL TEXTURE  1-5  I - 5.1  MATERIALS AND METHODS  The soil texture was recorded for each of the 352 plots (4 sites x 88 plots) because soil texture affects water relations. Percent sand (percentage of particles between 2 and 0.05 mm), percent silt (0.05 0.002 mm), and percent clay (< 0.002 mm) were determined by Norwest Labs (Langley, BC) using the hydrometer method (McKeaque, 1978). The information for each plot was also summarised to soil texture class, as specified by Wittneben (1986). The samples used for the textural analysis were a combination of three separate plot samples collected in October 1995. The samples were collected for the soil moisture experiment (see methods, Chapter 2), oven dried, combined by plot number, and then analysed for soil texture. Mean percent of sand, silt and clay content were determined for Kennedy Bench, East Osoyoos Lake, Water Tower, and for each block at Orchard. ANOVA was used to determine if there was a significant difference in soil texture between sites. Because of the correlation between soil texture components, only sand particle content was used in the analyses.  I - 5.2  RESULTS  Clay was a minor soil component at all sites, accounting for only 1.98 + 0.13 % to 2.55 + 0.13% (+ 1 s.e.) of the soil texture (Figure 2).  18  Figure 2:  Percent of soil texture components, sand, silt, and clay, at East Osoyoos Lake (EOL), Orchard Block 2 (OR2), Orchard Block 1(OR1), Water Tower (WT), and Kennedy Bench (KB). Percent sand means with the same letter are not significantly different (p < 0.05). • %Sand 100% .  B%Silt  1.98  2.40  3.34  4.50  94.68  93.10  80%  H  %Clay  2.36  2.55  2.41  10.08  19.58  28.13  60% 40%  87.59  77.88 69.45  20%  a  0%  a I  EOL  0R2  b l  0R1  d  C I  WT  I  I  KB  Site  There was a difference in sand content between all sites (Table 8) except East Osoyoos Lake (94.68 + 0.19% s.e. sand) and Orchard Block 2 (93.10 +_0.75%) (p < 0.05; critical t=3.878, df=347) (Tukey Multiple Range Test, SAS 1988). On average, East Osoyoos Lake and Orchard Block 2 contained 7% more sand particles than Orchard Block 1,17% more sand than Water Tower and 26% more sand than Kennedy Bench.  Table 8:  ANOVA Model: Differences in soil texture between sites based on sand particle content at the soil surface SOURCE OF VARIATION  Model Error Corrected Total  , DF  4 347 351  SUM OF SQUARES  22133.397 11289.257 33422.654  MEAN SQUARE  5533.350 32.534  F  170.08  PR>F  0.0001  R-Square =0.662227 SOURCE OF VARIATION  DF  SITE  4  TYPE I  SUM OF SQUARES  22133.397  MEAN SQUARE  5533.350  F  170.08  PR>F  0.0001  19  East Osoyoos Lake, Orchard Block 2 and Orchard Block 1 held the coarsest soils with a sand surface texture (> 8 5 % sand particle content), followed by Water Tower with a loamy-sand surface texture (70 - 8 5 % sand), and Kennedy Bench with a sandy-loam texture (50 - 70% sand).  1-6  LIFEFORM COVER  I - 6.1  MATERIALS & METHODS  Non-vascular and vascular plant data, as well as percent cover of bare ground, litter, and animal droppings were collected from each plot in August 1995. Vascular and non-vascular data collected included: •  Moss, lichen, and liverwort species of the microbiotic crust;  •  percent cover microbiotic crust;  •  percent cover of the visible microbiotic crust lifeforms (moss, fruticose lichen, foliose lichen, and crustose lichen);  •  moss thickness (height in millimetres);  •  percent cover of shrub, broad-leaf herbs, grasses, and legumes;  •  shrub species and height (metres) in the shrubbed plots. Species information pertaining to the microbiotic crust, specifically species numbers, percent cover  of microbiotic crust, percent cover of the visible microbiotic crust lifeforms and moss thickness is presented here. The vascular plant, bare ground, litter, and animal droppings data are summarised in Appendix A. Species richness data were collected from the four study sites. Presence of lichen, moss, and liverwort species was recorded from the 40 (KB = 55) 0.1-m plots and the 88 1-m plots. A cumulative 2  2  total of lichen, moss, and liverwort species was determined based on the 1994 0.1-m plot and the 1995 12  20  m plot data and taxonomy was verified by experts. Dr. Terry Mcintosh verified moss species and lichen 2  species were confirmed by Trevor Goward, Dr. Teuvo Ahti, Dr. John Thomson, and Dr. Roger Rosentreter. Each plot was 1-m , but the plot area that met the effect (crust, shrub/crust, shrub, or bare soil) 2  requirements see (Table 6) could be as small as 15 cm . Therefore, the 1-m plot contained factors from 2  2  other treatments, such as microbiotic crust cover in a bare treatment plot or bare ground in a crust treatment plot, which necessitated the collection of data for all variables from each plot (Figure 3). All references to percent cover are based on data collected only from the 1-m plots. 2  Figure 3:  Example of a 1-m plot for a crust only plot 2  Percent cover of non-vascular plant components was collected using a continuous cover scale. The 1 -m plots were divided into 25,20-cm cells. Each cell accounted for 4 % of the plot. Percent cover 2  2  of each microbiotic crust variable in the plot was determined by adding the proportion of cover for each cell. The division of lichens into three forms was based on Goward et. al. (1994). Since lichen lifeforms were mutually exclusive, total lichen cover for each plot was determined by adding all lifeforms. Unless noted, percent total microbiotic crust cover was determined by adding moss and lichen.  Moss thickness  was a measure of the height of the moss within the plot. Tortula ruralis Hedw. was the most common moss  21  at all sites and was used for all moss thickness measurements. The moss height was recorded in millimetres and was the average of three haphazard plot measurements. Each measurement was taken by removing one T. ruralis gametophyte, and recording its length from the rhizoid collar to the apical tip. The microbiotic crust data were analysed using ANOVA (SAS 1988) to determine if there were significant differences in the percent cover of the crust characteristic between the factors, crust and shrub. All references to significant differences between factors or sites refer to p < 0.05. ANOVA tables for each site and crust variable included in this section are presented in Appendix B.  I - 6.2  RESULTS  I - 6.2.1  Microbiotic Crust Species Richness  A total of 43 species (26 lichens, 15 mosses, and two liverworts) were identified in the antelopebrush microbiotic crust. Two lichen and two moss species remain unidentified. When the average number of species per square meter surveyed was compared Orchard Block 2 was the richest in density of microbiotic crust species (Table 9).  22  Table 9:  The lifeform composition and number of microbiotic crust species recorded at each of the five study areas. SITE  Area surveyed (m )  #MOSS  # LICHEN  #  TOTAL #  # SPECIES/  SPECIES  SPECIES  LIVERWORT  SPECIES  SAMPLING AREA  SPECIES  2  EAST OSOYOOS LAKE WATER TOWER KENNEDY BENCH ORCHARD BLOCK 1 ORCHARD BLOCK 2  92 92 94 59 33  6 7 11 11 11  5 11  16 10 9  0 0 2 1 0  (MEAN # / M ) 2  11  0.12  18  0.20  29  0.31  22  0.37  20  0.61  More moss than lichen species were identified at three of the five sites. Lichen species outnumbered moss species at Water Tower and Kennedy Bench. A microbiotic crust species list for the antelope-brush habitat of the south Okanagan is included in Appendix C.  I - 6.2.2  Microbiotic Crust Cover  The average cover of microbiotic crust differed between sites (F = 3.03, df = 4, 347, p= 0.02, Table 1 Appendix B). In general, East Osoyoos Lake and Orchard Block 1 contained higher average crust cover than Water Tower, Kennedy Bench and Orchard Block 2 (Figure 4). In general, shrubs did not significantly affect microbiotic crust cover. East Osoyoos Lake was the only site that contained significantly more microbiotic crust in areas with both a shrub and crust cover (shrub/crust) as compared to areas with only a crust (crust) (t = -2.69, df = 29, p = 0.01) (Figure 5).  23  Figure 4:  Mean percent cover of microbiotic crust recorded at Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). % Microbiotic Crust Cover  45 40 35 30 25 20 15 10 5 0  Figure 5:  OR2  WT  KB  OR1  32  88  88  56  EOL  Mean percent cover of microbiotic crust in the 1 -m plots of shrub/crust and crust plots for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). Means with the same letter are not significantly different (p < 0.05). 2  • Crust m Shrub/ crust 70 Q) > O O  60  •ss 50 a  o o 40 o !a o 30 . o E 20  a a  a  T-  a  rlf  a  i  a?  c ro 10 CO  E  0 -  OR2  WT  KB  0R1  EOL  There were differences in lifeform composition of microbiotic crusts between sites (Figure 6). Moss cover was the largest component of the microbiotic crust at all sites, although at Kennedy Bench the microbiotic crust was composed almost equally of moss (55%) and lichens (45%).  24  Figure 6:  Mean percent microbiotic crust cover and proportion that is bryophyte (moss and liverwort) and lichen for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e. for total crust cover) • %Bryophyte n % L i c h e n  50 40  Kill  a  > 30 o 20 10  n =  OR2  WT  KB  OR1  32  88  88  56  EOL 88  There was significantly more moss cover in crusted plots at all sites (df = 1; EOL: F = 72.58, p = 0.0001, Table 2 Appendix B; WT: F = 24.57, p = 0.0001, Table 3 Appendix B; KB: F = 24.79, p = 0.0001, Table 4 Appendix B; OR1: F = 10.81, p = 0.002, Table 5 Appendix B; OR2: F = 45.58, p = 0.0001 Table 6 Appendix B). Moss cover was also significantly higher in shrubbed plots at East Osoyoos Lake (F = 7.4, df = 1, p = 0.008, Table 2 Appendix B), Kennedy Bench (F = 36.04, df = 1, p = 0.0001, Table 4 Appendix B), and Orchard Block 1 (F = 9.21, df = 1, p = 0.004, Table 5 Appendix B). Moss thickness or height varied greatly between the sites. On average, Kennedy Bench contained the tallest moss (11.66 + 0.71 mm, n = 88) and Water Tower the shortest (5.15 + 0.27 mm, n = 88). The moss at East Osoyoos Lake averaged 8.02 + 0.51 mm (n = 88), Orchard Block 1 averaged 7.82 + 0.62 mm 9 n = 56), and the mean at Orchard Block 2 was 9.41 + 0.65 mm (n = 32). Shrub/crust treatment contained the tallest moss at all sites except Water Tower (Figure 7).  25  Figure 7:  Mean moss thickness in crust, shrub/crust, shrub, and bare effect plots for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.)  • Bare SI Shrub • Crust m Shrub/ crust 2 0  I  1 1  Orchard Block 1 was the only site with significantly more lichen in crusted versus non-crusted plots (F = 4.31, df = 1, p = 0.04, Table 10 Appendix B) (Figure 8). The majority of Orchard Block 1 lichen was concentrated in areas with only a microbiotic crust (crust), which contained 11.89 + 3.85% lichen compared to 2.82 + 1.05% in areas without crust or shrub (bare). The effect of shrubs on lichen colonisation varied. There was no difference in lichen cover between shrubbed and non-shrubbed plots at Orchard Block 1 (F = 0.63, df = 1, p = 0.43,Table 10 Appendix B), Orchard Block 2 (F = 1.24, df = 1, p = 0.3, Table 11 Appendix B), or East Osoyoos Lake (F = 0.39, df = 1, p = 0.53, Table 7 Appendix B) (Figure 8). However, a shrub cover did have a positive effect on lichen establishment at Water Tower (F 14.85, df = 1, p = 0.0002, Table 8 Appendix B) and shrubbed plots held 26% more lichen than non-shrubbed areas. The lichen in the shrubbed plots at Water Tower was concentrated in the shrub effect (mean 5.29 + 1.43% versus 3.38 + 1.01% in shrub/crust effect). The situation was reversed at Kennedy Bench where there was significantly more lichen in the non-shrubbed plots (F = 44.55, df = 1, p = 0.0001; Tukey multiple range test (SAS 1990): p < 0.05, critical t = 2.81, df = 2,84). The majority of lichen was held in crust effect (26.79 + 3.84%) versus bare effect (14.05 + 2.59%).  26  Figure 8:  Mean percent cover of lichen in crusted and non-crusted plots and shrubbed and nonshrubbed plots at Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). Means with the same letter are not significantly different (p < 0.05)  • Crusted 25  • Shrubbed  n Non-crusted  • Non-shrubbed T b  £ 20  > o u  s  1 5  10  c  (0 c  1  5 OR2  WT  KB  a  OR1  a  EOL  OR2  WT  KB  OR1  EOL  The lichen cover consisted of three growth-forms; fruticose, foliose, and crustose. Fruticose lichens dominated the lichen cover at all sites followed by foliose lichens and small quantities of crustose lichens, which were recorded at Kennedy Bench, Water Tower and East Osoyoos Lake only (Table 9 and Figure 9). Figure 9:  Mean proportion of lichen cover comprised of fruticose, foliose, and crustose growth-forms for Orchard Block 2 (OR2), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and East Osoyoos Lake (EOL) (mean + 1 s.e.). ES Fruticose • Foliose • C r u s t o s e 14 12 +  £  c  cu u  10  8  2 6 CD > o o  o I Nfci OR2  WT  KB  OR1  EOL  27  The dominant fruticose lichens at all sites were Cladonia species and at Kennedy Bench, Orchard Block 1 and Orchard Block 2 the Cladonias were much more developed than those seen at Water Tower and East Osoyoos Lake. Lichen expansion requires time and development of secondary squamules occurs with age (Rogers 1990). Therefore, Cladonias at Water Tower and East Osoyoos Lake may be recent immigrants.  1-7  DISCUSSION OF THE CHARACTERISTICS OF THE MICROBIOTIC CRUST IN RELATION TO PRIMARY GRAZER AND SOIL TEXTURE This study was not designed to test the effect of primary grazer on the microbiotic crust community.  The microbiotic crust, soil texture, and primary grazer information was examined to determine if these site characteristics would influence the soil water relations examined in Chapter II.  I -7.1  EXTENT OF THE MICROBIOTIC CRUST  The relationship between type of primary grazer and extent of microbiotic crust cover that was seen in this study did not agree with previous research. There was no relationship between livestock grazing and crust cover though other research have concluded both extent and species diversity of the crust are reduced with increased grazing pressure (Rogers and Lange 1971, Nash et. al. 1977, Anderson et. al. 1982b, Brotherson et. al. 1983, Andrew and Lange 1986, Joh.ansen.and St. Clair 1986, Harper and Marble 1988, Marble and Harper 1989). East Osoyoos Lake held the highest average cover of microbiotic crust and was grazed by livestock. In addition, crust cover at Water Tower, which was also grazed by livestock, was similar to crust cover at Orchard Block 2 and Kennedy Bench, two wildlife grazed sites.  28  A clear relationship between extent of crust and soil texture was not evident in this study. Kennedy Bench and Water Tower held finer textured soils and did contain similar quantities of microbiotic crust, but East Osoyoos Lake and Orchard Block 2 were the sandiest sites and differed in crust cover. There have been studies on the colonisation of the microbiotic crust that have concluded crust extent and species richness are related to soil texture, specifically that finer textured soils possess a higher crust cover and a more diverse, lichen dominated community (Rogers 1972, Anderson et. al. 1982a, Dunne 1989, Metting 1991). However, mosses were the dominate crust lifeform at all sites and were found to increase with sand particle content (r = 0.46, p = 0.0001, n = 352) (Leach 1931, Richardson 1981, Shulten 1985, Mcintosh 1986). The colonisation ability and resilience of the dominant moss species may confound the relationship between the extent of the microbiotic crust, primary grazer, and soil texture. Tortilla ruralis Hedw. accounted for the majority of moss cover on the antelope-brush sites and based on a taxonomic study of the bryophytes in the southern interior of BC, Mcintosh (1986) classified T. ruralis an aggressive, weedy taxa of sandy soils, which was also more resistant to disturbance.  I - 7.2  LIFEFORM AND SPECIES DIVERSITY OF THE MICROBIOTIC CRUST  Mosses were the dominant lifeform at all antelope-brush sites surveyed. However, Kennedy Bench, the wildlife grazed, finer textured site held an almost equal quantity of moss and lichen. Both finer textured sites (Kennedy Bench and Water Tower) possessed more lichen than moss species, but lichen was concentrated in the non-shrubbed areas at Kennedy Bench, while occupying the shrubbed plots at Water Tower. The difference in lichen distribution would suggest they are more likely to colonise finer textured soils, which is similar to the findings of Rogers (1972) and Metting (1991). However, from this study it is also suggested lichens will only colonise areas that are also protected from soil disturbance. Areas not protected by a shrub are more susceptible to disturbance from livestock trampling or vehicles.  29  No clear relationship between species richness and soil texture was evident, but primary grazer did appear to influence species numbers. A higher number of species were identified at Kennedy Bench, which held the finest soil texture, but on average, the very sandy Orchard Block 2 contained more species per square meter. Both Kennedy Bench and Orchard Block 2 were grazed by wildlife. A comparison of livestock grazed sites found the finer textured Water Tower site held seven more species than East Osoyoos Lake. Both Water Tower and East Osoyoos Lake contained fewer microbiotic crust species overall and fewer species per meter than those grazed by wildlife. The findings suggest that microbiotic crust species richness may be higher on finer textured soils, possibly because more moss and lichen species are able to colonise the more stable soils. However, this study also suggests species numbers may decrease when the primary grazers are livestock. Tortula ruralis was the most common moss species colonising all sites, but Ceratodon  purpureus  Hedw., Bryum argenteum Hedw., Brachythesium albicans [Hedw.] B.S.G. and Polytrichum piliferum Hedw. were also prominent species. Leach (1931), Richardson (1981), Shulten (1985), and Mcintosh (1986) identified these species as coloniser mosses of sandy soils. Richardson (1981) and Shulten (1985) also concluded that some lichens, particularly Cladonia and Peltigera species were able to colonise generally unstable soils once mosses had established. Interestingly, Cladonia species were the most common lichen on all antelope-brush sites and accounted for 90% of the lichen cover at Kennedy Bench.  I - 7.3  CONDITION OF THE MICROBIOTIC CRUST  The relationship between moss thickness and primary grazer was not consistent. At all sites mosses were thicker in areas less susceptible to disturbance from grazing (shrubbed plots), but an examination of moss thickness in areas susceptible to disturbance (non-shrubbed plots) showed mixed results. The mosses in the non-shrubbed plots at Kennedy Bench and Orchard Block 2 (wildlife grazed) were taller than  30  the mosses in non-shrubbed areas at the livestock grazed sites (Water Tower and East Osoyoos Lake). However, Orchard Block 1 was also grazed by wildlife, but the moss thickness in the non-shrubbed areas closely resembled the moss thickness found at East Osoyoos Lake. East Osoyoos Lake and Orchard Block 1 were very sandy sites and the similarity in moss thickness, despite differences in primary grazer, may indicate that soil texture also influences moss growth. Unprotected areas on sandy sites would likely dry faster and experience more soil movement from wind and rain, conditions that may slow moss establishment and growth.  I - 7.4 SUMMARY  Relationships between the characteristics of the microbiotic crust and primary grazer and soil texture were not clear in this study and warrant further research. Crust extent did not differ by primary grazer or soil texture. However, there was a tendency for crust lifeform to vary with soil texture and lifeform also appeared to change, from lichen to moss, in areas susceptible to livestock grazing. Species richness also appeared to be lower in livestock grazing areas, but there was no relationship between species numbers and soil texture. If mosses were colonising sites disturbed by livestock grazing, you would expect the moss to be thin and distributed in small patches. However, this was not evident. Based on the findings in this study, it appears a determinant for microbiotic crust colonisation and growth is soil surface stability and stability may be affected by either physical disturbance, such as livestock grazing, or by the mechanical composition and degree of aggregation of the soil.  31  RATIONAL FOR SITE SELECTION FOR THE SOIL MOISTURE STUDY  1-8  Kennedy Bench and Water Tower were chosen for the soil moisture study discussed in Chapter II because they shared similar topographic and physical characteristics and received the same amount of precipitation during the August rain event. The sites were located on level benches with similar aspects and at comparable elevations [KB: 450 m a.s.l; WT: 475 m a.s.l.(Scudder unpublished GPS data)]. Both sites held finer textured soils and contained similar quantities of microbiotic crust. However, the sites did differ in primary grazer, Kennedy Bench was grazed primarily by wildlife while Water Tower was grazed by livestock. Differences in microbiotic crust lifeform and species composition, average moss thickness, distribution of lichen, shrub height, grass cover, litter distribution, and amount of bare ground were also identified between the two sites (see Appendix A for vascular plant information). These differences may have contributed to the significant difference between sites identified during a preliminary analysis of soil water content one day after the 8-mm rainfall (F = 41.05, df = 3, p = 0.0001; Tukey multiple range test: critical t = 3.669, df = 3,174, p < 0.05). The four sites (Kennedy Bench, Water Tower, Orchard Block 1, and Orchard Block 2) that received 8-mm of rainfall during the August rain event were included in the preliminary analysis. Because the preliminary analysis identified a significant difference between the two sites, despite similarities in site characteristics and precipitation, Kennedy Bench and Water Tower were analysed separately in Chapter II.  32  C H A P T E R II  The Effect of the Microbiotic crust o n Soil Moisture Content and Soil Moisture L o s s  II -1.0  INTRODUCTION  The microbiotic crust on soil surfaces is assumed to enhance surface storage, percolation, infiltration, and retention of soil moisture and alters the soil moisture evaporation rate (Booth 1941, Fletcher and Martin 1948, Shields and Durrell, 1964, Thomson 1967, Daubenmire 1968, Loope and Gifford 1972, Freidmann and Galun 1974, Blackburn 1975, Grime 1979, Brotherson and Rushforth 1983, Belnap and Gardner 1993, Belnap, 1994, Eldridge and Greene 1994) (see Chapter I, Section I-2). However, the effect of the microbiotic crust on soil moisture content is not well researched. Some studies have examined the water retention capability of specific crust species, but research measuring differences in soil moisture content between areas with a microbiotic crust and areas without a crust are virtually non-existent. Blue-green algal sheaths are known to absorb 12 to 13 times their volume in water, in six minutes (Shields and Durrell 1964) and Microcoleus vaginatus [Vauch.] Gom. can absorb eight times its weight in water almost instantaneously (Belnap and Gardner 1993). In the 1920's it was speculated that algal sheaths regulated soil moisture in sandy soils (Fritsch 1922) and a second soil moisture study concluded more moisture was held in the upper 2.5-cm of soil with prominent microbiotic crusts than in adjacent soils with no crusts (Booth 1941). Booth was studying an algal crust in oak savannah in Oklahoma and felt his soil moisture test was inadequate. Gold and Bliss (1995) examined plant community development in the high Arctic and found that soil water was consistently higher in areas with a microbiotic crust.- The Arctic crust was comprised of both lichens and moss. However, moisture retention properties were of little biological relevance because the Arctic subsoil was always saturated.  33  The South Okanagan experiences the driest climate in British Columbia and although the vascular plant communities are adapted to droughty conditions, a lack of soil water can limit both seed germination and plant growth (Weaver 1968, Walter and Stadelmann 1974, Fenner 1985, Bewley and Black 1994). Ecosystem components that reduce soil moisture loss may provide a vital link in the regeneration process of arid habitats. Previous studies have suggested that microbiotic crusts conserve soil water, but information on the role of the species in the crust and the extent of the water holding ability have not been studied. Two of the five sites discussed in Chapter One, Water Tower and Kennedy Bench, were chosen for the soil moisture study. Water Tower and Kennedy Bench were selected because they shared similar topographical and physical characteristics and received the same amount of precipitation during the August rain event (see Chapter I, Section I-8). Specific questions this research project addresses are: 1.  Does the microbiotic crust affect soil water content and loss on the antelope-brush sites?  2.  Does the soil moisture effect differ with microbiotic crust species?  3.  Is the role of the microbiotic crust on soil water more important than influences from other ecosystem components, specifically vascular plants, litter, bare ground, and sand particle content?  34  MATERIALS AND METHODS  11-2.0 II - 2.1  SITE PRECIPITATION MONITORING  One rain gauge was placed on each study site and used to record the total amount of precipitation received during the sampling season. Rain gauges were attached to three-foot wooden stakes secured in an open area. Mineral oil was added to the gauge to prevent evaporation of precipitation between data collection periods. The gauges were installed August 8th and precipitation amounts were collected on August 16, August 26, September 2, October 6, October 8, and October 10,1995 (Table 1).  Table 1:  Precipitation received by Kennedy Bench and Water Tower between August 8th and October 6,1995.  AMOUNT OF PRECIPITATION RECEIVED SAMPLING  DATE  PERIOD  PRECIPITATION RECEIVED  llllfl H^^^^H KENNEDY BENCH  WATER TOWER  August 8 to 16  August 10 & 12  1 mm  3 mm  August 17 to 18  August 17  8 mm  8 mm  August 19 to 26  no precipitation  0  0  August 27 to September 2  Sept. 2  4 mm  4 mm  September 2 to October 6  unknown  20 mm  14 mm  October 8  October 7  1 mm  0  October 10  October 9  2 mm  2 mm  35  II - 2.2  SOIL MOISTURE SAMPLING  Soil moisture samples were collected in August and October 1995. Sampling occurred over a five day period following the single rain event on August 17 and again during the sporadic showers between October 7 and October 11. Soil samples were collected every second day; August sampling occurred on August 18 (day 1), August 20 (day 3) and August 22 (day 5). October sampling dates were October 6 (day 1), October 8 (day 3), and October 10 (day 5). Soil moisture content was determined using the gravemetric procedure (Carter 1993). Soil samples were collected from 0 to 2-cm in depth. For crusted effects (crust and shrub/crust), the microbiotic crust was raised and soil directly beneath was collected. The lifeform of the crust species above the sample point was recorded and the aspect of the sample was noted for shrub effects. Soil samples were collected from a new area within the plot each sampling day, but both microbiotic crust lifeform and aspect were kept constant throughout. In a few cases, lifeform and/or aspect could not be retained and the new species lifeform and/or aspect were recorded on the sampling day they changed. The soil samples were put in tin soil cans and placed in an insulated cooler immediately following collection. The wet weight of all samples was recorded at the end of the field day to the nearest onehundredth (0.01) of a gram. Soil samples were oven-dried at 105°C for 48 hours and the dry weight was recorded to the nearest 0.01 g after a 2-hour cooling period. Samples were kept in the closed drying oven during cooling. The following equation was used to determine water loss in grams per 100 grams of dry soil (g H2O/100 g dw). Samples were standardised for 100 grams of dry soil to permit comparison.  (Wet weight - dry weight) x 100 Dry weight  36  RESULTS  11-3.0  II  - 3.1  II - 3.1.1  KENNEDY BENCH  Soil Moisture Content in AUGUST at KENNEDY BENCH  One day after the 8-mm rainfall, the average soil moisture content at Kennedy Bench was 6.25 + 0.2 g (s.e.) H2O/100 g dw. On average the site lost 67% of the day 1 soil moisture by day 3 and retained 2 3 % of the day 1 soil water five days after the rain event. On day 1 there was a significant interaction between the factors crust and shrub that accounted for 4 % of the total variation in soil moisture content (Table 2). The presence of both a shrub and microbiotic crust (shrub/crust effect) had a much greater effect on soil water than either shrub or crust alone and resulted in less water absorption and slower drying. Day one shrub/crust contained 2 9 % less soil moisture than the effect with only crust and 4 1 % less moisture than the effect with just shrub (Figure 1). On days three and five soil moisture content in shrub/crust effect resembled the crust only effect (T= - 0.6533, df = 42, p = 0.52), suggesting changes to the microbiotic crust resulting from a shrub cover affected rain water absorption and slowed drying. An examination of the microbiotic crust showed the shrub/crust effect contained more microbiotic crust and thicker mosses than the crust effect (Chapter I, Figure 5 and Figure 7). The presence of microbiotic crust at the sample point (crusted plots) had a significant effect on soil moisture from day 1 through day 5 (Table 2), but the effect of the crust changed during the first 3 sampling days. On day 1, areas covered with microbiotic crust contained an average of 3 0 % less soil moisture than plots without a crust cover (non-crusted plots). However, by day 3 plots with microbiotic crust contained an average of 64% more soil water than non-crusted areas and on day 5 crusted areas held 6 8 % more moisture (Figure 2A). 37  Table 2:  n=  ANOVA: Effect of microbiotic crust and shrub on soil moisture content at Kennedy Bench one (day 1), three (day 3), and five days (day 5) after the August rain event  44  •jSjEliNG if DAY if DAY R  2  = 0453  DAY R  2  2  3  = 0.578  DAY R  1  5  = 0.536  Figure 1:  8  SOURCE glARIATION  DF  TYPE  MEAN  III  I iii  Model Microbiotic crust (Crusted) Shrub (Shrubbed) Microbiotic crust x shrub Error Model Microbiotic crust (Crusted) Shrub (Shrubbed) Microbiotic crust x shrub Error Model Microbiotic crust (Crusted) Shrub (Shrubbed) Microbiotic crust x shrub Error  3  1 1 1 84 3 1 1 1 84 3 1 1 1 84  F  PR>F  %OF  SQUARE  SUM OF  TOTAL  Si VARIATION  SQUARES  143.699 111.049 19.521 13.129 173.666 100.544 99.588 0.831 0.125 73.551 53.377 48.133 3.904 1.340 46.186  47.90 111.049 19.521 13.129 2.067 33.51 99.588 0.831 0.125 0.876 17.79 48.133 3.904 1.340 0.55  23.17 53.71 9.44 6.35  0.0001 0.0001 0.0029 0.0136  38.28 0.0001 113.74 0.0001 0.95 0.3327 0.14 0.7068 32.36 87.54 7.10 2.44  0.0001 0.0001 0.0092 0.1222  35% 6% 4%  57% 0.5 0.07  % %  48% 4% 1%  Mean grams of soil water measured on day 1, day 3 and day 5 in bare, shrub, crust, and  shrub/crust effects. Measurements were taken in August at Kennedy Bench (mean + 1 s.e.). • Day 1  H  Day 3  • Day 5  o  10 ooOl CI c  Bare  Shrub  NON-CRUSTED  Crust  Shrub/ Crust CRUSTED  The effect of shrub on soil water content varied over the five days (Table 2). On day 1 areas without a shrub (non-shrubbed) held approximately 14% more soil moisture than areas with a shrub cover (shrubbed). Non-shrubbed samples from bare soil (bare effect) contained the most soil water on day 1,  38  holding 2 0 % more moisture than non-shrubbed sample points covered with only microbiotic crust (crust effect) (Figure 1). There was no significant difference in soil moisture content between areas with or without a shrub cover on day 3, but by day 5 the shrubbed areas contained an average of 2 6 % more moisture than non-shrubbed locations (Figure 2B). The effect with both shrub and microbiotic crust (shrub/crust) accounted for the increased day 5 soil water in the shrubbed plots, holding 6 9 % more moisture than areas with only a shrub (shrub effect) (Figure 1).  Figure 2:  Mean grams of soil water measured on day 1, day 3, and day 5 for crusted and non-crusted and shrubbed and non-shrubbed plots in August at Kennedy Bench (mean + 1 s.e.). B. HCRUSTED • NON-CRUSTED  Day 1  II - 3.1.2  Day 3 Sampling Day  SHRUBBED  B  • NON-SHRUBBED  Day 1  Day 3  Day 5  Sampling Day  Soil Moisture Loss in AUGUST at KENNEDY BENCH  There was a dramatic difference in the drying rate of crusted and non-crusted, and shrubbed and non-shrubbed plots in the three days following the August rain event (Table 3 - day 1/3). However, an interaction between the factors indicated drying patterns differed with the various combinations of microbiotic crust and shrub.  39  Table 3:  Repeated measures MANOVA: Effect of microbiotic crust and shrub on the grams of water lost between day 1 & day 3 and between day 3 & day 5 in August at Kennedy Bench  n = 4' TIME  *  SOURCE OF VARIATION  DF  iiil il' llliiiili-i iiiiiiil 1 i ill 1! ill 1 •! 1 i 1111 < ! l |J i!! 1111.' .•!1!'.! ^' i' r.-jS ^ ^ l u H l M ^ f e i . l i | ; j ' ! i rfeffiti<ll I i' !•!' 11H1 jtffljt[• 1.-. i'H1 i-bh^g^^^^^g^|^HH-!:|-|||:|-|:l.l:| 1H|IIH111111111 i 11 1 i II' "| 1' 11 ' 1' " ! ]'/ i r j-J'^S J  r  1  1  DAY  DAY  1  1  111 1  1  1M  L  r  1 /3 Time Time x Microbiotic crust Time x Shrub Time x Microbiotic crust x Shrub Error 3 /5 Time Time x Microbiotic crust Time x Shrub Time x Microbiotic crust x Shrub Error  1 1 1 1 84 1 1 1 1 84  TYPE III  MEAN  PR>F  SQUARE SQUARES  1609.19 1609.19 540.07 540.07 30.09 30.09 13.73 13.73 205.02 2.44 143.16 143.16 4.54 4.54 4.90 4.90 1.83 1.83 177.71 2.12  %OF TOTAL VARIATION  •  659.31 221.28 12.33 5.62  0.0001 0.0001 0.0007 0.0200  67% 23% 1 % 0.6 %  67.67 2.15 2.31 0.87  0.0001 0.1466 0.1319 0.3547  43% 1 % 2% 0.6 %  Soil moisture was lost at a very rapid rate in sample points not covered with microbiotic crust (noncrusted: shrub and bare effects). As mentioned, the non-crusted areas contained the highest soil water content on day 1, but held the lowest soil moisture by day 3 (Figure 2). During the three-day period, areas of bare soil with a shrub cover (shrub effect) lost an average of 8 3 % of the day 1 moisture and samples taken from bare ground (bare effect) lost an average of 8 5 % soil water (Figure 1). Crust covered sample points lost approximately one-half the soil water of non-crusted areas. However, the addition of a shrub cover to microbiotic crust (shrub/crust effect) resulted in a much slower rate of soil moisture loss than plots with crust alone (crust effect) (T = 3.83, df = 42, p = 0.0004). Shrub/crust effect lost 19% of the day 1 soil moisture by day 3, while crust effect lost 4 7 % of the soil moisture during this same period (Figure 1).  40  After three days of rapid drying, there was no difference in the percentage of moisture loss in areas with or without microbiotic crust or shrub (Table 3 - day 3/5), although crusted areas contained considerably more grams of water (Figure 1). Reasons why drying patterns did not differ between the factor levels may be quite different. Samples taken from bare soil (non-crusted plots: bare and shrub effects) lost the majority of their soil water within three days of the precipitation event (Figure 1). On average, the non-crusted areas held only 1.2 + 0.09 g H2O /100 g dw on day 3 and less than 1-g H2O /100 g dw remained after five days (Figure 2). By day 3 there was very little water remaining and it is likely moisture content at the air-soil interface was near equilibrium. In contrast, areas with microbiotic crust lost only 36% of their moisture in the first three days and still retained 4 3 % of the day 1 measurement after five days. On average, over 2-g H2O /100 g dw were present in the crusted areas on day 5. The low day 3 to day 5 drying rate in crusted areas cannot be attributed to low moisture content and likely reflects the ameliorating effects of the microbiotic crust. There was no difference in drying patterns and little difference in soil water content between shrubbed and non-shrubbed plots on either day 3 or day 5. The findings demonstrate the minimal impact of a shrub cover on its own. Grouping the data by shrub averaged the differences between crusted and non-crusted effects (Figure 2). Shrub/crust effect was particularly good at retaining soil water and was the only effect to lose approximately the same amount of moisture between day 3 and day 5 (22%) as between day 1 and day 3 (19%). After five days areas with both a shrub and crust still contained over one-half (59%) of the soil water measured on day 1 (Figure 1). The moisture content study suggested the thicker mosses and more extensive crusts found beneath a shrub lowered rainwater absorption and slowed drying (page 37). The moderated moisture loss evident in shrub/crust effect provided further evidence that microbiotic crust attributes resulting from a shrub cover were an important influence on the rate of soil water loss.  41  II - 3.1.3  Plot Conditions Associated with AUGUST Soil Moisture at KENNEDY BENCH  II - 3.1.3.1  Relationship between microbiotic crust in the 1 mplot surrounding the sample point and August soil moisture at Kennedy Bench 2  The extent, lifeform composition, and morphology of the microbiotic crust in the 1-m surrounding 2  the sample point were correlated with site soil moisture over the five-days. Plots containing large amounts of microbiotic crust comprised of thick mosses were associated with decreased soil moisture on day 1, but the same conditions were correlated with increased water content on days three and five (Figure 3, Figure 4, and Figure 5).  Figure 3:  Correlation between soil water content on day 1, day 3 and day 5 and the percentage of  microbiotic crust cover in the 1 m surrounding the sample point in August at Kennedy Bench. 2  DAY 1  DAY 3  r = 0.42 p = 0.0001 n =88  r = - 0.52 p = 0.0001 n =88  100  DAY 5  50 100 MICROBIOTIC CRUST (%) (arcsine sqrt)  100  42  Figure 4:  Correlation between soil water content on day 1, day 3 and day 5 and the percentage of moss cover in the 1-m2 surrounding the sample point in August at Kennedy Bench. DAY 3  DAY 1  50 MOSS Cover (%) (arcsine sqrt)  100  Figure 5:  DAY 5  100  100  Correlation between soil water content on day 1, day 3 and day 5 and average thickness of the moss in the 1-m surrounding the sample point in August at Kennedy Bench. 2  DAY 1 o  (A  at  o .cr o to a> c  DAY 5  DAY 3  r = - 0.38 p = 0.0003 n = 87  r = 0.38 p = 0.0003 n =87  c  CD c o o La  i  0  20  40  MOSS THICKNESS (mm)  Water infiltration appeared to be higher in lichen rather than mossy areas. There was a tendency for day 1 moisture content to be slightly higher if the microbiotic crust contained large amounts of lichen (r = 0.20, p = 0.06, n = 87). The relationship was particularly noticeable in areas with microbiotic crust at the sample point (crusted plots) and plots covered with more lichen contained more moisture than plots with smaller amounts of lichen (Figure 6). Lichen cover was not related to moisture content in crusted plots on day 3 (r = 0.06, p = 0.6, n = 43) or day 5 (r = - 0.12, p = 0.33, n = 43).  43  Figure 6:  Correlation between day 1 soil water content of crusted plots and the percent of lichen cover in the 1-m surrounding the sample point in August at Kennedy Bench. 2  20  § r 2  » o  15  o o  10  C J-  § ? !_  ° o  r = 0.48 p = 0.001 n =43  5 0  g < 0  20 40 LICHEN Cove r (%) (arcsine sqrt)  60  Attributes of the microbiotic crust surrounding the sample point were also associated with moisture loss. Thick, mossy crust was related to low water loss in the three days following the rain shower (Figure 7). After day 3 the relationship changed and extensive lichen crusts lost more water than other areas (Figure 8). However, in areas that did not have a crust cover at the sample point (non-crusted plots) the extent of the crust was more important than species composition for increasing water loss between day 3 and day 5 (Figure 9).  Figure 7:  Correlation between the grams of soil water lost between day 1 & 3 and the percent of microbiotic crust, moss, and thickness of the moss in the 1-m surrounding the sample point in August at Kennedy Bench. 2  15  D) V  o  —' C  1  10  ir if  ™ -a •o .—.C  •S •a<»I = 5  c °>  5 x> S  <fe ' 0  o" oo  -  -  Oo  o oo  r = - 0.54 p = 0.0001 n =86  r = - 0.56 p = 0.000 n =86 ooQ„  '  5  0 1—.  £ -s  <5&o o :—•Q--0--  4*  r = - 0.45 p =0.0001 n =85  co o o o So."  - ©-"oo—-D' oo<  0 50 100 MICROBIOTIC CRUST Cover (%) (arcsine sqrt)  50 MOSS Cover (%) (arcsine sqrt)  100  10  20  30  40  MOSS THICKNESS mm  44  Figure 8:  Correlation between the grams of soil water lost between days 3 & 5 and the percent of lichen cover in the 1-m surrounding the sample point in August at Kennedy Bench. 2  sen ay; soi sqr  =  •o ocn o >  —  o>  £CD o X  lost  IO  0)  >>  0.22 p = 0.04  4 8  U)  .a CO) vJ u  r=  8 o °S>f°<? o_  o = 86  oo CP  8o?  0  n  oo  oo,  "TJ  CB  •a •D  an  S •o  -4 20  60  40  LICHEN Cover(%) (arcsine sqrt) Figure 9:  Correlation between the grams of water lost between days 3 & 5 and the percent of microbiotic crust, moss, and lichen cover in the 1-m surrounding the sample point in the noncrusted plots in August at Kennedy Bench. 2  = 0.33  CO  >. CO  c  0) CD  5-  O  4  3 c ^  s  °  o  o °o ° ° °o %  co  CD  Z  p  O  t o  °  o  o °  OO  ° «  O  9,<rP O ~ r = 0.45  —d"""tr  oo o 8°, o o §°° ° _o__o. o 8"| o  oo  o o O O o o o o „O O6>O  pSTJj002""  ° 20 n =44 60 40 MICROBIOTIC CRUST Cover(%) (arcsine sqrt) 0  t  r = 0.31 p = 0.04 n =44  = 0.03  = 44  10 20 30 MOSS Cover (%) (arcsine sqrt)  20  40  LICHEN Cover (%) (arcsine sqrt)  60  The extent of the microbiotic crust surrounding the sample point appeared to be the most important factor ameliorating site soil water. Crust and shrub effects contained the same amount and thickness of moss (Figure 10 and Figure 11), but crust effect held an average of 5 3 % more microbiotic crust in the one square metre surrounding the sample point (Figure 12). Although shrub effect contained more soil water on day 1 crust effect held 5 8 % more soil water by day 5 (Figure 1).  45  Figure 10:  The mean percent cover of lichen and bryophyte (moss and liverwort) in bare, shrub, crust and shrub/crust effect at Kennedy Bench. • Bryophyte cover • Lichen Cover  50 40 o  30  1  20  i r  10  Bare  Shrub  NON-CRUSTED  Figure 11:  Crust  Shrub/ Crust  CRUSTED  The mean thickness of the moss in the bare, crust, shrub, and shrub/crust effect at Kennedy Bench (mean + 1 s.e.). 25 E  20  5  15  8  10  i  o E  Bare  Shrub  NON-CRUSTED  Crust  Shrub/ Crust  CRUSTED  46  Figure 12:  Mean percent cover of microbiotic crust in the bare, shrub, crust, and shrub/crust effects at Kennedy Bench, (mean + 1 s.e.). Note: variables were arcsine square root transformed for ANOVA analysis; means with the same letter are not significantly different.  §  60,  o o 20 E 5? 10 E  F = 28.22 p = 0.0001 df = 1,84  a  Bare  b b a  Shrub  NON-CRUSTED  Crust  Shrub/ Crust  CRUSTED  The findings suggest lichen tissue doesn't absorb water as well as moss tissue. Mosses are poikilohydrous and thicker moss mats would have absorbed more water than thin moss mats (Richardson 1  1981, Longton 1988). At the time of the data collection, the mosses were quite desiccated from the dry, hot weather of early August . 2  Lichens are also poikilohydrous, but their tissues are able to absorb atmospheric moisture and they will remain hydrated under environmental conditions dry enough to desiccate mosses (Longton 1988, Nash 1996). It is possible the lichens were not as dry as the mosses and required less moisture to fully hydrate, thus allowing more precipitation to penetrate the soil surface. The morphology of the Cladonia species of lichens, which was the most common genus at Kennedy Bench, may also have contributed to the increased soil water content. Cladonias possess two thalli, one prostrate and one upright. Together they  Poikilohydric plants are plants whose water content varies with the surrounding environment (Richardson 1981). Regional records indicate the last significant rainfall at the Oliver station occurred on August 6 when 3.8mm was received. The average temperature between August 6 and 17 was 17.7°C (Environment Canada Atmospheric Environment Service). 1  2  47  create a complex surface structure with a wide array of water storage sites, small fissures that open to the soil surface and funnels that direct precipitation (the upright thallus are shaped like goblets and can act as funnels). The unique morphological characteristics of the lichen could have concentrated water in local areas resulting in an increase in soil moisture.  II - 3.1.3.2  Relationship between bare ground in the 1-m2 plot surrounding the sample point and August soil moisture at Kennedy Bench  The amount of bare ground surrounding the sample point was related to soil water content and loss at Kennedy Bench. Extensive areas of bare ground were associated with high soil water on day 1 and low soil moisture on days 3 and 5 (Figure 13). The amount of bare soil surrounding the sample point contributed to the rapid rate of moisture loss in the three days following the rain (Figure 14). The bare and shrub effects contained the highest amount of bare ground ( Figure 15) and held the most soil moisture on day 1, but retained only 8% (bare effect) and 1 1 % (shrub effect) of the moisture after five days (Figure 1). In contrast, the shrub/crust effect, which held the lowest amount of bare ground, still retained 58% of the day 1 moisture five days after the rain event (Figure 1). Figure 13:  Correlation between grams of soil water on day 1, day 3, and day 5 and the percent of bare ground in the 1-m surrounding the sample point in August at Kennedy Bench. 2  DAY 1  aj |  0  20  DAY 3  40  60  0  20 40 BAREGROUND (%)  DAY 5  60  rj  20  40  60  (arcsine sqrt)  48  Figure 14:  Correlation between the grams of soil water lost between day 1 and day 3 and the percent of bare ground in the 1-m surrounding the sample point in August at Kennedy Bench. 2  r = 0.73 p = 0.0001 n =86  10  20  30  40  50  BARE GROUND (%) (arcsine sqrt)  Figure 15:  Mean percent cover of bare ground in the 1-m plots of bare, shrub, crust, and shrub/crust at Kennedy Bench (mean + 1 s.e.) Note: data were arcsine square root transformed for ANOVA analysis. 2  12  p = 0.0001 f = 148.39 df = 1,84  10  C D > O o  Bare  Shrub  NON-CRUSTED  11-3.1.3.3  Crust  Shrub/ Crust  CRUSTED  Relationship between vascular plant cover in the 1 -m surrounding the sample point and August soil moisture at Kennedy Bench 2  High vascular plant cover was associated with increased soil moisture content on day 1 (Figure 16), but was not related to soil water on subsequent sampling days (day 3: r = 0.09, p = 0.4, n = 87; day 5: r = 0.13, p = 0.22, n = 87). Vascular plants consisted of herbs and perennial bunchgrasses and were most prevalent in areas without a shrub cover (bare and crust effect) (Figure 16). In the non-shrubbed areas large amounts of vascular plants where related to increased soil water throughout the five sampling days  49  (Figure 17). The influence was particularly noticeable in areas where samples were collected from bare soil (bare effect) (Figure 18). Plots with an extensive covering of grass also lost more soil water over the sampling period, although the association with moisture loss between day 1 and day 3 was marginal (Figure 19).  Figure 16:  The distribution of vascular plant cover for bare, shrub, crust, and shrub/crust at Kennedy Bench and the correlation between vascular cover and soil moisture on day 1 in August (Grass: p = 0.0001, f = 34.28; Herb: p = 0.91, f = 0.01; df = 1, 84; means with the same letter are not significantly different) 25  M % Grass Cover  i  • % Herb Cover  50  •a  o  40  DAY 1  20  15  a> "Hi  30  r s 10 C CS  20  3 ~ 3  10  CS  3 Bare  Shrub  Crust  NON-CRUSTED  Figure 17:  20  Shrub/ Crust  60  40  80  VASCULAR PLANT Cover (%) (arcsine sqrt)  CRUSTED  The correlation between grams of soil water on day 1, day 3, and day 5 and percent of vascular plant cover in the 1-m surrounding the sample point of the non-shrubbed plots in August at Kennedy Bench. 2  DAY 3  DAY 1 o  in  25 20  o co  2 I s «  S S. ° I O  CD  r = 0.34 p = 0.03 n =43  = 0.40 = 0.007 = 43  o o  15  DAY 5 r = 0.36 p = 0.02 n =43  o  10 5  o CO  0 20  40  60  80  20  40  60  80  20  40  60  80  VASCULAR PLANT COVER (%) (arcsine sqrt)  50  Figure 18:  The correlation between the grams of soil water on day 1, day 3, and day 5 and the percent of vascular plants in the 1-m surrounding the sample point of the areas of bare soil not protected by a shrub (bare effect) in August at Kennedy Bench. 2  DAY 1  DAY 3  DAY 5  25  o in O)  20  o S r CN <"  o o  15  in <2 10 .E ~en in  r = 0.51 p = 0.02 n =22  1  CD CB  I ?  o o -rj  r = 0.48 p = 0.02 n =22  r = 0.58 p = 0.004 n = 22  o  o C  b o  0  20  40  60  80  &.  20  40  60  80  20  40  60  80  VASCULAR PLANT Cover (%) (arcsine sqrt)  5  Figure 19:  Correlation between the grams of water lost between day 1 & day 3 and between day 3 & day 5 and the percent of grass cover in the 1-m surrounding the sample point in August at Kennedy Bench. 2  DAY 1 TO DAY 3 =  *- o >< co •S cn CD CD  X  •o cn. in o <*>„ — >, <- re  0> TJ ra T3  3 g  15 10  ,°0. °X ° o o o 0 8  8 ° °c 0  o 6>p  o o ,o °- o w o „  DAY 3 TO DAY 5 r = 0.24 p = 0.03 n = 86 15  °  OQO  o°»o<p. .o 8>»  CO  o  ° r = 0.2 = 0.06, n = 86  20 40 GRASS COVER (%) (arcsine sqrt)  60  20  40  60  Vascular plants in bare effect plots increased at the expense of bare ground (r = - 0.54, p = 0.01, n = 22). As noted, day 1 soil moisture was higher in areas with large amounts of bare ground (Figure 13) and also increased with large amounts of vascular plants (Figure 16). Both bare soil and vascular plants contained attributes that resulted in increased soil water content. It is likely soil moisture was higher in areas with large amounts of bare soil because there was minimal rainwater interception. Vascular plant 51  cover may also have increased soil water by ameliorating the microclimate and slowing evaporation and/or by providing additional rainwater storage sites. Approximately 19 hours passed between the end of the August rain shower and collection of the day 1 soil sample. Plant cover would reduce airflow and shade bare soil, and wet vegetation may have cooled the surface temperatures; all of which would slow surface evaporation during the hours before data collection. Altered microclimate conditions may also account for the higher soil moisture in bare effect plots with large amounts of vascular plants on days three and five. It is also possible the release of rainwater captured by the vegetation increased the amount of precipitation received in the local area. Bunchgrasses were the dominant plant cover at Kennedy Bench and it is possible their natural funnel shape intercepted moisture that would otherwise have fallen outside the plot (Ndawula-Senyimba ef.a/. 1971).  II - 3.1.3.4  Relationship between litter in the 1-m surrounding the sample point and August soil moisture at Kennedy Bench 2  Litter in the 1-m plots did not influence soil moisture content on day 1 (r = - 0.12, p = 0.3, n = 87) or 2  day 5 (r = 0.14, p = 0.2, n = 87), but large amounts of litter were related to increased site soil water on day 3 (Figure 20). Litter was concentrated in plots with a shrub cover (shrubbed) (Figure 20), but was particularly important for day 3 soil water content in the non-shrubbed, bare effect (r = 0.55, p = 0.008, n = 22). Litter in bare effect was closely related to other plot variables, increasing with plant cover (r = 0.52, p = 0.01, n = 22) and at the expense of bare ground (r = - 0.71, p = 0.0002, n = 22). It is likely both vegetation and litter reduced evaporation in the area surrounding the bare soil sample location providing capillary water to the dry, bare ground. 3  Capillary water is soil water that moves to reduce a moisture gradient between adjoining areas (Fitzpatrick 1986). Capillary water can move in any direction. 3  52  Litter surrounding the sample point had a minimal influence on the amount of moisture loss. In the three days following the rain there was a tendency for reduced water loss in areas with large amounts of litter (r = - 0.21, p = 0.06, n = 86). However, litter was important in the shrub/crust effect, where grams of water lost between day 1 and day 3 increased with litter cover (Figure 21). In the shrub/crust effect litter increased at the expense of extensive, thick, mossy crusts (n = 22; extent of microbiotic crust: r = - 0.72, p = 0.0002; moss: r = - 0.63, p = 0.002; moss thickness: r = - 0.59, p = 0.004). The association between litter and increased water loss in the effect that commonly contained large amounts of moss crust would suggest the water retention abilities of the microbiotic crust are superior to that of litter.  Figure 20:  The mean cover of litter for bare, shrub, crust, and shrub/crust at Kennedy Bench (mean +1 s.e.) and the correlation between the grams of soil water in August and the percent of litter cover in the 1-m surrounding the sample point. Note: data were arcsine square root transformed for analysis; means with the same letter are not significantly different 2  DAY 3  100  F = 15.32 p = 0.0002 df=1,84  o  I" cr o co  80  2 •  60  s »  r = 0.27  20  p = 0.01 16  n =87  12  o  8  40  O  20 Bare  Shrub  NON-CRUSTED  Crust  Shrub/ Crust  CRUSTED  0  0  o °oo ° cP o o o O• o»o' u  4  25  1  50 75 LITTER Cove r (%) (arcsine sqrt)  100  53  Figure 21: The correlation between the grams of soil water lost between day 1 and day 3 and the percent cover of litter in the 1-m surrounding sample points covered by both a microbiotic crust and shrub (shrub/crust effect) in August at Kennedy Bench. 2  TJ  _  _  TJ  8  r= 0.54 p = 0.01 n = 21  .38.  c °> o> o 0  o o o  o  1 3'  o c.  0 4o  CP 50  II - 3.1.3.5  60  70 80 90 LITTER Cover (%) (arcsine sqrt)  100  Relationship between sand particle content in the 1-m surrounding the sample point and August soil moisture at Kennedy Bench 2  Sand particle content was related to the amount of water that was lost in shrubbed areas between day 3 and day 5. Increased sand content correlated to an increase in water loss (Figure 22), an expected response since sandy soils are rapidly drained. However, it is not clear why the relationship between sand content and water loss was restricted to the last three sampling days and only the shrubbed areas. Sand particle content was not related to any of the other plot variables. Figure 22:  Correlation between the grams of soil water lost between day 3 and day 5 and the percent of sand particles in the 1-m surrounding sample points covered by a shrub (shrubbed plots) in August at Kennedy Bench. 2  1° re » TJ cn • c o cr <D o cu ^ in 4  88 < 0  0)  II  c Z 3 re 0 I  0  * -.0  o o  r = 0.34 p = 0.02 n =44  «n  t- re CD T J  re TJ 3 g  40  50 60 70 SAND CONTENT (%) (arcsine sqrt)  80  54  11-3.1.4  Soil Moisture Content in OCTOBER at KENNEDY BENCH  It rained the day before the October day 1 data collection, but the exact amount received at the site is unknown. Precipitation received during the sampling period was recorded and l-mm of rainfall was received on day two and 2-mm of precipitation fell on day four. Kennedy Bench averaged 4.78 + 0.15 g H2O /100 g dw on the first collection day. Although the site received one millimetre of rainfall on day two, the day 3 site average fell to 4.62 + 0.17 g H 0 /100 g dw. 2  As a result of the rain shower on day four, the day 5 site average increased to 5.18 g H2O /100 g dw. The response of the effects to the additional soil water differed. In areas where samples were collected from bare soil (bare effect), bare soil covered by a shrub (shrub effect), and beneath microbiotic crust (crust effect) average soil water content increased, but the effect with both a shrub and microbiotic crust cover (shrub/crust) continued to lose soil water during this period (Figure 23).  Figure 23:  Mean grams of soil water measured on day 1, day 3 and day 5 for bare, shrub, crust and  shrub/crust effects. Measurements were taken in October at Kennedy Bench, (mean + 1 s.e.)  • Day 1 H Day 3 • Day 5  5  •a CD CN  4  rii  Gl  ro 2  Bare  Shrub  NON-CRUSTED  Crust  Shrub/ Crust CRUSTED  A microbiotic crust cover at the sample point was the most important factor affecting soil water content on days one and three, but shrub cover at the sample point was the only important factor on day 5 55  (Table 4). Areas with a microbiotic crust (crusted plots) contained 16% more soil water than non-crusted areas on day 1 and 21% more moisture on day 3 (Figure 24A). On day 5 there was no difference in soil water content between crusted and non-crusted effects. However, there was a difference in the day 5 water content between shrubbed and non-shrubbed plots (Figure 24B). On average, areas with a shrub cover (shrubbed plots) held 14% less soil water than areas without a shrub (non-shrubbed plots). Table 4:  n= 44  SAMPLING DAT  ANOVA - Effect of Microbiotic crust and shrub on October Soil Moisture Content on day 1, day 3, and day 5 at Kennedy Bench DF  SOURCE or VARIATION  TYPF III SUM OF  MEAN SQUARE  F  PR>F  " j OF TOTAL VARIATION  SQUARES  DAY1 R  2  = 0.231  DAY  3  R = 0.089 2  DAY R  2  5  = 0.147  Model  3  Microbiotic crust Shrub Microbiotic crust x shrub  77.97 1 1 1  66.160 2.917 8.893  25.99 66.160 2.917  Error  84  260.265  8.893 3.10  Model  3  55.007  18.336  Microbiotic crust Shrub Microbiotic crust x shrub  1  54.508  Error  560.006  0.0008 6.667  Model  3  36.886  12.30  Microbiotic crust  Shrub  Microbiotic crust x shrub Error  1 1  1 84  1.674 34.811 0.401  214.586  1.674 34.811  0.401 2.55  0.0001  21.35 0.94 2.87 2.75  54.508 0.499  1 1 84  0.499 0.0008  8.39  0.0001 0.3347 0.0939  20% 0.86  %  3%  0.0478 8.18 0.07  0.00 4.81  0.0054  0.7851 0.9911  9% 0.08  %  0.0001 %  0.0038 0.66  13.63  0.16  0.4205 0.0004 0.6929  0.67  %  14% 0.16%  56  Figure 24:  Mean grams of soil water measured on day 1, day 3, and day 5 for crusted and non-crusted and shrubbed and non-shrubbed plots in October at Kennedy Bench (mean + 1 s.e.). B. n Crusted  5  • Non-Crusted  I Shrubbed  •Non-Shrubbed  T3  in cn  |  4  4  o  a  rI  w  o> 2  c  cs o>  Day  1  Day 3  Day 5  Day  1  Day 3  Day 5  Crusted areas retained their soil water between day 1 and day 3, while non-crusted plots dried. However, the crusted areas absorbed little of the 2-mm rain shower that occurred on day four, while noncrusted plots absorbed enough of the light precipitation to increase the average moisture level to that of the crusted plots. Soil beneath crusted areas absorbed less precipitation but was able to retain the moisture longer than soil in non-crusted areas. This trend was also evident during the August rain event. The importance of shrub to day 5 soil water was misleading. Non-shrubbed areas contained considerably more moisture on day 5 because of the rapid absorption that occurred in the bare effect (Figure 23).  57  II - 3.1.5  Soil Moisture Loss in OCTOBER at KENNEDY BENCH  Changes in soil moisture in response to the October showers demonstrated the influence of a microbiotic crust cover. In general, the light shower received between days one and three did not affect moisture content and all plots lost soil water during this period. However, the two-millimetre rainfall between days three and five had mixed results depending on the effect. Soil moisture increased in bare, crust, and shrub effects and decreased in shrub/crust effect. On average, crusted plots only lost 0.1 g H2O /100 g dw in the first three days of the sampling period and non-crusted plots lost slightly more, an average of 0.39 g H2O /100 g dw (Figure 24A). In the three day period, average water loss in the shrubbed and non-shrubbed plots was 0.22 g H2O /100 g dw and 0.12 g H2O /100 g dw, respectively (Figure 24B). Undoubtedly the light showers that occurred during this period slowed the rate of soil water loss in all plots.  Table 5:  Repeated Measures MANOVA: Effect of crust and shrub on the grams of water loss between day 1 & day 3 and between day 3 & day 5 in October at Kennedy Bench  SOURCE OF VARA I TO IN DAY 1 /3 Time  1  Timex Microbiotic crust  1  Time x Shrub  1  Time x Microbiotic crust x Shrub Error  DAY 3/5  DF  Time Time x Microbiotic crust Time x Shrub Time x Microbiotic crust x Shrub  Error  1  84 1 1 1 1 84  TYPE III SUM OF IIIIIRES  MEAN SQUARE  1.596  PR>F F  49  0.282  1.596 0.282  0.501  0.09  0.501  0.16  4.361  24.769  4.361 271.682 24.769 18.539 13.488  3.234  18.539  241.983  0.1 % 0.2 %  1.35  0.6947 0.2489  2%  8.60  0.0043  8%  0.0130  6%  6.44 4.68  0.183  0.06  2.881  0.6 %  0.7685  13.488  0.183  0.4843  TOTAL VARIATION!  0.0333 0.8018  5% 0.06 %  58  Between day 3 and day 5 the percentage of water lost differed for both crusted and non-crusted and shrubbed and non-shrubbed plots (Table 5, day 3-5). The two millimetre rainfall that occurred on day four resulted in an average increase in soil water of 1.45 g H2O /100 g dw in the non-crusted plots, but there was no change in soil water content in crusted areas (Figure 24A). Both shrubbed and non-shrubbed plots increased in average soil water between days three and five, however the non-shrubbed areas increased by 0.87 g H2O /100 g dw, while shrubbed increased by only 0.25 g H2O /100 g dw (Figure 24B). The shrubs in shrubbed plots likely intercepted a portion of the light rain shower.  il - 3.1.6 Plot conditions associated with OCTOBER soil moisture at KENNEDY BENCH  II - 3.1.6.1  Relationship between microbiotic crust in the 1-m2 surrounding the sample point and October soil moisture at Kennedy Bench  Crust conditions were not related to water content or loss in the first three sampling days. However, after the 2-mm rain shower, both water content and loss varied with extent and lifeform composition of the microbiotic crust. Across the site day 5 soil water content was lower in areas with large quantities of moss and was higher in areas with large quantities of lichen (Figure 25). This relationship was particularly noticeable in plots with microbiotic crust at the sample point (crusted plots) (Figure 26). The extent of the microbiotic crust and the thickness of the moss surrounding the sample point were also related to water content in crusted plots. Plots with large amounts of thick, mossy microbiotic crust contained less soil water (Figure 27).  59  Figure 25:  The correlation between the grams of soil water on day 5, after the 2-mm rainfall, and the percent cover of moss and lichen in the 1-m surrounding the sample point in October at Kennedy Bench. 2  r = 0.3 p = 0.005 n =86  100  50  MOSS cover (%) (arcsine sqrt)  Figure 26:  100  LICHEN cover (%) (arcsine sqrt)  The correlation between the grams of soil water on day 5 moisture and the percent cover of moss and lichen in the 1-m surrounding sample points covered with a microbiotic crust (crusted plots) in October at Kennedy Bench . 2  10  o  a cn c o u  cr  o  c  °o o o  in a>  <> "  ° °  o o 0  r = - 0.47 p = 0.001 n =43 o o  ° 8  o o  0  <8> Oo  0  8o°  r = 0.34 p = 0.03 n =43  o o °  o  o  0°  CD  50 MOSS cover (%) (arcsine sqrt)  100  20 40 LICHEN cove r (%) (arcsine sqrt)  60  60  Figure 27:  The correlation between the grams of soil water on day 5 and the percent cover of microbiotic crust and thickness of the moss in the 1-m surrounding sample points covered with a microbiotic crust (crusted plots). in October at Kennedy Bench . 2  cn o —.  10  r = - 0.32 p = 0.04 n =43  ° r 1"  o  cr u)  ^o  S •  otf>^°°oa  3 « -  O <D  o  o  oo  8  2 o°r  cP o  » ° o o o o  "or  l^o*  o®  a  5  *  r = - 0.32 p = 0.04 n =43 o o  o  s.  -^3  0  O  50  100  10  MICROBIOTIC CRUST cover(%) (arcsine sqrt)  20  30  40  MOSS THICKNESS (mm)  Water loss between day 3 and day 5 was related to lifeform of the microbiotic crust, but the relationship was limited to specific effects. Shrubbed plots lost more water if the surrounding plot contained extensive areas of moss (Figure 28).  Figure 28:  The correlation between the grams of soil water lost between day 3 and day 5 and the percent of microbiotic crust cover and moss cover in the 1-m surrounding sample points covered by a shrub (shrubbed plots) in October at Kennedy Bench . 2  13  s?  CO T3 >> —  _  2  re o  TJ  o o  CO  O  c cn «> o  o —-  re .re  O  0  0  0  0  0  0  & 8  -2  ,  0  „°  0  0 0  0  0  f> 0  0  °  <s>  0  (> 0  0  0  0  0 0  0  <©o  ^ 0  u  0  r = 0.36 p = 0.02 n =44  0  0  0 0  0  1!  o •° -#0)co-» cna i  o  0  Q  Oo oo  0  0 0  O  0  0 0  r = 0.44 p = 0.003 n =44  Tj  20 40 60 80 MICROBIOTIC CRUST cover (%) (arcsine sqrt)  100  20  40 60 MOSS cover (%) (arcsine sqrt)  80  100  61  A relationship between lichen crusts and increased soil water following a rain shower was identified during the August sampling. However, the October data suggested the type of rain shower also influenced water absorption. In August, soil beneath both lichen and moss crusts absorbed portions of the 8-mm rain event, although soil beneath lichen dominated crusts had a tendency to absorb more water. In October the light 2-mm shower was absorbed by soils covered with crusts that were predominately lichen (crust, bare, shrub effects), but the showers were not absorbed by soils beneath moss crusts (shrub/crust effect) (Figure 23 and Figure 10). It is possible the moss leaves absorbed the rainwater, although lower air temperatures and intermittent rain showers would suggest extensive re-hydration was not necessary, making interception (water storage) the likely mechanism. If mosses were partially hydrated it is possible lichen crusts were fully hydrated and allowed more rainwater through to the soil. Between days one and three Kennedy Bench lost soil moisture, however soil water in the crust effect rose from an average of 5.25 + 0.16 on day 1 to 5.40 + 0.37 on day 3 (n = 22). Moisture interception by both moss and lichen followed by slow percolation into the soil from the numerous water storage sites may have caused this phenomenon. However, there was a very close relationship between lichen and grass cover (r = 0.60, p = 0.0001, n = 87) and this correlation may confound the interpretation of crust lifeform effects.  II - 3.1.6.2  Relationship between bare ground in the 1-m surrounding the sample point and October soil moisture at Kennedy Bench 2  Across the site, a large amount of bare ground in the area surrounding the sample point was correlated to reduced soil water content on day 1 and day 3 (Figure 29). Bare ground was not related to  62  water loss between day 1 and day 3, but there was a tendency for sample points surrounded by large amounts of bare ground to lose less moisture between day 3 and day 5 (Figure 30).  Figure 29:  The correlation between the grams of soil water on day 1, day 3, and day 5 and the percent cover of bare ground in the 1-m surrounding the sample points in October at Kennedy Bench. 2  DAY  o in cn  10  1  DAY  r = - 0.35 p = 0.0011 n =86  o o  Q°o  DAY  r = - 0.27 p = 0.01 n =86  o  6  3  5  r = - 0.01 p =0.9 n =86  °  ~. cr § •  B* o  m  C  0  o o  § ?  0  1_  0)  20  •  I Figure 30:  40  20 40 BARE GROUND (%) (arcsine sqrt)  60  60  40  60  The correlation between grams of water lost between day 3 and day 5 and the percent of bare ground in the 1 m surrounding the sample point in October at Kennedy Bench . 2  13  5? ~ n o cr v>  CO TO >. "O  c a>  (/>  O) o c  1; (A O  r = - 0.38 p = 0.0002 n =88  .4  Ol ~-  ni ^  «_  13  20  40  60  BARE GROUND (%) (arcsine sqrt)  During the August sampling, areas of bare soil (bare effect) absorbed water readily, but over time they were the driest on the site (Figure 1). This trend was also evident in October.  63  II • 3.1.6.3  Relationship between grass cover in the 1-m surrounding the sample point and October soil moisture at Kennedy Bench 2  In August vascular plants were associated with increased site soil water content the day after the rain and to increased soil moisture in areas not protected by a shrub, particularly the bare effect, throughout the sampling period. In October the dominant vascular plant cover, grass, was associated with increased site soil moisture after the 2-mm rain shower, but not the shower that occurred before the day 1 collection (Figure 31). However, grass cover in the non-shrubbed areas was related to increased soil water following both showers (Figure 32). Grass cover was not related to moisture loss in October. The association between grass and soil water on day 3 was unclear. Grass cover was not correlated to day 3 moisture content in non-shrubbed areas (Figure 32, day 3), but grass was associated with increased soil water in plots with high amounts of bare ground (bare effect) and this relationship continued into day 5 (Figure 33).  Figure 31:  The correlation between grams of soil water on day 1, day 3 and day 5 and percent of grass cover in the 1 m surrounding the sample point in October at Kennedy Bench. 2  DAY 1  I  1  10  DAY 3  r=0.14,p = 0.2,n = 86  r = 0.04, p = 0.7, n = 86  o  P P 8  ™ J-  s ~ c o  so re 5  r = 0.36, p = 0.0007, n = 86  o  at  c0) c n ~— °3Su c re  DAY 5  8m&g°<x CP o cr^ o °o '0 o  ' f  o  O  o 20  40  60  e % ° c o > 0  P -  n  o  o  0  " P  9  o  o 9 b 9 0  o  ^W o %  O o  O  20 40 GRASS cover (%) (arcsine sqrt)  60  20  40  60  64  Figure 32:  The correlation between grams of water on day 1, day 3, and day 5 and the percent grass cover in the 1m surrounding the sample points not covered by a shrub (non-shrubbed plots) in October at Kennedy Bench. 2  DAY  10  T3  1  DAY  r = 0.33 p = 0.03 n =43  O (A Ol  3  DAY  r = 0.35 p = 0.02 n =43  r = 0.22 p =0.14 n =43  o  5  jo o  o° o  oo°°«« 8  !£ o  oQPf°o o o  0  v>  O  OO  oo  o^oo  o  So o  0  o  2 • 3 »  o  20  Figure 33:  40  o  20 40 GRASS cover (%) (arcsine sqrt)  60  60  20  60  40  The correlation between grams of soil water on day 1, day 3, and day 5 and the percent of grass cover in the 1 m surrounding sample points of bare soil not covered by a shrub (bare effect) in October at Kennedy Bench. 2  DAY  5  DAY  10  •D O to Ol  o . o  2  1  °°  ° n  Q  o  O  o  o o o  a8 <<*b  O  CQ o  o^o  o  o £> o P°o OCT  o  OO  5  r = 0.47 p = 0.03 n =22  o  _c  o  DAY  r = 0.47 p = 0.03 n =22  r = 0.25 p = 0.25 n =22  <0  3  „ c  ro  *J  c a *^  c o o &.  0)  £  o  20  40  60  20  o  40  60  20  40  60  GRASS cover (%) (arcsine sqrt)  Grass requires full light for optimum growth (Weaver 1968). Across the site, grass increased with lichen cover (r = 0.6, p = 0.0001, n = 87), which also favoured the high light in the non-shrubbed areas  65  (Figure 10, page 46) and decreased with thick mosses (moss thickness: r = - 0.31, p = 0.004; moss: r = 0.39, p = 0.0002; n = 87), which preferred the shaded sites. The delayed release of moisture intercepted by the grasses may explain the postive association immediately following a rain shower. However, lichen are also correlated with increased soil water immediately after rain events and have ample storage sites to intercept and slowly release water. Because both grass and lichen contain attributes that lend themselves to water conservation and because the variables are strongly correlated, it is impossible to determine the impact of either ecosystem component, without further study. Differences in the way vascular plant or grass cover related to soil water in August and October may indicate the importance of the herb layer. Herbs increased in areas protected by a shrub (shrubbed plots) that contained thick mosses (r = 0.31, p = 0.04, n= 43) and these areas absorbed moisture during the August rain event, but did not absorb water during the October shower after the herb layer had died back. Herb roots may affect rainwater absorption by providing a conduit through the moss mat.  II - 3.1.6.4  Relationship between sand particle content in the 1-m surrounding the sample point and October soil moisture at Kennedy Bench 2  There was a minor association between soil texture and both soil water content and water loss. Although there were no site trends, sandy soils were correlated to increased day 1 soil moisture in crusted plots (Figure 34), particularly in areas with only microbiotic crust (crust effect). In addition, sand content was related to increased soil moisture on day 3 in areas with only a shrub cover (shrub effect) (Figure 35). Water in the shrub effect was lost quickly however, and despite the 2-mm rain shower, between day 3 and day 5 water loss increased with sand particle content (Figure 36).  66  Figure 34:  Correlation between grams of soil water measured on day 1 and the percent of sand particles in the 1m surrounding sample points covered with a microbiotic crust (crusted plots) in October at Kennedy Bench. 2  10  X  r = 0.37 p = 0.01 n = 43  _  S * *7 ~° c  a  co o  2  20  40  60  80  SAND PARTICLE CONTENT (%) (arcsine sqrt)  Figure 35:  The correlation between grams of soil water and the percent of sand particle content on day 1 in crust effect and on day 3 in shrub effect in October at Kennedy Bench DAY  =  10  Si  oi  o o  t  1  DAY 3  Crust Effect  r = 0.57 p = 0.007 n = 21  3 T CM S  rl 15  r = 0.46 p = 0.04 n = 21  ce  Shrub Effect  5  ODO  o  o°  o J&o  2co  3  o  20  40  60  SAND PARTICLE CONTENT (%) (arcsine sqrt)  80  20  40  60  80  SAND PARTICLE CONTENT (%) (arcsine sqrt)  67  Figure 36:  Correlation between grams of soil water lost between day 3 and day 5 and the percent of sand particles in the 1 m surrounding sample points of bare soil that were covered by a shrub (shrub effect) in October at Kennedy Bench. 2  •a >. .•= ra o  r = 0.49 p = 0.02 n =22  o  o o  o  00  °  20 40 60 SAND PARTICLE CONTENT (%) (arcsine sqrt)  80  Sand content was not related to any plot variables and it is unclear why sand would be associated with higher soil water content in specific areas (crust effect on day 1 and shrub effect on day 3). Infiltration of rainwater is higher in sandy soils, however, sands drain quickly and hold less soil water than finer textured soils (Brady 1990). It is possible the microbiotic crust and shrubs on Kennedy Bench were capturing sand particles moved by wind. Sands are the largest soil particles and are the first to settle with a decrease in wind speed. As well, because sand particles are heavier they will roll across the soil surface in lighter winds and could be caught by the rough surface of the microbiotic crust (Miicher et. al. 1988, Longton 1992). These tendencies may have resulted in concentrations of sand particles beneath shrubs and microbiotic crust and the relationship between sand and higher soil water on day 1 and day 3 may be an indication of this phenomenon rather than a direct effect of sand on soil water. This study did not identify a significant difference in sand particle content between effects (p = 0.43, F = 0.92, df = 3,84), but a more in-depth study would be required before a proper conclusion could be reached.  68  The majority of bare soils beneath shrubs absorbed moisture during the 2-mm rain shower. However, in some cases soils continued to dry during this three-day period. A dense shrub cover may have intercepted the rainfall depriving the bare soil of moisture. The positive correlation between sand particle content and soil water loss between day 3 and day 5 may indicate the tendency for sandy soils deprived of moisture to dry-out faster than other soils.  69  II - 3.2  WATER TOWER  11-3.2.1  Soil MoistureContent in AUGUST at WATER TOWER  Kennedy Bench and Water Tower received 8-mm of rain on August 17. The day following the rain event the average soil moisture content at Water Tower was 3.7 + 0.12 g H2O/100 g soil dw, approximately one-half the day 1 average measured at Kennedy Bench. A full 7 1 % of the soil water at Water Tower was lost by day 3 and, on average, only 15% of the day 1 soil moisture remained five days after the rain event. At Kennedy Bench in August and October soil beneath crusted plots received less rainwater, but retained more of the moisture than soil in non-crusted plots (Figure 2A - day 1, page 39 and Figure 24A day 5, page 57). Soil beneath crusted sample points did not absorb less water than non-crusted plots at Water Tower. However, plots with a microbiotic crust cover did retain more soil water throughout the sampling period, although the differences between factor levels were not as pronounced as those seen at Kennedy Bench (Figure 37 and Figure 2, page 39). There was a significant interaction between the factors crust and shrub on day 3. The interaction accounted for 3.7% of the total variation in soil moisture content (Table 6), indicating water content varied with different combinations of shrub and crust. Although there was no difference in soil water content between effects on day 1, areas with both a shrub and microbiotic crust (shrub/crust) held more soil water on day 3 (Figure 38). Shrub/crust contained 38% more moisture than microbiotic crust alone (crust) and 74% more water than shrub alone (shrub effect). By day 5 there was no difference in average water content between the crust and the shrub/crust effects (T = -1.48, df= 42, p = 0.15). The effect of microbiotic crust and shrub on moisture content varied over the five-day period. A crust or shrub cover did not affect soil water on day 1, but both crust and shrub were important on day 3.  70  On day 5, microbiotic crust at the sample point was the only important factor affecting soil water content (Table 6).  Table 6:  | SAMPLING DAY day 1 R = 0.038 2  n=44 day 3 R = 0.525 2  n = 44,43 day 5 R = 0.481 2  n=44  ANOVA -  Effect of microbiotic crust and shrub on August soil moisture  content at Water Tower on day 1, day 3, and day 5  SOURCE OFV; VRIATION ^1fHlTN::!lp:| ;  DF  ;;  :Tn«iB-|' |j n i |.|ij.|i|.|.|.fflirt]!jmiMjia jmm ljlliji ;  TYPE HI SUM OF SQUARES  MEAN SQUARE  F  %OF TOTAL VARIATION  :™ i!!!|PR>F !  III  Model Microbiotic crust Shrub Microbiotic crust x shrub Error  3 1 1 1 84  10.90 6.116 1.127 3.657 278.960  3.63 6.116 1.127 3.657 3.32  1.09 1.84 0.34 1.10  0.3562 0.1784 0.5617 0.2970  2.0% 0.4% 1.3%  Model Microbiotic crust Shrub Microbiotic crust x shrub Error  3 1 1 1 83  240.075 204.301 16.422 16.959 217.122  80.025 204.301 16.422 16.959 2.616  30.59 78.10 6.28 6.48  0.0001 0.0001 0.0142 0.0127  45% 3.6% 3.7%  Model Microbiotic crust Shrub Microbiotic crust x shrub Error  3 1 1 1 84  40.033 38.333 1.316 0.385 43.157  13.344 38.333 1.316 0.385 0.514  25.97 74.61 2.56 0.75  0.0001 0.0001 0.1133 0.3892  46% 1.6% 0.5%  At Kennedy Bench a shrub and microbiotic crust cover (shrub/crust effect) intercepted more water initially, letting less water through to the soil, but the soils beneath the shrub and crust dried slower than other areas. Interception of the rainwater was not evident at Water Tower, but the shrub and crust combination did slow drying (Table 7). Water conditions in shrub/crust effect resembled the crust more than the shrub effect at both sites. At Kennedy Bench it was suggested that the addition of a shrub to microbiotic crust changed crust characteristics and the changes enhanced water retention in the soil. However, because there was no difference in water interception patterns of the effects at Water Tower it is unlikely crust characteristics differed. It is possible a shrub cover enhanced the ability of the crust to retain moisture on day 3 by shading the sample area and reducing evaporation.  71  Figure 37:  Mean grams of soil water measured on day 1, day 3 and day 5 for crusted and non-crusted and shrubbed and non-shrubbed plots. Measurements were taken in August at Water Tower, (mean +1 s.e.). B.  • CRUSTED  m  • NON-CRUSTED  Day 1  • NON-SHRUBBED  Day 5  Day 1  Day 3  Day 5  Mean grams of soil water measured on day 1, day 3, and day 5 for bare, shrub, crust, and shrub/crust effects in August at Water Tower, (mean + 1 s.e.).  ?  •a  5  -I  o  in 4 o> o o O 3  I  CM  i  X  ra c CD  2  c o 1 o i_  &  wal  Figure 38:  Day 3  SHRUBBED  c co  0  Bare  Shrub  NON-CRUSTED  Crust  Shrub/ Crust  CRUSTED  72  II - 3.2.2  Soil Moisture Loss in AUGUST at WATER TOWER  Throughout the sampling period drying patterns at Water Tower differed between crusted and noncrusted plots, but not between shrubbed and non-shrubbed areas (Table 7) (Figure 37). Over the five days there was also an interaction between shrub and crust, but the effect of the shrub/crust combination changed after day 3.  Table 7:  Tl  Repeated Measures MANOVA: Effect of crust and shrub on the grams of water lost between day 1 & day 3 and between day 3 & day 5 in August at Water Tower  SOURCE OF VARIATION  L  DAY 1  DAY 3  /3  /5  Time Time x Microbiotic crust Time x Shrub Time x Microbiotic crust x Shrub Error Time Time x Microbiotic crust Time x Shrub Time x Microbiotic crust x Shrub Error  DF  1 1 1  1 83 1 1  1 1 83  TYP|ji|i SUM OF SQUAKESI 1290.68 142.31 4.14 17.41 162.795 79.97 33.03 4.27 6.17 95.93  MEAN SQUARE  |§§§ii  F  PK  F  1290.68 142.31 4.14 17.41 1.96  658.04 72.56 2.11 8.88  0.0001 0.0001 0.1499 0.0038  79.97 33.03 4.27 6.17  69.19 28.58 3.70 5.34  0.0001 0.0001 0.0580 0.0234  1.16  MB  TO IVARMIONI 80% 9% 0.3 % 1 % 36% 15% 2% 3%  Between day 1 and day 3 plots without a microbiotic crust cover (non-crusted) dried faster than plots with a crust (crusted) and the addition of a shrub to crusted areas slowed the drying. This trend was also recognised at Kennedy Bench. On average, crusted plots lost 54% of the day 1 moisture in the three days following the rain. The crusted areas lost an additional 26% of the day 1 soil water between day 3 and day 5 and 2 0 % of the day 1 moisture remained five days after the rain event (Figure 37). Drying patterns of the effects with microbiotic crust (crust and shrub/crust) differed. Between day 1 and day 3 shrub/crust effect lost 4 2 % of the soil water compared to a 66% loss in crust effect (Figure 38). Between day 3 and day 5 the shrub/crust effect lost the majority of the moisture, 36% compared to a 16%  73  loss in crust effect. By day 5 there was no difference in the average soil water content of the two crusted effects (Figure 38). Non-crusted plots held slightly more water on day 1, but lost an average of 8 7 % of the moisture by day 3 (Figure 37). Moisture loss in the non-crusted plots between day 3 and day 5 averaged 3% and only 10% of the day 1 soil water remained after five days. In all plots the majority of water loss occurred within three days of the rain event, but water loss was less in areas with a crust and moderated considerably in areas with both a shrub and crust cover. The fact that crusted plots lost more water than non-crusted areas between day 3 and day 5 does not indicate that microbiotic crust cover leads to high water loss, rather it highlights the difference in available water between these plots three days after a rain.  II - 3.2.3 Plot conditions associated with soil moisture in AUGUST at WATER TOWER  II - 3.2.3.1  Relationship between microbiotic crust in the 1-m plot surrounding the sample point and August soil moisture at Water Tower 2  Similar to Kennedy Bench, characteristics of the microbiotic crust in the 1-m surrounding the 2  sample point influenced soil water content and loss across the site and the impact of the association changed during the sampling period. Large amounts of crust (Figure 39) comprised of thick mosses (Figure 40, Figure 41) were associated with reduced site soil moisture content on day 1, but increased water content on day 3 and day 5.  74  Figure 39:  Correlation between the grams of soil water on day 1, day 3 and day 5 and the percent of microbiotic crust in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  1  DAY  3  18 —  o  r 12 cr  o o O <9 <5> o  CA CD  5  = 0.36 = 0.0006 = 87  CA O)  o o  DAY  r = 0.27 p = 0.01 n =87  ©o°o ° o  8° C  S  o  0  20  o  40  60  80  0 20 40 60 80 MICROBIOTIC CRUST Cover (%) (arcsine sqrt)  CD  Figure 40:  20  40  60  80  Correlation between the grams of soil water on day 1, day 3 and day 5 and the percent of moss in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  1  DAY  18  o  CA  o„  o o^,o  DAY  5  r = 0.30 p = 0.005 n =87  r = - 0.44 p = 0.0001 n = 87  Q  cn  3  o o^cPo ^ > ° o  20  CD  40  60  20 40 60 MOSS Cover (%) (arcsine sqrt)  80  i  Figure 41:  80  Correlation between the grams of soil water on day 1, day 3 and day 5 and the average thickness of the moss in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  18  o CA  s» 3'« CD  i  TJ  S  8 8  DAY  3  DAY  5  r = - 0.37 p = 0.0004 n =87  o  o .—. 12 ^ cr o in  o  1  r = 0.30 p = 0.005 n =87  6goo o  9  o  O  10  88, 15  0  5  10  15  10  15  MOSS THICKNESS mm  75  The association between lichen and soil water content also resembled the relationship seen at Kennedy Bench; soil water was higher on day 1 in plots with large amounts of lichen, but there was no relationship between moisture content and lichen cover on days three or five (Figure 42 and Figure 6).  Figure 42:  The correlation between soil moisture on day 1, day 3 and day 5 and the percent of lichen cover in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  o  in u>  1  DAY 3  DAY  18  r = 0.08 p =0.5 n =87  o o  v—  12  <  1  §o$ o°o8°° 0  r = 0.28 p = 0.008 n =87  o  O)  S ^ 6  °  10  20  5  30  8o  o  §o §fio& %o> 10 20 LICHEN cove r (%) (arcsine sqrt)  10  o o o % ° a  ° 0 0  o ° 8 oo  20  30  The influence of microbiotic crust characteristics on water loss at the sample point also changed after day 3. Between day 1 and day 3 soil water was conserved if the sample point was surrounded by large amounts of microbiotic crust comprised of thick mosses (Figure 43). After day 3 soil water loss increased in areas with the same conditions (Figure 44). Microbiotic crust at the sample point retained soil water that continued to be lost after day 3, when non-crusted points were dry. Microbiotic crust surrounding the sample point had the same effect. If plots contained extensive thick, mossy crusts they continued to lose water after day 3, in contrast to areas with more bare soil that had dried by this time. Microbiotic crust at the sample point or in the vicinity of the sample point improved soil moisture conditions.  76  Figure 43:  The correlation between the grams of water lost between day 1 & day 3 and the percent cover of microbiotic crust, moss, and thickness of the moss in the 1-m surrounding the sample point in August at Water Tower. 2  T-  =  o  ra •a O) *J c oo r O 0)  12 9  I  o —  2 O)  5  8  2  o  0  0  8  o  Q  o  0 20 40 60 80 MICROBIOTIC CRUST Cover (%) (arcsine sqrt)  Figure 44:  o  o e  -»o- - —J' o  5  re  8  8 8o§  8  <o  0  r = - 0.51 p = 0.0001 n=87  o ° *o °  o  ' ra c o '—• >. C " re  TJ TJ TJ C  r = - 0.57 p = 0.0001 n =87  oo. ° <?<o oooxp  0  20 40 60 MOSS Cover(%) (arcsine sqrt)  5  10  15  MOSS THICKNESS (mm)  Correlation between grams of water lost between day 3 and day 5 and percent cover of microbiotic crust and moss and thickness of the moss in the 1-m surrounding the sample point in August at Water Tower. 2  r = 0.26 p = 0.02 n = 87 o  0 20 40 60 80 MICROBIOTIC CRUST Cover (%) (arcsine sqrt)  II - 3.2.3.2  o  "  20 40 60 MOSS Cover (%) (arcsine sqrt)  80  0  5  10  15  MOSS THICKNESS (mm)  Relationship between bare ground in the 1-m surrounding the sample point on August soil moisture at Water Tower 2  At Kennedy Bench areas of bare soil absorbed more precipitation and lost the water rapidly. There was no difference in the absorption patterns of soil in the bare or crusted sample points at Water Tower (Figure 38), but bare ground did dry quickly and sample points surrounded by bare ground lost large amounts of moisture between day 1 and day 3 (Figure 45). Plots with less than 10% bare ground in the  77  surrounding 1-m lost an average of 2.51 + 0.82 g H2O/100 g dw, while plots with 40 to 5 1 % bare ground 2  lost an average of 6.64 + 0.39 g H2O/100 g dw during the same period. Water Tower plots with extensive bare ground contained less soil water on days three and five, similar to the findings at Kennedy Bench (Figure 46). Water loss decreased during this time, presumably because little moisture remained in the area (Figure 45).  Figure 45:  The correlation between the grams of water lost between day 1 and day 3 and between day 3 and day 5 and the percent of bare ground in the 1-m surrounding the sample point in August at Water Tower. 2  DAY 1-DAY 3  DAY 3 - DAY 5  0.44 P = 0.0001 r=-  n = 87  0  20  40  BARE GROUND (%) (arcsine sqrt)  60  0  20  40  60  BARE GROUND (%) (arcsine sqrt)  78  Figure 46:  The correlation between grams of soil water on day 1, day 3 and day 5 and the percent of bare ground in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  o  1  DAY  3 r = - 0.52 p = 0.0001 n = 87  18  It)  DAY  O)  r = - 0.41 p = 0.0001 n =87  o  r" o  5  cr  m 12  S s 3 "5  S S3. °o l^ u. 0)  20  40  60  20  40  60  0  20  40  60  BARE GROUND (%) (arcsine sqrt)  £  Although large amounts of bare ground were associated with lower soil moisture the condition at the sample point was a more important determinant for water content. Shrub and crust effects held similar amounts of bare soil (Figure 47), but areas sampled under a microbiotic crust (crust effect) held 5 8 % more soil water on day 3 and 4 2 % more moisture on day 5 than samples taken from bare soil (shrub effect) (Figure 38). Crust at the sample point had a greater influence on soil water content than the bare soil surrounding the point.  79  Figure 47:  Mean percent cover of bare ground in the 1-m surrounding the sample point of bare, shrub, crust, and shrub/crust effect at Water Tower, (mean + 1 s.e.) 2  50 T3 C 3  40  Ol  30  o p  20 10  Bare  Shrub  NON-CRUSTED  II - 3.2.3.3  Crust  Shrub/ Crust  CRUSTED  Relationship between vascular plant cover in the 1-m plot surrounding the sample point on August soil moisture at Water Tower 2  Grasses and herbs were associated with site soil water content on day 1 and water loss between day 1 and day 3, although the components had counteracting effects. Site water content increased with grass cover and decreased with large amounts of herbs (Figure 48). However, the association between grass and herbs in the surrounding area and water content did not override conditions at the sample point, as day 1 soil water content did not differ between the factors or factor levels (Table 6 and Figure 38).  80  Figure 48:  The correlation between grams of soil water on day 1 and the percent cover of grass and herbs in the 1-m surrounding the sample point in August at Water Tower. 2  q=- 18 ' o  r = 0.26 p = 0.02 n =87  cr  in  2 • 12 5 -2.  0 ocfoo^  0 0  o°  0  r = - 0.39 p = 0.0002 n =87  o  0 o8g  ° °o  0  o  0  O  —-  O CD  |  T>  O  »  20 40 GRASS Cover(%) (arcsine sqrt)  60  10 20 30 HERB Cover (%) (arcsine sqrt)  40  The relationship between the vascular plant components and water loss also varied, as did the extent of their association. Grass cover surrounding the sample point influenced all plots and high amounts of grass were correlated with higher water loss. Herb cover was only associated with water loss in plots without shrubs. In the non-shrubbed plots water loss was lower if the surrounding area contained many herbs (Figure 49).  Figure 49:  The correlation between the grams of water lost between day 1 and day 3 and the percent of grass cover in all plots and the percent of herbs in plots without a shrub (non-shrubbed) in August at Water Tower. ALL PLOTS  CO  12  o in cn  §* ° <n  2.  sa ia  — co co —' M co S TJ  NON-SHRUBBED PLOTS  TJ  9 6 3 .8  O EPoCfloOOQ bo ^O O o Q-S^J oo  ft  r = 0.22 p = 0.04 n = 87  l°  °  u  20 40 GRASS Cover(%) (arcsine sqrt)  i  oo,oo o o o o  °° «  8 o oo o o o  o  r = - 0.37 p = 0.01 n =43  60  10 20 30 HERB COVER (%) (arcsine sqrt)  40  81  Grasses were not related to water content on day 3 or day 5, but increased herb cover continued to correlate to low moisture content in areas without a crust cover, specifically bare effect (n = 22; day 3: r = - 0.47, p = 0.03; day 5: r = - 0.48, p = 0.02). Neither grasses or herbs were associated with water loss between day 3 and day 5. The association between vascular plant components and water content and loss may actually be a reflection of other effects. Across the site grasses dominated areas of silty, bare soils with low amounts of litter (n = 88; silt: r = 0.32, p = 0.003; bare: r = 0.21, p = 0.05; litter: r = - 0.24, p = 0.02), while herbs increased in sandy areas that held extensive crusts of thick moss (n = 88; sands: r = 0.41, p = 0.0001; extent of microbiotic crust: r = 0.32, p = 0.002; moss: r = 0.34, p = 0.001; moss thickness: r = 0.33, p = 0.002). Grasses were associated with increased water content on day 1 and increased water loss between day 1 and day 3, very similar to the pattern of soil moisture in bare soils. Similarly, the relationship between herbs and water content and loss closely resembled the moisture patterns identified in microbiotic crust areas. It is possible the association between vascular plants and moisture conditions are indications of the powerful influences of microbiotic crust and bare soil, but further study would be required for a proper conclusion.  II - 3.2.3.4  Relationship between litter in the 1 -m plot surrounding the sample point on August soil moisture at Water Tower 2  Litter was related to moisture content and loss. Areas with large amounts of litter lost fewer grams of water in the three days following the rain and the plots contained more water than areas with low amounts of litter on day 3 and day 5 (Figure 50, day 1- 3 and Figure 51). Because the littered areas contained more available water, evaporation was higher during the last three sampling days (Figure 50, day 3- 5).  82  Figure 50:  The correlation between grams of water lost between day 1 and day 3 and between day 3 and day 5 and the percent cover of litter in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  1  TO DAY  3  DAY  12  3 TO DAY 5  r = - 0.38 p = 0.0003 ° "«=87  5 TJ  0^8  r = 0.38 p = 0.0003 n = 87  0  O)  S cr * o CN  o.  c a  o B n  3 j  al.  y  5  Figure 51:  °Oo  o  -+-  20  40 60 80 LITTER Cover (%) (arcsine sqrt)  100  20  40  60  80  100  Litter Cover (%) (arcsine sqrt)  The correlation between grams of soil water on day 1, day 3, and day 5 and the percent cover of litter in the 1-m surrounding the sample point in August at Water Tower. 2  DAY  o  18  in cn  o .—.  2• O  CO  r = - 0.03 p = 0.8 n=87°  1  DAY  3  DAY  r = 0.43 p = 0.0001  o  5  r = 0.30 p = 0.004 n =87  <fi  12  3'«  °o o O  TJ  50  100  50 LITTER Cove r (%) (arcsine sqrt)  100  50  100  83  II - 3.2.3.5  Relationship between sand particle content in the 1-m plot surrounding the sample point on August soil moisture at Water Tower 2  The soil texture at Water Tower contained an average of 77.88 + 1.22 % sand and the high sand content was related to reduced site soil water on day 1 (Figure 52) and reduced moisture content in noncrusted plots on day 5 (Figure 53).  Figure 52:  The correlation between grams of soil water and percent of sand in the soil of the 1-m plot surrounding the sample point in August at Water Tower 2  DAY  o  18  CO  en 12  1  DAY  r = - 0.42 p = 0.0001 n =87  DAY  5  r = -0.15 p=0.2 n =87  r = 0.03 p =0.8 n =87  o cn rc £  3  6°  08°®  <9°o  6  o  n  V •*-»  c o o  0)  50  100  50  £  100  50  100  SAND content (%) (arcsine sqrt)  Figure 53:  The correlation between grams of soil water on day 5 and the percent of sand in the 1 -m plots of the non-crusted areas in August at Water Tower. o  2  r = - 0.63  CO  cn _ 6 o er  2 10cu 4  p = 0.0001 n = 44  o  cn _c  co  = ™. 2 o o cu  £  T3  20  40 60 80 SANDcontent(%) (arcsine sqrt)  100  84  Between day 1 and day 3 less soil water was lost in areas with sandy soils (Figure 54), an unexpected result since water and air usually pass freely through the large pores that separate the sand grains (Brady 1990). The positive association of sand content on water conservation was particularly evident in crust effect and there was a strong relationship between sand, total microbiotic crust (r = 0.54, p = 0.0001, n = 88), moss cover (r = 0.64, p = 0.0001, n =88), and moss thickness (r = 0.55, p = 0.0001, n = 88). It was likely the nature of the crust species colonising the sandy soil (mosses rather than lichens), rather than the sand itself that was responsible for conserving soil water.  Figure 54:  The correlation between the grams of water lost between day 1 and day 3 in all plots and in the crust effect and the percent of sand in the soil of the 1-m surrounding the sample point in August at Water Tower. 2  ALL PLOTS  CRUST EFFECT  12 r = - 0.53 p = 0.01 n = 21  o  °  8  o  ^ o °o oO  40  50 60 70 SAND Content (%) (arcsine sqrt)  80  0  20 0 40 60 Sand Content (%) (arcsine sqrt)  80  85  __n-Z.2A  Soil Moisture Content in OCTOBER at WATER TOWER  Between September 2 and October 10, Water Tower received 16 mm of rainfall, seven millimetres less than that received by Kennedy Bench during the same period (Table 1). Throughout the October data collection Water Tower contained roughly one-half the soil moisture of Kennedy Bench. The average day 1 soil moisture at Water Tower was 2.10 + 0.1 g H2O /100 g dw, by day 3 average soil moisture had dropped to 1.76 + 0.1 g H2O /100 g dw. Two millimetres of rain fell on day 4 and on day 5 Water Tower averaged 2.36 + 0.08 g H 0 / 100 g dw. 2  Although the average amount of soil moisture at Water Tower was much less than Kennedy Bench the effect of the microbiotic crust and shrub on water content was similar. A microbiotic crust cover at the sample point was the most important factor affecting soil moisture on days one and three and shrub cover was important on day 5 (Table 8). However, in contrast to the October findings at Kennedy Bench, a microbiotic crust cover was also important for soil water content at Water Tower on day 5.  Table 8:  ANOVA -  n= 44  Effect of Microbiotic crust and shrub on October soil moisture content at Water Tower on day 1, day 3, and day 5  P  %OF TOTAL  DAY1 R = 0.182 2  DAY  3  R = 0.208 2  DAY 5 R = 0.314 2  Model Microbiotic crust Shrub  53.732 1 1 1 84  17.91 44.176 4.059 5.497  44.176 4.059 5.497 240.817  2.867  Model  60.425  20.142  Microbiotic crust Shrub Microbiotic crust x shrub Error  55.951 3.031 1.443 230.586  2.745  64.054  21.351  Microbiotic crust x shrub Error  84  Model Microbiotic crust Shrub Microbiotic crust x shrub Error  1 1 1 84  27.267 31.107 5.680 140.104  55.951 3.031 1.443  27.267 31.107 5.680  6.25  0.0007  15.41 1.42 1.92 7.34  0.0002 0.2374 0.1698  15% 1% 2%  0.0002 20.38 1.10 0.53  12.80 16.35 18.65 3.41  0.0001 0.2964 0.4704  19% 1% 0.5 %  0.0001 0.0001 0.0001 0.0685  13% 15% 3%  1.668  86  On average, crusted plots contained more soil water than non-crusted plots on days one and three and less soil water on day 5 after the 2-mm rainfall (Figure 55A). Crusted plots held a mean of 2.45 + 0.14 g H 0 / 100 g dw on day 1,29% more soil water than non-crusted plots (1.75 + 0.11 g H 0/ 100 g dw). 2  2  All plots lost soil water between days one and three, but crusted areas retained 3 2 % more moisture than the non-crusted during this period (crusted: 2.09 + 0.10 g H2O/100 g dw; non-crusted: 1.43 + 0.16 g H2O/ 100 g dw). Crusted plots continued to lose soil moisture after the 2-mm rain shower and average soil water content dropped slightly to 2.07 + 0.10 g H2O/100 g dw on day 5. In contrast, there was a dramatic increase in soil water in the non-crusted plots as a result of the rain shower. Interception of the rainwater by the microbiotic crust in the crusted plots and rapid absorption of rainwater by the bare soil in noncrusted areas was also seen at Kennedy Bench.  Figure 55:  Mean grams of soil water measured on day 1, day 3 and day 5 for crusted and non-crusted and shrubbed and non-shrubbed plots. Measurements were taken in October at Water Tower (mean +1 s.e.). B.  5* TJ  O  • Crusted  B Shrubbed  • Non-Crusted  • Non-Shrubbed  ID Ol  Day  1  Day  3  Day  5  Day  1  Day  3  Day  5  There was no difference in soil water content between shrubbed and non-shrubbed plots on days one and three, but the non-shrubbed areas contained 2 3 % more soil water than shrubbed on day 5 (Figure  87  55B). In response to the 2-mm rainfall, average soil moisture in the non-shrubbed plots increased 3 0 % from 1.87 + 0.17 g H 0/100 g dw to 2.67 + 0.12 g H 0/ 100 g dw. Bare effect experienced the largest 2  2  increase in soil water; 4 9 % compared to only a 5% increase in crust effect (Figure 56). After the day 4 rain average soil water content also increased in the shrubbed plots, but only by 19%, from 1.66 + 0.10 g H2O/ 100 gdw to 2.05+ 0.1 gH 0/ 100gdw. 2  The impact of the rainfall on the day 5 soil moisture measurement of the effects in the shrubbed plots was dramatic. Soil moisture in the shrub effect increased by 42%, while soil water dropped 8% in the shrub/crust effect during the same day 3 to day 5 period (Figure 56). The dramatic difference demonstrated the effect was more the result of a microbiotic crust versus bare soil than a shrub influence.  Figure 56:  Mean grams of soil water measured on day 1, day 3, and day 5 for each of the four effects (mean + 1 s.e.). October measurements at Water Tower. •  Bare  Shrub  D1  B D 3  Crust  L7JD5  Shrub/ Crust  NON-CRUSTED  II - 3.2.5  CRUSTED  Soil Moisture Loss in OCTOBER at WATER TOWER  The pattern of October rain showers differed between Kennedy Bench and Water Tower. Water Tower did not receive rain on day 2, but similar to Kennedy Bench, Water Tower did receive 2-mm of  88  rainfall on day 4 of the data collection cycle. Regardless of differences in the shower pattern, results of the soil moisture testing were very similar between sites; plots dried between days one and three and plots without microbiotic crust (non-crusted) increased in soil moisture after the 2-mm rain fall. The effect of the 2-mm rain on soil moisture in plots with a microbiotic crust cover was also similar between sites. Plots with just microbiotic crust (crust effect) increased in soil water, but sample points covered by both a crust and a shrub (shrub/crust effect) continued to lose soil water during this period. The amount of moisture lost at Water Tower did not differ between crusted and non-crusted or shrubbed and non-shrubbed plots between days one and three (Table 9, day 1 - 3). Plots with a microbiotic crust cover (crusted) lost an average of 0.36 g H2O /100 g dw and plots without a crust cover (non-crusted) lost a mean of 0.32 g H2O /100 g dw. Similar amounts of soil water loss were seen in shrubbed and non-shrubbed areas, which lost averages of 0.33 g H2O /100 g dw and 0.34 g H 0 /100 g 2  dw, respectively.  Table 9:  T I M E ;  Repeated Measures MANOVA: Effect of crust and shrub on the grams of water lost between day 1 & day 3 and between day 3 & day 5 in October at Water Tower  ^^  DAY 1 /3  isoupeEiGFtvap,A,T,oM  Mi  Time  1  Time x Microbiotic crust  1  Time x Shrub  1  Time x Microbiotic crust x Shrub Error  DAY 3/5  TYPE III  1 84  Time  1  Time x Microbiotic crust  1  Time x Shrub  1  Time x Microbiotic crust x Shrub  1  Error  84  lillsl HIPH 22.77 0.347  10.77  0.347 0.037  0.037  0.02  0.653  0.653  0.31  22.77  177.501  73.94  0.16  2.113 73.94  44.42  80.668  80.668  48.46  7.359  7.359  4.42  0.699  0.699  0.42  139.827  0.0015  11 %  0.6862  0.2 %  0.8944  0.02 %  0.5796  0.32 %  0.0001  24%  0.0001  27%  0.0385  2%  0.5189  0.2 %  1.665  Both microbiotic crust and shrub affected the loss of soil water between days three and five (Table 9, day 3-5). The amount of water lost differed for crusted and non-crusted and shrubbed and non89  shrubbed areas (Figure 55). The day 4 rain shower increased soil water in non-crusted areas by 1.22 g H2O /100 g dw, while crusted plots continued to lose an average of 0.02 g H2O /100 g dw soil water during the same period. A small increase in soil moisture occurred between day 3 and day 5 in both shrubbed and nonshrubbed plots (Figure 55B). Similar to the findings at Kennedy Bench the sample points with both a shrub and microbiotic crust continued to dry despite the rain shower. The shrub/crust effect at Water Tower lost an average of 0.15 g H2O /100 g dw between day 3 and day 5. Bare effect gained the most soil water at both sites with an average increase of 1.51 g H 0 /100 g dw between day 3 and day 5 at Water Tower 2  (Figure 56).  II - 3.2.6  Plot conditions associated with soil moisture in OCTOBER at WATER TOWER  II - 3.2.6.1  Relationship between microbiotic crust in the 1-m2 surrounding the sample point on October soil moisture at Water Tower  Microbiotic crust was not related to soil moisture content on day 1 or day 3, but after the 2-mm rainfall, soil moisture decreased if the area surrounding the sample point contained extensive amounts of mossy, microbiotic crust (Figure 57). There was also a tendency for soil moisture to be less if mosses in the crust were thick (day 5: r = - 0.27, p = 0.01, n = 88). The relationship between microbiotic crust characteristics and soil moisture was particularly noticeable in areas with just a microbiotic crust cover (crust effect) and moss thickness appeared to play a more important role in these areas (Figure 58).  90  Figure 57:  The correlation between grams of soil water on day 5 and the percent of total microbiotic crust and percent of moss in the 1-m surrounding the sample point in October at Water Tower. 2  oo o  15 cr  in  cu 10  2 a>  <«-> c Si c 2. o o s CD  5 6  T3  20  O  in CD  Figure 58:  40  60  80  MICROBIOTIC CRUST (%)  MOSS cover (%) (arcsine sqrt)  (arcsine sqrt)  The correlation between grams of soil water on day 5 and the percent of total microbiotic crust, moss and thickness of the moss in the 1-m surrounding sample points covered only by a microbiotic crust (crust effect) in October at Water Tower. 2  15  cn  o —  S T~ ~ cr o  m 10  °o o  2•  • -2- = r = - 0.55 8  I  |  -  u ~ 0) o  oo 6o o o o °  p = 0.008 n =22  DO  r =- 0.57 p = 0.006 n =22  °0°  CUD  o  O O  r = - 0.57 oro p = 0.006 o o o o „ _ o o n -22 no  h—  o  20 40 60 MICROBIOTIC CRUST (%) (arcsine sqrt)  20 40 MOSS cover (%) (arcsine sqrt)  60  5  10  15  MOSS THICKNESS (mm)  The soil beneath lichen dominated crusts received more water than soil beneath moss crusts in August and October at Kennedy Bench and in August at Water Tower. Lichens also allowed more water through to the soil in the October sampling at Water Tower but the correlation was only evident in the crust effect after the 2-mm rainfall (Figure 59).  91  Figure 59:  The correlation between grams of soil water on day 5 and the percent cover of lichen in the 1m surrounding sample points covered only by microbiotic crust (crust effect) in October at Water Tower. 2  15  cn  o — o  in  2*  - — if 5 £ | i_  r  10  o o o  o o  r = 0.47 p = 0.03 n =22  5  ~  o> o  10 20 LICHEN COVER (%) (arcsine sqrt)  30  Characteristics of the surrounding microbiotic crust did not influence drying during the first three sampling days, but between day 3 and day 5 there was a tendency for areas with extensive, mossy crusts to continue losing water despite the rain shower that occurred on day four.  Figure 60:  The correlation between grams of water lost between day 3 and day 5 and the percent cover of moss and microbiotic crust in the 1-m surrounding the sample point in October at Water Tower. 2  r = 0.21 p = 0.05 n =88  o  o o  o° 5> 8  °  0 20 40 60 MOSS COVER(%) (arcsine sqrt)  -4-9  h  1  20 40 60 MICROBIOTIC CRUST (%) (arcsine sqrt)  1  80  92  II - 3.2.6.2  Relationship between bare ground in the 1 -m surrounding the sample point and October soil moisture at Water Tower 2  Large amounts of bare ground did not dry faster than other areas at Water Tower, however similar to earlier findings bare soil was related to increased water absorption following the rain shower. The ability of bare soil to absorb precipitation quickly was evident across the site (Figure 61). On average, areas with greater than 3 0 % bare ground in the 1-m2 plot surrounding the sample point held 1.32 grams more soil water on day 5 than areas with less than 3 0 % bare ground (Figure 62).  Figure 61:  The correlation between grams of soil water on day 5 and the percent of bare soil in the 1-m surrounding the sample point in October at Water Tower.  2  15  T  o — O  10  CO  a•  o °  co  ~0  o  8 ^ o °  3 «  r = 0.48 p = 0.0001 n =88  §i mo  0 20 40 BAREGROUND (%) (arcsine sqrt)  Figure 62:  60  Average grams of soil water on day 5 in plots with an average of less than 3 0 % (n = 54) and more than 3 0 % (n = 34) bare ground in the 1-m surrounding the sample point in October at Water Tower, (mean + 1 s.e.) 2  o o  10 C icr  DAY 5  §"  ra = — o <=  CD 4-*  «  — .  Sk. i—  q> O  £  0 - 29.9%  30-51%  BARE GROUND (%) (arcsine sqrt)  93  Bare ground and areas covered with a microbiotic crust lost the same percentage of moisture during the first three October sampling days. It is possible the lower average air temperature during the sampling period (12.6°C Environment Canada, Atmospheric Environment Service) and the increased humidity from the frequent rain showers moderated drying rates across the site.  II - 3.2.6.3  Relationship between vascular plant cover in the 1-m plot surrounding the sample point and October soil moisture at Water Tower 2  Grasses were associated with soil water content in areas with microbiotic crust on day 1 and both grasses and herbs were correlated with site soil moisture on day 5. Similar to the August findings the relationship between vascular plant components and soil water was not clear. On day 1, a higher grass cover was associated with more soil water in crust effect (r = 0.66, p = 0.0008, n = 22), but less moisture in shrub/crust (r = - 0.44, p = 0.04, n = 22). On day 5 soil moisture across the site increased with grass cover, but increased herb cover was associated with reduced moisture (Figure 63).  Figure 63:  The correlation between grams of soil water on day 5 and the percent cover of grass and herbs in the 1-m surrounding the sample point in October at Water Tower. 2  cn  o  15  v>  in  ° 10  Sa 8  I  O  o  o.  O  <b' 0  Oo  8 o  °P8  r = 0.36 p = 0.0006 n =88 20 40 GRASS Cover(%) (arcsine sqrt)  O O °  °  60  r = - 0.31 p = 0.004 n =88  10 20 30 HERB Cove r (%) (arcsine sqrt)  40  94  Grasses surrounding the sample point were also related to moisture loss. In the shrub/crust effect water loss was lower between day 1 and day 3 if the area contained large amounts of grass (Figure 64, Day 1 - Day 3). The reverse was true between day 3 and day 5 where increases in both grasses and herbs were related to higher water loss (Figure 64, Day 3 - Day 5).  Figure 64:  The correlation between the grams of water lost between day 1 and day 3 and the percent cover of grass in the 1-m surrounding sample points covered by both a shrub and microbiotic crust (shrub/crust effect) and the correlation between the grams of water lost between day 3 and day 5 and the percent cover of vascular plants (grass and herbs) in the 1-m plots of the shrub/crust effect in October at Water Tower. 2  2  DAY 1 - DAY 3  cn —  O  6  r = - 0.59 p = 0.0041 n =22  L-  o cr *- co o cu - - co cn o 1  4 z  ~ £  ° V  ° *  2  5 g  4  &= ° co  DAY 3 - DAY 5  '°  r = 0.44 p = 0.04 n =22  803  °o  1  10 20 30 GRASS COVER (%) (arcsine sqrt)  40  0  °  W  o  1  1  1  20 40 60 80 VASCULAR PLANT Cover (%) (arcsine sqrt)  The conflicting effects of the vascular plant components would suggest other variables were more important for influencing soil moisture. For example, grasses in the crust effect increased at the expense of thick moss (r = - 0.57, p = 0.006, n = 22), while grasses in shrub/crust increased in silty, bare soils (n = 22: Silts: r = 0.47, p = 0.03; bare: r = 0.46, p = 0.03) at the expense of microbiotic crust, particularly thick mossy crusts (n = 22: microbiotic crust: r = - 0.42, p = 0.05; moss thickness: r = - 0.42, p = 0.05). Because grasses had counteracting effects on soil water it is likely the characteristics of the microbiotic crust associated with the grasses were influencing moisture conditions within the plots.  95  There was no obvious explanation for the herb and soil moisture correlation and it is possible the observed relationship reflects the impact of other plot conditions. For example, herbs increased in sandy soils (r = 0.41, p = 0.0001, n = 88) and extensive, thick mossy crusts (n = 88: extent of microbiotic crust: r = 0.32, p = 0.002; mosses: r = 0.34, p = 0.001; moss thickness: r = 0.35, p = 0.0007). Extensive mossy crusts were associated with reduced soil water, after a rain shower, at both Kennedy Bench and the August measurement at Water Tower. Sandy soils were also related to reduced soil water at Water Tower in August. Therefore, the association between herbs and soil water may actually reflect the relationship between soil moisture, the microbiotic crust and soil texture variables.  II • 3.2.6.4  Relationship between litter in the 1-m surrounding the sample point in October soil moisture at Water Tower 2  Litter was related to site soil water loss between day 3 and day 5 and site moisture content on day 5. Between day 3 and day 5 water loss was higher in areas with high litter cover (Figure 65) and on day 5 the areas with large amounts of plant litter contained less soil moisture (Figure 66A).  Figure 65:  The correlation between the grams of water lost between day 3 and day 5 and the percent cover of litter in the 1 -m surrounding the sample point in October at Water Tower. 2  s?  "d  r = 0.21 p = 0.05 n =88  to CO  a  2 x  4-1  cr  in  f  cn if 0  Z  &?-2  0) " 0 T3 g  C  -4  20  40 60 LITTER COVER (%) (arcsine sqrt)  80  100  The litter may have intercepted the 2-mm rain shower and limited the amount of moisture reaching the soil surface. However, contrary to the site findings large amounts of litter in the areas of bare soil (bare 96  effect) were related to increased soil water (Figure 66B). It is difficult to interpret the contradictory findings in bare effect. On average, bare effect held the lowest amount of litter and the highest amount of bare ground. It is possible the correlation between litter cover and soil water may reflect the ability of the bare soil surrounding the litter to absorb more of the light rain shower.  Figure 66:  The correlation between grams of soil water on day 5 and the percent cover of litter in all plots and in the plots where samples were taken from bare soil (bare effect) in October at Water Tower. A L L PLOTS cn o —  I" cr <A  2•  CD O  10  3 «  ° 41 o  BARE EFFECT  15  ° r O  B.  o  CO rP  CO* o r = - 0.31 p = 0.00 n =88 20  40  & f  o  3  ° o  60  0  o  O  oo  ft FP °°o -8° 0  LITTER Cover  II - 3.2.6.5  „oo o cl °  O^O  80  8  r = 0.46 p = 0.03 n =22 100  20  40  60  (%) (arcsine sqrt)  Relationship between sand particle content in the 1 -m surrounding the sample point and October soil moisture at Water Tower 2  Sand particle content was not related to water loss but was associated with water content on day 5. Soil moisture across the site decreased if areas contained a high sand component (Figure 67A). This tendency was particularly strong in areas without a shrub cover (bare and crust effect) (Figure 67B). Sandy soils were also related to low soil water after the rain shower in August (Figure 52).  97  Figure 67:  The correlation between grams of soil water on day 5 and the percent of sand in all plots and in the plots where samples were not protected by a shrub (non-shrubbed) in October at Water Tower.  A.  0  SITE  20  40  60  B.  80  SAND Content (%) (arcsine sqrt)  II - 4.0  II - 4.1  100  0  NONS -HRUBBED  20  40  60  Sand Content (%) (arcsine sqrt)  DISCUSSION  E F F E C T OF MICROBIOTIC CRUST AND SHRUB ON SOIL MOISTURE  Air temperature and precipitation patterns differed between the August and October sampling, but there was little difference in the effect of the microbiotic crust and shrub on soil moisture. At both Kennedy Bench and Water Tower plots with microbiotic crust at the sample point (crusted plots) initially received less rainwater, but soils beneath the crust retained the moisture that was received for longer periods than areas sampled in bare soil (non-crusted plots). Five days after the August rain event areas with a microbiotic crust cover at both sites retained approximately twice the soil moisture of bare ground. In October the difference between the effects was reduced and crusted plots held approximately 2 5 % more moisture than non-crusted areas. The periodic  98  showers during the October data collection were likely responsible for reducing soil moisture differences between the crusted and non-crusted plots. The findings are consistent with previous studies that show higher soil moisture beneath algae-crusted soils (Fritsch 1922, Booth 1941) and also beneath moss and lichen (Johnson and Thomas 1978, Wight et al. 1992, Gold and Bliss 1995). Shrub cover occasionally affected soil water content and loss, but it was usually the presence or absence of microbiotic crust in the shrubbed plots that influenced moisture conditions. Shrub cover was included as an effect because of the ability of shrubs to bring moisture from deep within the soil profile to the upper soil layers where it could be utilised by shallow rooted species (Caldwell 1990, Caldwell ef al. 1991, Wan et al. 1993). It was thought the ability to hydraulically lift soil water might alter moisture conditions at the soil surface. However, differences in water content because of a shrub cover in and of itself were not apparent in this study. The soil samples were collected from the 0 to 2-cm depth where vascular plant roots are rarely found. It is quite likely the hydraulic effect of the dryland shrubs was influencing water conditions deeper within the soil profile. Microbiotic crust beneath a shrub was more extensive and thicker and these conditions increased microbiotic crust effects on water infiltration and retention by soils beneath the microbiotic crust. The thicker the mosses in the crust the less water that was received by underlying soils, to the point that light rain showers were not measurable in the soil. In contrast, soils beneath lichen dominated crusts received the light showers. Lichens generally occupied open areas and were only found beneath the shrub canopy at Water Tower. The percent cover of lichens in the shrubbed plots at Water Tower was negligible and did not appear to affect moisture conditions.  99  II -  4.2  E F F E C T OF ENVIRONMENTAL CONDITIONS ON SOIL MOISTURE  Soil moisture at the sample point varied with environmental conditions in the local area, but it was difficult to separate their influence from the influence of sample point conditions. In general, the presence of a microbiotic crust in the surrounding 1-m plot, as well as the crust characteristics in this area had the 2  most consistent influence on soil moisture at the sample point. However, it is believed conditions at the sample point were paramount. If a microbiotic crust covered the sample point (crusted plots), the positive effect on soil moisture conditions was enhanced by surrounding crust characteristics, such as extent, dominant lifeform and moss thickness. However, if the sample point was bare (non-crusted plots) the ability of the surrounding microbiotic crust to moderate soil moisture varied by site.  II - 4.2.1  Effect of surrounding microbiotic crust on moisture conditions at sample points also covered with microbiotic crust  In August and October, at both Kennedy Bench and Water Tower, the water conditions in plots where soil samples were collected from beneath a microbiotic crust (crusted plots) changed with lifeform and morphology of the dominant crust species in the surrounding area. Soils beneath thick mosses received less precipitation than soils beneath lichen dominated crusts. Soils beneath thick mosses also dried slower than soils under lichens. At Kennedy Bench soil in plots dominated by lichen received more moisture, but it was the surrounding thick mosses that continued to be associated with higher water content five days after the August rain. The higher water content five days after the rain suggested the moss acted like a sponge, releasing the water slowly to the soil. Some moisture would likely move from the soil beneath the moss to the soil beneath the lichen, benefiting the entire crusted area. The positive relationship between lichen cover, moss thickness and soil moisture was not evident at Water Tower,  100  perhaps because the crusted areas contained so little lichen (2.44%) and the mosses were only one-half as thick as the mosses at Kennedy Bench. Other studies that examined the water holding capacity of mosses of different thickness were not found. However, in contrast to this study Brotherson and Rushforth (1983) concluded lichen and algal crusts reduced infiltration while mosses enhanced infiltration. Their experiment was somewhat misleading as infiltration was assessed as the time it took for 50-ml of water within a cylinder to be absorbed and no soil moisture measurements were taken in association with the test. They did acknowledge the moss appeared to be acting as a sponge, but did not expand on the possibility that the 50-ml of water was absorbed by the moss thalli. If they had measured the soil moisture content beneath the different crust species I expect the soils beneath the moss would have been drier than those beneath the lichens, which points to lower infiltration through a moss mat. They also did not provide information on the composition of the crust they were studying, so it is possible absorption patterns vary by species and/or their condition. An examination of infiltration through a crustose lichen dominated microbiotic crust was also done. They found higher infiltration through the crust as compared to bare soils (Loope and Gifford 1972). This study found the highest infiltration occurred in bare soils. The Loope and Gifford study was done in an area with soils that averaged 4 5 % silt and clay and these finer textured soils would have been more susceptible to surface sealing from rain impact than the sandy soils in this study. The relationship between water loss and surrounding microbiotic crust in crusted areas was similar among sites, but differed between August and October. In August little moisture was lost in the three days following the rain if the area surrounding the sample point held large amounts of thick moss, but after day 3 drying patterns were influenced by vascular plant and litter cover, and amount of bare soil. In October microbiotic crust surrounding the sample point did not affect water loss at either site. However, plots at Kennedy Bench that contained more lichen lost less water between day 3 and day 5. The relationship  101  between lichen and reduced water loss supports the earlier observation that soils beneath lichen crusts received moisture from the light rain shower that occurred on day four.  II - 4.2.2  Effect of surrounding microbiotic crust on moisture conditions at bare sample points  The effect of microbiotic crust on moisture conditions in areas that did not have microbiotic crust at the sample point (non-crusted plots) varied. Surrounding crust conditions were correlated to soil moisture conditions at both sites in August, but crust conditions were not related to soil water in October. At Water Tower moss cover surrounding the sample point was associated with reduced moisture on day 5. This is a deceptive association and appears to be more an indication of the drying patterns of bare, sandy soil than surrounding crust conditions. In contrast, the August rain identified several differences in the water absorption and drying patterns of soils in non-crusted areas at Kennedy Bench. Surrounding microbiotic crust conditions immediately following the August rain did not influence soil moisture in the non-crusted areas. However, bare sample points surrounded by an extensive mossy crust held more water three days after the rain. In addition, throughout the sampling period water loss at the bare sample points was moderated by the extensive moss. Kennedy Bench and Water Tower contained the same average amount of microbiotic crust and at both sites more moisture was retained for longer periods in areas covered by a crust. However, the differences between the sites in plots where the sample points were bare suggest the microbiotic crust at Kennedy Bench was better at ameliorating soil moisture than the crust at Water Tower. It is also possible there were other factors that were also influencing water retention at Kennedy Bench.  102  At both sites sample points of bare ground absorbed more water than neighbouring crusted areas, but while the bare ground at Water Tower lost moisture rapidly regardless of surrounding microbiotic crust conditions, a crust could slow the water loss in the bare soils at Kennedy Bench. It is possible differences in the microbiotic crust response between the sites could be attributed to the quality of the crust at Kennedy Bench, which contained greater species diversity and thicker mosses. However, the findings would suggest that although the microbiotic crust plays an important role, other ecological conditions also contribute to moisture conditions. In addition to possessing a higher quality microbiotic crust, Kennedy Bench also contained a higher cover of taller shrubs, more grasses and litter, a more silty soil and less bare ground than Water Tower.  II - 4.2.3  Effect of bare ground, vascular plants, litter and sand content on soil moisture  The amount of bare soil, vascular plant and litter cover, and sand content were also associated with moisture conditions at the two sites. Bare ground increased at the expense of microbiotic crust. The bare soil areas absorbed precipitation rapidly, but at both sites the abundant moisture was lost within three days of the rain event. This finding was similar to Brotherson and Rushforth (1983) who found the highest water infiltration rates occurred on bare soils, but differed from Loope and Gifford (1972) who documented the highest infiltration in lichen crusted soils. Rapid absorption of precipitation and drying was most noticeable in areas of extensive bare soil, but the rapid change in moisture conditions was moderated if vascular plants or litter sheltered the bare ground. However, vascular plants and litter were not as effective at water conservation as the microbiotic crust. Vascular plants did contribute to increased infiltration of the precipitation in all plots, regardless of sample point condition and it is likely the canopy of caespitose bunchgrasses were capturing and concentrating rainwater in the local area as reported by NdawulaSenyimba etal. (1971). 103  Sandy soils drain quickly (Brady 1990) and on Water Tower they were associated with reduced soil moisture. The association between sand and soil moisture was not as clear at Kennedy Bench. In August, sands at Kennedy Bench were only related to reduced moisture in shrubbed areas three days after the rain and in October they were related to increased soil water in crusted areas on day 1 and increased loss between day 1 and day 3. Differences in sand particle content between crusted and non-crusted plots were not identified, nor was sand content related to any other crusted plot variable. However, sand content did increase with moss thickness in plots without a shrub cover (r = 0.32, p = 0.04, n = 43). Microbiotic crusts are known to capture and concentrate silt particles (Evenari et al. 1971, Loope and Gifford 1972, Kleiner and Harper 1977, Anderson et al. 1982a, Brotherson and Rushforth 1983). It is possible the crusts in the Antelope-brush sites were capturing and accumulating wind-blown sand. A closer examination of changes in soil texture at the soil surface of crusted areas would be required to verify the relationships seen in this study.  II - 4.3  SUMMARY  1. Effect of microbiotic crust Immediately following a rain event and regardless of surrounding plot conditions, soils beneath a microbiotic crust received less rainwater than areas of bare soil, but the crusted soils retained the soil moisture for longer periods of time than the bare soil.  2. Effect of shrub: Shrub cover modified soil moisture conditions, but it was usually the presence or absence of microbiotic crust in the shrubbed plots that was associated with soil water. Microbiotic crust in shrubbed plots was more extensive and the mosses were thicker than areas that were not covered by a shrub.  104  3.  Effect of other site variables: In the absence of a microbiotic crust vascular plant and litter cover moderate moisture conditions, but they are not as effective at conserving soil moisture as a microbiotic crust cover.  4.  Effect of microbiotic crust lifeform: Soils beneath moss absorbed less water than soils beneath lichens, but the soils beneath lichens dried faster than those covered by moss. Soil beneath thick mosses did not receive light rain showers, but soils beneath lichens did.  5.  Effect of microbiotic crust surrounding the sample point: When sample points were covered with a microbiotic crust, more soil water was retained as the extent of the microbiotic crust surrounding the sample point and the thickness of the moss in the crust increased. Microbiotic crust surrounding bare soil only moderated the precipitation absorption and retention patterns of the bare soil when the site contained extensive microbiotic crust comprised of both lichens and thick mosses, a good cover of grass and litter, little bare ground, and relatively silty soils.  6.  Effect of bare ground: In the antelope-brush sites bare ground increased at the expense of microbiotic crust, absorbed precipitation rapidly and lost the moisture quickly. More precipitation was absorbed and it was lost more quickly as the amount of bare ground increased.  7.  Effect of season: The microbiotic crust was more effective at retaining soil water in hot, dry weather than in cool weather that was experiencing periodic showers.  105  8. Variability within sites: Soil moisture varied considerably across a site. Areas of bare soil lost their soil water within three days of a precipitation event in hot weather, while neighbouring areas with microbiotic crust retained over one-half their moisture five days after the rain.  106  C H A P T E R III  III  -1.1  M A N A G E M E N T IMPLICATIONS  D O E S THE MOISTURE HOLDING CAPACITY OF MICROBIOTIC CRUSTS HAVE ECOLOGICAL V A L U E ?  Increased water retention in soils covered by a microbiotic crust may provide an important link in the colonisation process of antelope-brush communities by aiding both floral and faunal species. Microbiotic crusts could be extremely valuable for reducing moisture stress in desert ecosystems (Brotherson and Rushforth 1983). In xeric environments, such as the south Okanagan, adequate moisture is the most limiting factor to plant growth (Walter and Stadelmann 1974) and is critical for germination and seedling establishment (Fenner 1985, Bewley and Black 1994). Available moisture between summer storms also moderates surface conditions for small mammals and invertebrates and many of the small mammals found in antelope-brush habitats are essential seed dispersers (Shatford 1997).  Ill-1.1.1 Vascular plant germination  There are mixed views on whether more soil moisture or longer periods of moisture are necessary for increased vascular plant germination in semiarid environments (MacMahon and Schimpf 1981, Bewley and Black 1994). Up to 16 days of imbibition for germination of an Artemisia species was reported (Gutterman 1993), but minimum precipitation events of between 6-mm and 25-mm have also been linked to increased germination (Juhren et al. 1956, MacMahon and Schimpt 1981). Work on the germination of bluebunch wheatgrass {Elymus spicata), a native perennial grass of antelope-brush habitats, concluded germination was both delayed and reduced when seeds experienced periods of dehydration during the germination process (Allen ef. al. 1994). 107  The microbiotic crust has been associated with increased (St. Clair et. al. 1984, Eldridge and Greene 1994) and decreased (Mcllvanie 1942, Fenner 1985, Schlatterer and Tisdale 1969, Keizeref. al. 1985) germination. Indirectly, the crust was also linked to increased soil moisture in a sage brush study. The study concluded sage brush and its associated litter, which included a microbiotic crust, extended the time plant-available water remained near the soil surface for 2 weeks compared to the open areas between the shrub that did not contain microbiotic crust or other litter (Wight et al. 1992). It is likely effects of the microbiotic crust on vascular plant germination and establishment will depend on both the type of seed tested and the species composition of the crust. It has been suggested that increased germination linked to the microbiotic crust may only be valid for larger seeded species or seeds adapted to penetrate the soil (Lesica and Shelly 1992). Ill -1.1.2 Vascular plant survival  Water in arid environments is also critical for seedling survival, plant growth, reproduction, and plant persistence (Dasberg 1971, MacMahon and Schimpt 1981, Lesica and Shelly 1992, Gutterman 1993). In some cases the microbiotic crust does not enhance vascular plant recruitment, but does increase seedling establishment and survival (Schlatterer and Tisdale 1969, Harper and St. Clair 1985, Keizer et al. 1985, Lesica and Shelly 1992).  108  Ill -1.1.3 Nutrient cycling  Water is the most limiting resource to plant growth, but in years of high precipitation soil nitrogen can be limiting (Romney etal. 1978). Algae in the microbiotic crust are known to produce nitrogen usable by plants (Shields 1957, Snyder and Wullstein 1973, Belnap and Harper in press, Evans and Ehlringer 1993) and moisture increases the activity of the soil algae (West and Skujins 1978).  Ill  - 1.2  HOW DO WE RETAIN THE MICROBIOTIC CRUST?  _ _ _  Microbiotic crust was important for moderating soil moisture conditions on antelope-brush sites and both the amount of crust present and the quality of the crust species contributed to the positive effect. The antelope-brush sites contained coarse soils and it is likely mosses and Cladonia lichens will always dominate the microbiotic crust in this habitat. Cladonia lichens allowed better infiltration and moss better retention of soil water. Hence, a microbiotic crust with a mix of both lifeforms would contribute to increased absorption of precipitation (lichens), as well as increased retention of soil water (mosses). The water retention ability of mosses increased in thicker mosses (increased mulch effect) and it is likely infiltration would be enhanced by well-developed lichens, specifically Cladonia species that contained both a primary and secondary thallus (increased surface storage sites). The amount of microbiotic crust in a local area was also important, especially for water retention. Extensive areas of contiguous microbiotic crust increased the length of time soil moisture remained in the local environment. An extensive, diverse, and well-developed microbiotic crust takes time to develop and a long-term management plan would be required to achieve such a community. Management practises aimed directly at conserving and enhancing the microbiotic crust would include strategies that limit soil disturbance and 109  reduce trampling of existing crusted areas during hot, dry weather when the crust is susceptible to disintegration. In this study, microbiotic crust improved moisture conditions at antelope-brush sites, but the crust's moderating effects were enhanced on sites in good ecological condition. The ability of the microbiotic crust to improve precipitation absorption and retention increased when the site contained little bare ground and a well-developed vascular plant community. In this study improved absorption and retention occurred on the site with finer textured soils and the water holding capacity of these soils would have contributed to the soil moisture conditions. However, the amount of water retained at this site surpassed the increase in soil water expected because of finer soils (Salter and Williams 1965, Herbel and Gibbens 1987). Antelope-brush sites would benefit from a comprehensive management strategy aimed at restoring the ecosystem. Two basic management objectives would benefit antelope-brush habitats of the South Okanagan: (1) reduce soil disturbance, and; (2) reduce the amount of bare ground. Achieving the objectives would include: (1) restricting livestock grazing on very coarse textured sites; (2) limiting livestock grazing to periods when crust species are fully hydrated, and; (3) increasing the recruitment and survival of native vascular plant species (seeding, planting, and mulching) and the microbiotic crust (seeding with crust propagules).  110  REFERENCES  Allen, P. S., S. B. Debaene-Gill, and S. E. Meyer. 1994. Regulation of germination timing in facultatively fall-emerging grasses. Proceedings - Ecology and management of annual rangelands. S. B. Monsen and S.G. Kitchen (eds). Intermountain Research Station, Forest Service, USDA: Ogden, Utah: 215-219 Anderson, D.C., K. T. Harper, and R.C. Holmgren. 1982a. Factors influencing development of cryptogamic soil crusts in Utah deserts. Journal of Range Management. 35:180-185. , K.T. Harper and S.R. Rushforth 1982b. Recovery of cryptogamic soil crusts from grazing on Utah winter ranges. Journal of Range Management. 35:355-359  Andrew, M. H., and R.T. Lange. 1986. Development of a new piosphere in arid chenopod shrubland grazed by sheep. I. Changes to the soil surface. Australian Journal of Ecology. 11:395-409. Baver, L.D. 1956. Soil physics. Third Edition. John Wiley & Sons: New York. Belnap, J . 1993. Recovery rates of cryptobiotic crusts: inoculant use and assessment methods. Great Basin Naturalist. 53,1:89-95. . 1994. Potential role of cryptobiotic soil crusts in semiarid rangelands. Proceedings: Ecology and management of annual rangelands. S.B. Monsen and S.G. Kitchen (eds). U.S.D.A. Intermountain Research Station: 179-185. and J . S. Gardner. 1993. Soil microstructure in soils of the Colorado Plateau: the role of the  cyanobacterium Microcoleus vaqinatus. Great Basin Naturalist 53:40-47.  and K.T. Harper. 1995. The influence of cryptogamic soil crusts on elemental content of tissue in  two desert seed plants. Arid Soil Research and Rehabilitation. 9:107-115..  Bewley, J . D. and M. Black. 1994. Seeds: Physiology of development and germination. Second Edition. New York: Plenum Press. Beymer. R.J. and J.M. Klopatek. 1991. Potential contribution of carbon by microphytic crusts in pinyonjuniper woodlands. Arid Soil Research and Rehabilitation. 5:187-198. Blackburn, W.H. 1975. Factors influencing infiltration and sediment production of semiarid rangelands in Nevada. Water Resources Research. 11:929-937. Booth, W.E. 1941. Algae as pioneers in plant succession and their importance in erosion control. Ecology 22:38-46. Brady, N.C. 1990. The nature and properties of soils. Tenth edition. MacMillan Publishing: New York. BC Conservation Data Centre. 1997. Tracking lists for vascular plants and plant communities. Wildlife Branch, BC Environment, Victoria, BC. 15pp.  111  Brotherson, J.D. and S.R. Rushforth. 1983. Influence of cryptogamic crusts on moisture relationships of soils in Navajo National Monument, Arizona. Great Basin Naturalist 43:73-79. Bryan, A. 1996. Vaseux Lake habitat management plan. South Okanagan Conservation Strategy. BC. Environment, Penticton, BC. 172pp. Caldwell, M.M. 1990. Water parasitism stemming from hydraulic lift: A quantitative test in the field. Israel Journal of Botany. 39:395-402. , J.H. Richards, and W. Beyschlag. 1991. Hydraulic lift: Ecological implications of water efflux from roots. Plant Root Growth. D. Atkinson (ed.). Blackwell Scientific Publications: Oxford. 423-436. Carter, M. R. (ed). 1993. Soil sampling and methods of analysis. Boca Raton : Lewis Publishers. Chilton, R.R.H. 1988. A summary of climatic regimes of British Columbia. Province of BC: Victoria, BC. Dasberg, S. 1971. Soil water movement to germinating seeds. Journal of Experimental Botany. 22: 9991008. Davidson, A.J., J.B. Harborne, and R.E. Longton. 1990. The acceptability of mosses as food for generalist herbivores, slugs in the Arionidae. Botanical Journal of the Linnean Society. 104:99-113. Daubenmire, R. 1968. Plant Communities. New York: Harper and Row Publishers. Dunne, J . 1989. Cryptogamic soil crusts in arid ecosystems. Rangelands. 11:180-182. Eldridge, D.J. 1993. Cryptogams, vascular plants, and soil hydrological relations: some preliminary results from the semiarid woodlands of eastern Australia. Great Basin Naturalist 53:48-58. and R.S.B. Greene. 1994. Microbiotic soil crusts: A review of their roles in soil and ecological  processes in the rangelands of Australia. Australian Journal of Soil Research. 32:389-415.  Evans, R. D. and J . R. Ehleringer. 1993. A break in the nitrogen cycle in aridlands? Evidence from 8  15  of soils. Oecologia. 94:314-317.  N  Evenari, M. and L. Shanan, N. Tadmor. 1971. The Negev. The challenge of a desert. Harvard University Press: Cambridge, Massachusetts. Fenner, Michael. 1985. Seed ecology. Chapman and Hill: London. Fitzpatrick, E.A. 1986. An introduction to soil science. Longman Scientific & Technical: New York. Fletcher, J.E. and W. P. Martin 1948. Some effects of algae and molds in the rain-crust of desert soils. Ecology. 29:95-100.  112  Frankland, C.J. 1974. Decomposition of lower plants. In Biology of plant litter decomposition I. (CH. Dickinson and J.F. Pugh (eds.). Academic Press: London. 3-36. Freidmann, E.I. and M. Galun. 1974. Desert algae, lichens, and fungi. In Desert Biology. G.W. Brown (ed.). volume 2. Academic Press: London. 165-212. Fritsch, F.E. 1922. The terrestrial algae. Journal of Ecology. 10: 220-236. Greene, R.S.B. and D.J. Tongway. 1989. The significance of (surface) physical and chemical properties in determining soil surface condition of red-earths in rangelands. Australian Journal of Soil Research 27:213-225. Gold, W. G. and L. C. Bliss. 1995. Water limitations and plant community development in a polar desert. Ecology. 75:1558-1568. Goward, T. B., McCune, and D. Meidinger. 1994. The lichens of British Columbia. Illustrated Keys. Part 1 - Foliose and squamulose species. Province of BC, Ministry of Forests Research Program: Victoria, BC. Grime, J.P. 1979. Plant strategies and vegetation processes. Chichester: John Wiley & Sons. Gutterman, Y. 1993. Seed germination in desert plants. Springer-Verlag: Berlin. Harper, K. T. and L.L. St. Clair. 1985. Cryptogamic soil crusts on arid and semi-arid rangelands in Utah: Effects on seedling establishment and soil stability. Bureau of Land management Report. Utah State Office: Salt Lake City, Utah. , and J.R. Marble. 1988. A role for nonvascular plants in management of arid and semiarid rangelands. Vegetation science applications for rangeland analysis and management. P.T. Tueller (ed). 135-169. and J . Belnap. unpublished. Nutrient uptake of two seed plants grown on soils with and without a  cyanobacterial amendment under greenhouse conditions.  , and R.L. Pendleton. 1993. Cyanbacteria and cyanolichens: Can they enhance availability of essential minerals for higher plants? Great Basin Naturalist. 53:59-72.  Herbel, Carlton H. and Robert P. Gibbens. 1987. Soil water regimes of loamy sands and sandy loams on arid rangelands in southern New Mexico. Journal of Soil and Water Conservation. 42:442-447. Johansen, J . R. and L. L. St. Clair. 1986. Cryptogamic soil crusts: Recovery from grazing near Camp Floyd State Park, Utah USA. Great Basin Naturalist. 46:632-640. Johnson, C D . and A.G. Thomas. 1978. Recruitment and survival of seedlings of a perennial species in a patchy environment. Canadian Journal of Botany. 56:572-580.  Hieracium  113  Juhren, M., F.W. Went, and E. Phillips. 1956. Ecology of desert plants. IV. Combined field and laboratory work on germination of annuals in the Joshua Tree National Monument, California. Ecology. 37: 318-330. Keizer, P. J . and B.F. van Torren, H.J. During. 1985. Effects of bryophytes on seedling emergence and establishment of shrot-lived forbs in chalk grassland. Journal of Ecology. 73:493-504. Kendrick, Bryce. 1992. The fifth kingdom. Second edition. Mycologue Publications: Newburyport, MA. Kleiner, E.F. and K.T. Harper. 1972. Environment and community organization in the grasslands of Canyonlands National Park. Ecology. 53: 299-309. and  . 1977. Soil properties in relation to cryptogamic ground cover in Canyonlands National  Park. Journal of Range Management. 30:202-205. Knutsen G. and B. Metting. 1991. Microalgal mass culture and forced development of biological crusts in arid lands. In Semiarid Lands and Deserts: Soil Resource and Reclamation. J . Skujins (ed.). Marcel Dekker: New York. 487-506. Krannitz, P. 1994. South Okanagan grassland conservation research. Progress Report, (unpublished). Lawrey. J . D. 1986. Biological role of lichen substances. The Bryologist. 89:111-122. . 1989. Lichen secondary compounds: evidence for a correspondence between antiherbivore and antimicrobial function. The Bryologist. 92: 326-328. Leach, W. 1931. On the importance of some mosses as pioneers on unstabel soils. Journal of Ecology. 19:97-102. Lesica, Peter and J . Stephen Shelly. 1992. Effects of cryptogamic soil crust on the population dynamics of>4rab/sfecu77da(Brassicaceae). American Midland Naturalist. 128: 53-60. Longton, R.E. 1988. The biology of polar bryophytes and lichens. Cambridge University Press: Cambridge. . 1992. The role of bryophytes and lichens in terrestrial ecosystems. In Bryophytes and lichens in a changing environment. Jeffrey W. Bates and Andrew M. Farmer (eds.) Clarendon Press: Oxford. 33-76. Loope, W.L and G.F.Gifford. 1972. Influence of a soil microfloral crust on select properties of soils under pinyon-juniper in southeastern Utah. Journal of Soil and Water Conservation. 27:164-167. Loria, M. and I. Herrnstadt. 1980. Moss capsules as food for the harvester ant, Messor. The Bryologist. 83: 524-525.  114  MacMahon, James A. and David Schimpt. 1981. Water as a factor in the biology of North American desert plants. In Water in Desert Ecosystems. Daniel D. Evans and John L. Thames (eds). Dowden, Hutchinson and Ross Inc.: Stroudsburg Pennsylvania. 114-171. Marble, J . R. and K. T. Harper. 1989. Effect of timing of grazing on soil-surface cryptogamic communities in a Great Basin low-shrub desert: A preliminary report. Great Basin Naturalist. 49:104-107. Mcintosh, T. T. 1986. The bryophytes of the semi-arid steppe of South-Central British Columbia. PhD Thesis, Department of Botany, University of British Columbia: Vancouver, B.C. Mcllvanie, S.K. 1942. Grass seedling establishment and productivity - Overgrazed vs. protected range soils. Ecology 23: 228-231. McKeaque, J.A. (ed.). 1978. Manual on soil sampling and methods of analysis. 2  n d  edition.  Metting, B. 1991. Biological surface features of semiarid lands and deserts. Semiarid lands and deserts soil resource and reclamation. J.Skujins (ed). Marcel Dekker: New York. 257-294. Miicher, H. J . and C.J. Chartres, D.J. Tongway, R.S.B. Greene. 1988. Micromorphology and significance of the surface crusts of soils in rangelands near Cobar, Australia. Geoderma. 42: 227-244. Nash, Thomas H. III. 1996. Lichen biology. Ccambridge University Press : Great Britian , S.L. White, and J.E. Marsh. 1977. Lichen and moss distribution and biomass in hot desert  ecosystems. The Bryologist. 80:472-479.  Ndawula-Senyimba, M.S., V.C. Brink, and A. McLean. 1971. Moisture interception as a factor in the competitive ability of bluebunch wheatgrass. Journal of Range Management. 24:198-200. Richardson, D.H.S. 1981. The biology of mosses. John Wiley and Sons Ltd.: New York. Rogers, R.W. 1972. Soil surface lichens in arid and subarid south-eastern Australia. III. The relationship between distribution and environment. Australian Journal of Botany. 20:301-316. . 1990. Ecological strategies of lichens. Lichenologist. 22:142-162. . and R.T. Lange. 1971. Lichen populations on arid crusts around sheep watering places in South  Australia. Oikos 22:93-100.  Romney, E.M. and A. Wallace, R.B. Hunter. 1978. Plant response to nitrogen fertilization in the northern Mojave Desert and its relationship to water manipulation. In nitrogen in desert ecosystems. N.E. West and J.J. Skujins (eds). US/IBP Synthesis Series. Vol.9. Dowden, Hutchinson and Ross: Stroudsburg, Pa. 232-243. Salter, P.J. and J.B. Williams. 1965. The influence of texture on the moisture characteristics of soils. II. Available-water capacity and moisture release characteristics. Journal of Soil Science. 16:310-317. 115  SAS Institute. 1988. SAS/STAT user's guide. Release 6.03. Cary, N.C.:SAS Institute Inc. Schlatterer, E. F. and E.W. Tisdale. 1969. Effects of litter of Artemisia, Chrysothamnus, and Tortula on germination and growth of three perennial grasses. Ecology 50:869-873. Schluter, A., T. Lea, S. Cannings, and P. Krannitz (1994). Antelope-brush ecosystems. Ecosystems in British Columbia at risk. BC Conservation Data Centre, Royal BC Museum, Environment Canada, and World Wildlife Fund: Victoria, BC. Schulten, J . A. 1985. Soil aggregation by cryptogams of a sand prairie. American Journal of Botany. 72: 1657-1661. Scudder, G.G.E. 1991. Threatened and endangered invertebrates of the South Okanagan. In Community action for endangered species. S. Rautio (ed.). Federation of BC Naturalists and N.W. Wildlife Preservation Society, Vancouver, BC. 47-58. Seaward, M.R.D. 1988. Contribution of lichens to ecosytems. In CRC handbook of lichenology. II. M.Galum (ed.). CRC Press: Boca Rataon, Florida. 107-129. Shatford, J.P.A. 1997. Seed dispersal, seed attributes and edaphic factors: their role and impact on the regeneration of antelope bitterbrush {Purshia tridentata, Rosaceae). M.Sc. Thesis. Department of Forestry. University of British Columbia, Vancouver, BC. Shields, L. M. 1957. Algae and lichen floras in relation to nitrogen content of certain volcanic and arid range soils. Ecology. 38:661-663. and L.W. Durrell. 1964. Algae in relation to soil fertility. Botanical Review. 30:92-128. Snyder, J.M. and L.H. Wullstein. 1973. The role of desert cryptogams in nitrogen fixation. American Midland Naturalist. 90:257-265. St. Clair, L.L., B.L. Webb, J.R. Johansen, and G.T. Nebeker. 1984. Cryptogamic soil crusts: Enhancement of seedling establishment in disturbed and undisturbed areas. Reclamation and Revegetation Research. 3:129-136. and J . R. Johansen. 1993. Introduction to the symposium on soil crust communities. Great Basin  Naturalist. 53:1-4.  Sveinbjornsson, B. and W.C. Oechel. 1992. Controls on growth and productivity of bryophytes: environmental limitations undercurrent and anticipated conditions. In Bryophytes and lichens in a changing environment. J.W. Bates and A.M. Farmer (eds). Clarendon Press: Oxford. Chapter 3. Thomson, J.W. 1967. The lichen genus Cladonia in North America. Toronto: University Press. Walter, H. and E. Stadelmann. 1974. A new approach to the water relations of desert plants. In Desert biology, volume II. G.W. Brown (ed.). Academic Press: New York.  116  Wan, C. and R.E. Sosebee, B.L. McMichael. 1993. Does hydraulic lift exist in shallow-rooted species? A quantitative examination with a half-shrub Gutierrezia sarothrae. Arid Soil. 153:11-17. West, N., E. 1990. Structure and function of microphytic soil crusts in wildland ecosystems of arid to semiarid regions. Advances in Ecological Research. 20:179-223. and J.J. Skujins. 1978. (eds). Nitrogen in desert ecosystems. US/IBP Synthesis Series. Vol. 9.  Dowden, Hutchinson and Ross : Stroudsburg, Pa.  Weaver, J . 1968. Prairie plants and their environment. A fifty year study in the mid-west. University of Nebraska Press: Lincoln. Wight, J.R., F.B. Pierson, C L . Hanson, and G.N. Flerchinger. 1992. Influence of sagebrush on the soil microclimate. In Proceedings - Symposium on ecology and management of riparian shrub communities. Gen. Tech. Rpt. INT-289. Warren P. Clary, E. Durant McArthur, Don Bedunah, and Carl L. Wambolt (compilers). Intermountain Research Station, Forest Service, USDA: Ogden, Utah. 181-185. Wittneben, U. 1986. Soils of the Okanagan and Similkameen Valleys. MOE Technical Report 18. Surveys and Resource Mapping Branch, BC Ministry of Environment, Victoria, BC.  117  APPENDIX A : Summary of vascular plant, bare ground, litter, and animal droppings data  Materials & Methods  Vascular plant data, as well as percent cover of bare ground, litter, and animal droppings were collected from each plot in August 1995. Vascular plant data collected included: • percent cover of shrub, broad-leaf herbs, grasses, and legumes; • shrub species and height (metres) in the shrubbed plots. All cover data were collected using a continuous cover scale. The 1-m plots were divided into 25, 20-cm cells. Each cell accounted for 4 % of the plot. Percent cover of each environmental variable in the plot was determined by adding the proportion of cover for each cell. 2  2  The lifeform system used to classify the vascular and non-vascular plants was broadly based on the lifeform categories devised by Kiichler (1967) and reviewed in Kent and Coker (1992) (Figure 68). Species classified as shrubs were based on descriptions by Hitchcock and Cronquist (1973) and included perennial woody shrubs and sub-shrubs, such as pasture sage (Artemisia frigida Willd.). Vertical shrub height was estimated to the nearest decimetre. The grass lifeform class included native and exotic grasses, both annual and perennial growth forms. Data collection followed the completion of the annual grass lifecycle. However, annual grasses were included as grass cover, rather than litter, if they were still attached to a root system supported by the soil and for the most part, the grass stems were vertical. Data collected for broad-leaf herbs and legumes [Leguminosae) included both exotic and native species, the majority had biennial and perennial growth forms. Annual broad-leaf herbs and legumes that fit the requirements outlined for annual grasses were also included as broad-leaf cover. The percent cover of bare ground was the area of soil within the plot not covered by vegetation, litter or animal droppings. Litter consisted of dead and decomposing leaves and shrub branches, as well as dead broad-leaf herbs and grass stems that were detached from their root systems and laying, horizontal, on the ground surface. Domestic livestock and wildlife droppings, found within the 1-m plot, were also recorded. Type of dropping was not noted, but it ranged from rabbit, deer and California bighorn sheep on Kennedy Bench to horse and cattle on East Osoyoos Lake. Mean percent cover of each environmental variable was computed and compared. 2  118  Figure 68:  Vascular and non-vascular plant lifeform categories used for data collection  Vascular Plant Cover  Orchard (43%) and Kennedy Bench (42%) averaged a much higher cover of antelope-brush than Water Tower (21%) or East Osoyoos Lake (23%) (Krannitz 1994). The mean shrub height also differed between sites, with shrubs at Orchard Block 1 (1.71 + 0.09 m) and Orchard Block 2 (1.71 + 0.11 m) and Kennedy Bench (1.55 + 0.10 m) averaging more than 50 cm taller than those at Water Tower (1.07 + 0.06 m) and East Osoyoos Lake (0.99 + 0.09 m) (Figure 69). 4  The average percent shrub cover for each block at Orchard is not available.  119  Figure 69:  The average height of the shrubs in the shrubbed plots of East Osoyoos Lake (EOL), Water Tower (WT), Kennedy Bench (KB), Orchard Block 1 (OR1), and Orchard Block 2 (OR2) 2.0  E  r a o  rfi  1.5  rb  2 i.o 3 0.5  0.0  n =  EOL  WT  KB  0R1  0R2  44  44  44  28  16  Grass cover was highest at Kennedy Bench with 28.12 +_1.92%, followed by East Osoyoos Lake with 23.66 + 1.65%. Orchard Block 1 and Water Tower contained similar quantities of grass cover with 20.29 + 2.3% and 1-9.21 + 1.13%, respectively. Orchard Block 2 contained considerably less grass cover with an average of only 8 + 2.62%. For all sites, grass cover was higher in the non-shrubbed plots (Figure 70  Figure 70:  Mean percent grass cover for the shrub/crust, shrub, bare and crust effects at East Osoyoos Lake (EOL), Water Tower (WT), Orchard Block 1 (OR1), and Orchard Block 2 (OR2), and Kennedy Bench (KB), (mean + 1 s.e.)  EOL  S Shrub/ Crust  • Shrub  • Bare  ^ Crust  WT  0R1  0R2  KB  120  Herbs and legumes were a minor component of the vascular plant cover at all sites. However, Water Tower and Orchard Block 2 held considerably more herbs than the other three sites. Herb cover averaged 7.4 + 0.79% at Water Tower and 7.39 + 1.28% at Orchard Block 2, which was almost double the herb cover found at Kennedy Bench (4.16 + 0.84%), Orchard Block 1 (4.25 + 0.81%) or East Osoyoos Lake (3.1 + 0.32%). The most common herb noted at Water Tower and Orchard Block 2 was diffuse knapweed (Centaurea diffusa Lam.).  Other Cover Categories The percent cover of litter, bare ground and feces within each 1-m plot was also recorded. Litter was a large component of the ground cover at all sites. Water Tower (57.85 + 2.48%), Orchard Block 1 (45.57 + 3.4%), Orchard Block 2 (57.63 + 4.75%), and East Osoyoos Lake (51.20 + 2.2%) had similar litter cover, but Kennedy Bench (82.97 + 1.47%) contained considerably more. 2  At all sites the highest average percent cover of litter was found in the shrubbed plots. However, differences in litter accumulation between shrubbed and non-shrubbed, varied by site. The largest differences in litter cover between shrubbed and non-shrubbed plots were found at East Osoyoos Lake and Water Tower, with 3 8 % and 36% more litter in shrubbed plots, respectively. Differences narrowed slightly on Orchard where shrubbed plots in both Block 1 and 2 contained 2 5 % more litter than the nonshrubbed plots. At Kennedy Bench, the difference in litter cover between shrubbed and non-shrubbed plots was only 1 1 % (Figure 71).  Figure 71:  Mean percent cover of litter in the shrubbed and non-shrubbed plots at Orchard Block 1 (OR1), East Osoyoos Lake (EOL), Water Tower (WT), Orchard Block 2 (OR2) and Kennedy Bench (KB) (mean + 1 s.e.). • Shrubbed • Non-Shrubbed  100 -r  0R1  EOL  WT  0R2  KB  121  The average cover of bare ground found on the sites ranged from 4 % to 2 2 % . Water Tower and East Osoyoos Lake contained the highest amount of bare ground with 21.71 + 1.68% and 20.5 + 1.81%, respectively. Orchard Block 1 held 16.2 + 2.01% bare ground and Orchard Block 2 slightly less with 14.39 + 2.59%, while Kennedy Bench contained the lowest amount of bare ground with an average of only 4.35 + 0.63%. Bare ground was more prevalent in the non-shrubbed 1-m plots than the shrubbed plots, however differences between the plot types were minimal at Kennedy Bench (Figure 72). 2  Figure 72: Mean percent cover of bare ground in the shrubbed and non-shrubbed plots at Kennedy Bench (KB), Orchard Block 2 (ORBL2), Orchard Block 1 (ORBL1), Water Tower (WT), and East Osoyoos Lake (EOL)(mean + 1 s.e.). • Shrubbed  KB  0RBL2 0RBL1  WT  EOL  Wildlife and livestock droppings were a minor component at all study sites. Water Tower contained the highest average feces cover with 2.49 + 0.4%, followed by East Osoyoos Lake with a mean of 1.36 + 0.38% and Orchard Block 2 with 1.5 + 0.54%. Of the remaining sites, Kennedy Bench held a mean of 1.16 + 0.2% and Orchard Block 1 contained the lowest average feces cover with only 0.5 + 0.18%. At all sites there was a higher cover of feces in the non-shrubbed effects.  Changes in Vascular Plants, Litter, and Bare Ground  The shrubs on the antelope-brush sites that were grazed by livestock (Water Tower and East Osoyoos Lake) were shorter and provided less ground cover than the sites grazed by wildlife, Kennedy Bench and Orchard Blocks 1 and 2. No relationship between grass and herb cover was identified. Data collection did not differentiate between native and exotic grasses or herbs. Results may have differed if palatable species were separated from colonising weed species. Litter was a large component of all sites and there was no difference in average cover between the sites except for wildlife grazed Kennedy Bench, which held approximately 3 6 % more litter than the remaining four sites. At all sites litter was concentrated in the shrubbed plots, but the difference between 122  litter cover in the shrubbed and non-shrubbed areas was greater in the sites grazed by livestock. The large difference in the distribution of litter on the sites grazed by livestock may indicate less biomass in the open areas. There were also differences in the amount of bare ground at each site and the differences appear to be related to both type of primary grazer and soil texture. Kennedy Bench and Orchard Block 1 and Orchard Block 2 were grazed by wildlife and held on average 79%, 23%, and 32% less bare soil than the sites grazed by livestock, respectively. In addition, there was little difference in the average amount of bare ground between Water Tower and East Osoyoos Lake, the sites grazed by livestock. However, there was a relationship between bare ground and sand content (r = 0.19, p = 0.0003, n = 352). Consequently, antelope-brush soils with high sand content appear to be more susceptible to pedon disturbances, but bare soil could also occur on finer textured soils that are subjected to grazing by livestock. Grazing has been linked to reduced biomass of vascular plants, reduced litter, and increased amounts of bare ground (Tisdale 1947, Ellison 1960, Blackburn 1983,1984, Takar ef a/., 1990, Vallentine 1990, Milchunas and Lauenroth 1992, Fleischner 1994). However, this study concentrated on the microbiotic crust and conclusions about the effects of primary grazer on the vascular plant community are speculative.  123  Appendix A References Blackburn, W.H. 1983. Livestock grazing impacts on watersheds. Rangelands. 5:123-125. . 1984. Impacts of grazing intensity and specialized grazing systems on watershed characteristics and responses. In. Developing strategies for rangeland management. Westview Press: Boulder, Colorado: 927-983. Ellison, Lincoln. 1960. Influence of grazing on plant succession of rangelands. The Botanical Review 26:178. Fleischner, T.L. 1994. Ecological costs of livestock grazing in Western North America. Conservation Biology. 8: 629-644. Hitchcock, C L . and A. Cronquist. 1973. Flora of the Pacific Northwest. Univeristy of Washington Press: Seattle. Kent, M. and P. Coker. 1992. Vegetation description and analysis. A practical approach. John Wiley & Sons: Chichester. Kuchler, A.W. 1967. Vegetation mapping. Ronald Press: New York. Milchuna, D. G. and W. K. Lauenroth. 1992. Quantitative effects of grazing on vegetation and soils over a global range of environments. Ecological Monographs. 63:327-366. Takar, A. A., J . P. Dobrowolski, and T.L. Thurow. 1990. Influence of grazing, vegetation lifeform, and soil type on infiltration rates and interrill erosion on a Somalion rangeland. Journal of Range Management. 43:486-490. Tisdale, E.W. 1947. The grasslands of the southern interior of British Columbia. Ecology. 28:346-382. Vallentine, J . F. 1990. Grazing management. Academic Press : San Diego.  124  APPENDIX B: ANOVA tables for non-vascular plant lifeform cover TABLE 1: DIFFERENCES IN TOTAL MICROBIOTIC CRUST (TTC) BETWEEN SITES Class Levels Values SITE 5 1 2 345 Number of observations in data set 352 :  Dependent Variable: TTC Sum of Source DF Squares Model Error CT. Total R-Square 0.033  Mean Square F Value Pr>F  4 2697.84 674.46 347 77246.63 222.61 351 79944.47 C.V. 45.85  Root MSE 14.920213  3.03  0.0178  TTC Mean 32.544523  Source SITE  DF Type ISS Mean Square F Value Pr > F 4 2697.84 674.4605603 3.03 0.0178  Source SITE  DF Type III SS Mean Square F Value Pr>F 4 2697.84 674.4605603 3.03 0.0178  TABLE 2: Difference in moss cover between crusted and non-crusted and shrubbed and non-shrubbed plots at East Osoyoos Lake Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in by group = 88  TABLE 3: Difference in moss cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Water Tower Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in by group = 88 Dependent Variable: TMOSS Source DF Sum of Squares F Value Pr > F Model 3 3849.17721767 8.70 0.0001 Error 84 12388.70508838 Corrected Total 87 16237.88230605 R-Square 0.237049 Source CRUST SHRUB  C.V. 44.34421  TMOSS Mean 27.3864904  DF 1 1 1  Type I SS F Value Pr > F 3623.05860581 24.57 0.0001 218.98075657 1.48 0.2264 7.13785530 0.05 0.8264  DF Source CRUST 1 SHRUB 1 1 CRUST*SHRUB  Type HISS F Value Pr>F 3623.05860581 24.57 0.0001 218.98075657 1.48 0.2264 7.13785530 0.05 0.8264  CRUSrSHRUB  Dependent Variable: TMOSS Source DF Sum of Squares F Value Pr>F Model 3 12250.9646552 26.91 0.0001 Error 84 12745.7686318 Corrected Total 87 24996.7332871 R-Square 0.490103  C.V.  34.43802  TMOSS Mean 35.7688705  Source DF CRUST 1 SHRUB 1 CRUST*SHRUB 1  Type I SS F Value 11012.2822410 72.58 1122.3645270 7.40 116.3178872 0.77  Pr > F 0.0001 0.0079 0.3838  Source DF CRUST 1 SHRUB 1 CRUST*SHRUB 1  Type III SS F Value Pr > F 11012.2822410 72.58 0.0001 1122.3645270 7.40 0.0079 116.3178872 0.77 0.3838  125  TABLE 4: Difference in moss cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Kennedy Bench Class Levels CRUST 2 SHRUB 2 Number of observations  Values 01 01 in by group = 88  Dependent Variable: TMOSS Source DF Sum of Squares F Value Pr > F Model 3 13221.7631074 27.87 0.0001 Error 84 13283.5522039 Corrected Total 87 26505.3153113 R-Square 0.498834 Source CRUST SHRUB CRUSrSHRUB  DF 1 1 1  C.V. 63.27652  TMOSS Mean 19.8735251  Type I SS F Value Pr > F 3920.70362578 24.79 0.0001 5700.00414368 36.04 0.0001 3601.05533794 22.77 0.0001  Source DF Type III SS F Value Pr>F CRUST 1 3920.70362578 24.79 0.0001 SHRUB 1 5700.00414368 36.04 0.0001 CRUST*SHRUB 1 3601.05533794 22.77 0.0001  TABLE 5: Difference in moss cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Orchard Block 1 Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in data set = 56 Dependent Variable: TMOSS Sum of Mean Source DF Squares Square F Value Pr > F Model 3 3765.239 1255.079 7.00 0.0005 Error 52 9327.626 179.377 Cor. Total 55 13092.865 R-Square 0.287579  C.V. Root MSE 42.16339 13.393186  TMOSS Mean 31.764961  Source DF Type I SS Mean Square F Value Pr>F CRUST 1 1939.0374 1939.037 10.81 0.0018 SHRUB 1 1651.393 1651.393 9.21 0.0038 CR*SHR 1 174.809 174.809 0.97 0.3281 Source DF Type III SS Vlean Square F Value  Pr>F  CRUST 1 939.0374 SHRUB 1 1651.393 CR*SHR 1 174.809  0.0018 0.0038 0.3281  1939.037 1651.393 174.809  10.81 9.21 0.97  126  TABLE 6: Difference in moss cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Orchard Block 2 Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in data set = 32 Dependent Variable: TMOSS Mean Sum of Square F Value Pr > F Source DF Squares 1199.287 15.41 0.0001 Model 3 3597.861 77.808 Error 28 2178.616 Cor. Total 31 5776.477 R-Square 0.622847  TABLE 7: Difference in lichen cover between crusted and non-crusted and shrubbed and non-shrubbed plots at East Osoyoos Lake. Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in by group = 88 Dependent Variable: TLICHEN Source DF Sum of Squares F Value Pr > F Model 3 26.09428498 0.63 0.6000 Error 84 1166.71585401 Corrected Total 87 1192.81013900  C.V. Root MSE TMOSS Mean 34.17405 8.8208682 25.811596  R-Square 0.021876  C.V. TLICHEN Mean 276.8492 1.34616901  Source DF ype I SS Mean Square F Value Pr>F CRUST 1 3546.681 3546.681 45.58 0.0001 SHRUB 1 32.546 32.546 0.42 0.5231 CR*SH 1 18.634 18.634 0.24 0.6284  Source CRUST SHRUB  DF 1 1 1 CRUST*SHRUB  Type I SS F Value 4.30443183 0.31 5.45579715 0.39 16.33405600 1.18  Pr>F 0.5792 0.5325 0.2813  Source DF Type III SS Mean Square F Value Pr>F CRUST 1 3546.681 3546.681 45.58 0.0001 SHRUB 1 32.546 32.546 0.42 0.5231 CR*SH 1 18.634 18.634 0.24 0.6284  Source CRUST SHRUB CRUSTSHRUB  Type III SS F Value 4.30443183 0.31 5.45579715 0.39 16.33405600 1.18  Pr>F 0.5792 0.5325 0.2813  DF 1 1 1  127  TABLE 8: Difference in lichen cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Water Tower Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in by group = 88  TABLE 9: Difference in lichen cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Kennedy Bench Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in by group = 88  Dependent Variable: TLICHEN  Dependent Variable: TLICHEN  Source DF Sum of Squares F Value Pr > F Model 3 733.51797959 5.35 0.0020 Error 84 3840.11757721 Corrected Total 87 4573.63555680  Source DF Sum of Squares F Value Pr > F Model 3 7107.21126360 18.64 0.0001 Error 84 10675.36699101 Corrected Total 87 17782.57825461  R-Square 0.160380  C.V. TLICHEN Mean 101.3605 6.67058217  R-Square  C.V. 0.399673  TLICHEN Mean 66.20424  17.0280915  Source CRUST SHRUB CRUSFSHRUB  DF 1 1 1  Type I SS F Value Pr > F 31.96987449 0.70 0.4054 678.96260002 14.85 0.0002 22.58550508 0.49 0.4841  Source CRUST SHRUB CRUSFSHRUB  DF 1 1 1  Type I SS F Value Pr > F 112.23205948 0.88 0.3500 5661.74478152 44.55 0.0001 1333.23442260 10.49 0.0017  Source CRUST SHRUB CRUSFSHRUB  DF 1 1 1  Type III SS F Value Pr > F 31.96987449 0.70 0.4054 678.96260002 14.85 0.0002 22.58550508 0.49 0.4841  Source CRUST SHRUB CRUSFSHRUB  DF 1 1 1  Type III SS F Value Pr > F 112.23205948 0.88 0.3500 5661.74478152 44.55 0.0001 1333.23442260 10.49 0.0017  128  TABLE 10: Difference in lichen cover between crusted and non-crusted and shrubbed and non-shrubbed plots Orchard Block 1 Class Levels CRUST SHRUB Number of observations  Values 2 01 2 01 in data s e t 56 :  Dependent Variable: TLICHEN Sum of Mean Square F Value Pr > F Source DF Squares Model 3 642.942 214.314 2.36 0.0822 Error 52 4725.419 90.873 Corr. Total 55 5368.362 R-Square 0.119765  C.V. Root MSE TLICHEN Mean 89.27022 .5327572 10.678541  TABLE 11: Difference in lichen cover between crusted and non-crusted and shrubbed and non-shrubbed plots at Orchard Block 2 Class Levels Values CRUST 2 01 SHRUB 2 01 Number of observations in data s e t 32 :  Dependent Variable: TLICHEN Sum of Mean Source DF Squares Square F Value Pr > F Model 3 146.937 48.979 0.89 0.4608 Error 28 1549.526 55.340 Corr Total 31 1696.463 R-Square 0.086614  C.V. ot M S E TLICHEN Mean 70.62218 7.4391005 10.533660  Source DF Type I S S Mean Square F Value Pr > F 391.828 4.31 0.0428 CRUST 1 391.828 SHRUB 1 57.030 57.030 0.63 .4318 194.084 2.14 0.1499 CR*SHR 1 194.084  Source CRUST SHRUB CR*SHR  DF Type I S S Mean Square Value Pr > F 1 39.843 39.843 0.72 0.4033 1 68.541 68.541 1.24 0.2752 1 38.553 38.553 0.70 0.4110  Source DF Type III S S Mean Square 391.828 CRUST 1 391.828 57.030 SHRUB 1 57.030 194.084 CR*SHR 1 194.084  Source CRUST SHRUB CR*SHR  DF Type III S S Mean Square F Value Pr > F 1 39.843 39.843 0.72 0.4033 1 68.541 68.541 1.24 0.2752 1 38.553 38.553 0.70 0.4110  F Value Pr > F 4.31 0.0428 0.63 0.4318 2.14 0.1499  129  APPENDIX C: Microbiotic crust species of the Antelope-brush shrub steppe  LICHENS  Species verified by T. Goward, Dr. J . Thomson, Dr. T. Ahti, and Dr. Roger Rosentreter.  Aspicilia sp. Caloplaca tominii Savicz Catapyrenium squamulosum (Ach.) Breuss Cladonia borealis S. Stenroos Cladonia cariosa (Ach.) Sprengel Cladonia cervicornis (Ach.) Flotow Cladonia chlorophaea (Sommerf.) Sprengel Cladonia fimbriata (L.) Fr. Cladonia macrophyllodes Nyl. Cladonia phyllophora Hoffm. Cladonia pleurota (Florke) Fr. Cladonia pocillum (Ach.) Grognot Cladonia pyxidata (L.) Hoffm. Cladonia rei Schaerer Cladonia symphycarpia (Florke) Fr. Cladonia verruculosa (Vainio) Ach. Collema tenax (Sw.) Ach. s. lat. Collema tenax (Sw.) Ach. var. crustaceum (Kremp.) Degel Diploschistes muscorum (Scop.) R. Sant. Endocarpon pusillum Hedwig Leptogium affin. tenuissimum (Dickson) Korber  BRYOPHYTES  Species verified by T. Mcintosh Mosses  Barbula unguiculata Hedw. Brachythesium albicans (Hedw.) B.S.G. Bryoerythrophyllum columbianum (Herm. Et Lawt.) Zander Bryum argenteum Hedw. Bryum caespiticium Hedw. Ceratodon purpureus (Hedw.) Brid. Didymodon vinealis Zander Encalyptra rhaptocarpa Schwaegr Phascum cuspidatum Hedw. Polytrichum juniperium Hedw. Polytrichum piliferum Hedw. Pseudocrossidium revolutum (Brid. In Schrad.) Zander Pterygoneurum ovatum (Hedw.) Dix. Tortula ruralis (Hedw.) Gaertn. Weissia brachycarpa C M . Hepatics  Cephaloziella divaricata (Sm.) Schiffn. Mannia fragrans (Balbis) Frye et Clark  Massalongia microphylliza (Hasse) Henssen Peltigera didactyla (With.) J.R. Laundon Peltigera ponojensis Gyelnik Psora cfr globifera (Ach.) A. Massal. Trapeliopsis granulosa (Hoffm.) Lumbsch  130  

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.831.1-0088593/manifest

Comment

Related Items