NITROGEN INPUTS BY BIOLOGICAL SOIL CRUSTS IN GRASSLANDS OF INTERIOR BRITISH COLUMBIA by Kasia Caputa B.Sc. Honours, University of Northern British Columbia, 2008 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN NATURAL RESOURCES AND ENVIRONMENTAL STUDIES (ENVIRONMENTAL SCIENCE) UNIVERSITY OF NORTHERN BRITISH COLUMBIA March 2013 © Kasia Caputa 2013 1+1 Library and Archives Canada Bibliotheque et Archives Canada Published Heritage Branch Direction du Patrimoine de I'edition 395 Wellington Street Ottawa ON K1A0N4 Canada 395, rue Wellington Ottawa ON K1A 0N4 Canada Your file Votre reference ISBN: 978-0-494-94152-2 Our file Notre reference ISBN: 978-0-494-94152-2 NOTICE: AVIS: The author has granted a non­ exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distrbute and sell theses worldwide, for commercial or non­ commercial purposes, in microform, paper, electronic and/or any other formats. L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'lnternet, preter, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conform em ent a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these. W hile these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada ABSTRACT The relationship between climate and inputs of nitrogen and carbon by biological soil crusts (BSCs) was examined in British Columbia’s Chilcotin and Thompson River grasslands. Variation in nitrogen and carbon of BSCs and associated soils was examined in relation to gradients in temperature and moisture availability. Patterns of N-fixation activity by BSCs were also examined in relation to climate, and used to model N- fixation rates throughout the year. BSCs were found to significantly contribute to soil carbon and nitrogen, particularly in the labile N fraction. Two peak periods were observed for N-fixation - one during the spring snowmelt, and one during midsummer rainy periods. N-inputs appeared to be sensitive to changes in precipitation patterns. Annual estimates of N-inputs for BSCs in this ecosystem could be as high as 83 kg N ha' 1but the fate of this N (i.e. plant uptake, denitrification or volatization) is still unknown. TABLE OF CONTENTS ABSTRACT................................................................................................................................... ii LIST OF TABLES......................................................................................................................... vi LIST OF FIGURES...................................................................................................................... vii ACKNOWLEDGEMENTS............................................................................................................ix BACKGROUND............................................................................................................................. 1 Grassland ecosystems of Interior British Columbia..................................................................1 Biological Soil Crusts............................................................................................................... 4 Carbon and Nitrogen Cycling in Grassland and Shrub-steppe Soils.........................................7 Biological N 2 Fixation............................................................................................................ 12 Biological Soil Crusts and Climate Change in Grasslands and Shrub-steppe........................15 RESEARCH OBJECTIVES.......................................................................................................... 19 CHAPTER 1: NUTRIENT COMPOSITION OF BIOLOGICAL SOIL CRUSTS AND ASSOCIATED SOILS ACROSS AN ELEVATION GRADIENT IN LAC DU BOIS GRASSLANDS, BRITISH COLUMBIA.....................................................................................21 INTRODUCTION.....................................................................................................................21 METHODS................................................................................................................................23 Study Site...............................................................................................................................23 Field Methods.........................................................................................................................25 Data Analysis.........................................................................................................................26 RESULTS..................................................................................................................................26 29 DISCUSSION CHAPTER 2: SEASONAL PATTERNS OF NITROGEN FIXATION IN BIOLOGICAL SOIL CRUSTS FROM BRITISH COLUMBIA’S CHILCOTIN REGION.................................. 34 INTRODUCTION.....................................................................................................................34 METHODS............................................................................................................................... 36 Study Site.............................................................................................................................. 36 Microclimate Measurements.................................................................................................. 37 Acetylene Reduction Analysis............................................................................................... 40 1^-Incubations......................................................................................................................42 Modeling BSC N-fixation activity under field conditions.....................................................43 RESULTS................................................................................................................................. 45 Farwell Canyon BSC Communities....................................................................................... 45 Acetylene Reduction Assays..................................................................................................47 BSC Microclimate Data.........................................................................................................47 Predicted BSC N-fixation....................................................................................................... 51 DISCUSSION............................................................................................................................57 Response of nitrogenase activity to microclimate and seasonal patterns............................... 57 Potential landscape-level nitrogen inputs............................................................................... 60 Seasonal patterns of nitrogen inputs by Chilcotin BSC communities.................................... 62 SUMMARY AND CONCLUSIONS............................................................................................64 REFERENCES LIST OF TABLES Table 1.1 Summary of climate normals at Kamloops Airport (Environment Canada 2012) and climate estimates at different elevations in Lac du bois Grasslands, British Columbia*.............. 24 Table 1.2: Selected Chemical Properties of Biological Soil Crusts and underlying mineral soils at Lac du Bois Grasslands, British Columbia. The mean differences reported are significantly different based on a paired two-sample t-test (P<0.05).................................................................27 Table 2.1. Climatic characteristics of the very hot dry bunchgrass biogeoclimatic subzone (BGxh3) for the study site (Steen and Coupe 1997)......................................................................36 Table 2.2. Summary of climate normals at Williams Lake Airport (from Environment Canada 2012) and at Farwell Canyon, British Columbia (estimated from Climate B.C. modelling using Wang et al. (2012)..........................................................................................................................37 Table 2.3. Summary of conversion ratio for ethylene production to 15N-incorporation (±SE) by biological soil crusts incubated for 48 hours at 21°C and 200 pmol PAR m * V . n=10................43 Table 2.4. Species identified in BSC samples from Farwell Canyon study site. Species marked with an "X" include those found in the dark BSC communities sampled for ARA incubations. 46 Table 2.5. Summary of estimated seasonal N-inputs per hectare by BSCs in a shrub-steppe in Farwell Canyon, B.C. The estimates are based on conversion from modeled C2 H2 inputs using both an experimentally derived C2 H2 :N2 conversion ratio of 0.06 and a commonly used conversion ratio of 3. Landscape level N-input estimates were determined using a BSC cover estimate of 52.7%...........................................................................................................................57 LIST OF FIGURES Figure 1.1 Views of the upper elevation grasslands (top left), lower elevation shrub-steppe (bottom left) and the representative biological soil crust (right) at the Lac du Bois Grasslands, British Columbia (Photos - P. Sanborn)........................................................................................25 Figure 1.2: Total Carbon, Total Nitrogen and Mineralizable Nitrogen in the biological soil crusts (BSC) and underlying soil (1-1 Ocm) across an elevation gradient in Lac du Bois Grasslands, British Columbia. Error bars represent standard error and letters denote significant differences (p<0.05) across soil type and elevation class................................................................................. 28 Figure 1.3: Ratio of Total C:N and 1SN natural abundance in the biological soil crusts (BSC) and underlying soil (1-1 Ocm) across an elevation gradient in Lac du Bois Grasslands, British Columbia. Error bars represent standard error and letters denote significant differences (p<0.05) across soil type and elevation class................................................................................................ 29 Figure 2.1 View of the vegetation and landscape representative of the study site, at Farwell Canyon, British Columbia..............................................................................................................38 Figure 2.2. Calibration curve for mean relative impedance response (n=8) and gravimetric moisture content (MC) of field collected BSC (n=8-12). Error bars represent standard error...... 39 Figure 2.3. Temperature and Light Response of acetylene reduction assay (ARA) fitted with multivariate curves by BSCs collected in (A) March and (B) July, n =10 for each datapoint.....48 Figure 2.4. (A) Relative ARA response to moisture content for BSCs pre-treated at optimum conditions and incubated at 21°C and 200 pmol PAR m 'V 1 (B) Relative ARA response over time starting at 2 hours after hydration to 30% moisture content.................................................. 48 Figure 2.5. Summary of BSC microclimate from 11 November 2009 to 1 October 2011 at Farwell Canyon, British Columbia.................................................................................................49 Figure 2.6. View of biological soil crusts melting out from under snow cover in early March at the Farwell Canyon study site. The width of the crust segment in the center of the image is ca. 9 cm.................................................................................................................................................. 50 Figure 2.7. BSC microclimate on south-facing aspect from 27 Feb. to 4 March 2011. From top: predicted rates of acetylene reduction (ARA) from microclimate model, air and BSC surface temperature, and incident photosynthetically active radiation (PAR) (pmol PAR m-2s-l ) at the BSC surface................................................................................................................................... 51 Figure 2.8.(A-D) Time (in hours) spent by hydrated (> 7.5% MC w/w) BSCs and (E-H) the estimated amount of C2 H2 fixed (in mmol of C2 H2 -m‘2) by hydrated BSCs at temperatures plotted in 5°C class intervals for winter (October to April) and summer (May to September), for the periods November 2009- September 2010 (Year 1) and October 2010 to September 2011 (Year 2). Calculations exclude periods of missing data (from 1 Nov - 11 Nov Year 1 and 9 Mar -13 Apr Year 2).................................................................................................................................... 52 Figure 2.9. Predicted ARA rates based on BSC microclimate measurements during a spring snowmelt event at Farwell Canyon in March 2011. The symbols represent ARA rates measured in situ with standard error (Y-error bars) and incubation time (X-error bars)................................53 Figure 2.10. Predicted ARA rates based on BSC microclimate measurements during a rain event at Farwell Canyon in May 2011. The symbols represent ARA rates measured in situ with standard error (Y-error bars) and incubation time (X-error bars).................................................. 54 Figure 2.11. Predicted ARA rates based on BSC microclimate measurements during a rain event at Farwell Canyon in July 2011. The symbols represent ARA rates measured in situ with standard error (Y-error bars) and incubation time (X-error bars).................................................................55 Figure 2.12. Monthly summary of BSC hydration time and estimated ARA activity at a shrubsteppe and shrub-steppe in Farwell Canyon, B.C. Asterisks denote time periods with missing data. ...................................................................................................................................................... 56 ACKNOWLEDGEMENTS Financial support for this study was provided by the Provincial Government of British Columbia as part of the Future Forests Ecosystems Scientific Council Grant. Thanks goes out to Ray Coupe and Rick Tucker at the Williams Lake Range Branch, for assistance in site selection, and to Riske Creek Ranch, for allowing us to work on their property. I would like to thank Burkhard Budel, and Curtis Bjork for identification of biological soil crust cyanobacteria and lichen species. Myles Stocki for providing 15N analysis and Clive Dawson for the remaining soil chemical analyses. Thank you to Chuck Bulmer for providing a digital elevation model for Lac du Bois park. Many thanks Katherine Stewart, Sarah Campbell, Kaela Perry and Chris Fetterly for assistance in the field and/or laboratory. None of this would have been possible without my co-advisors Darwyn Coxson and Paul Sanborn, for providing endless support, knowledge and advice, both in the field and during the analysis and writing stages. Also thank you to Lauchlan Fraser, for assistance and advice as a Committee member. Special thanks to my husband, family and friends, for your support and companionship on this great journey. BACKGROUND Grassland ecosystems of Interior British Columbia The grasslands and shrub-steppe in the interior plateau region of British Columbia are unique ecosystems with high ecological and recreational value and a critical resource for the Province's ranching industry. Roughly 90% of grasslands in British Columbia are used for grazing domestic livestock (GCC 2004) and have been important to British Columbia's ranching industry since the mid-1800s (Van Ryswyk and McLean 1989). Grasslands and shrub-steppe occupy less than 1% of the total land base of the province, yet provide a number of important ecosystem services, including water storage and filtration, carbon sequestration and habitat to almost 30% of threatened and/or endangered species in B.C. (GCC 2004). Most of the these areas in British Columbia seriously deteriorated through intensive use for livestock grazing during the 19th and early 20th centuries (Tisdale 1947, Wikeem and Wikeem 2004). Restoration efforts and improved range management beginning in the mid 20th century have led to the slow recovery of some grassland areas, but many are still highly threatened by conversion to urban and agricultural land, encroachment by forest and invasive non-native species and overuse by poorly managed grazing and recreational activities (GCC 2004). Climate change may introduce additional impacts to these sensitive ecosystems and intensive study is needed to gain an understanding of these impacts. Physiography, Climate and Soils The interior plateau of central British Columbia is located between the Coast Ranges and Columbia Mountains. The topography is rugged and deeply dissected by the Fraser and Thompson Rivers and their tributaries, which have steep-sided valleys 600 to 900 meters below the plateau (Valentine 1978). Conditions are generally warm to hot and dry in the summer, and moderately cold and dry in the winter. Most of the precipitation occurs in December to January as either snow or rain, with a second peak of rainfall in June, while the driest months are generally March and April (Meidinger and Pojar 1991). The climatic influence by the rainshadow effect of the mountain ranges to the west is most intense in the deeper valleys (Meidinger and Pojar 1991). The grasslands and shrub-steppe of the interior plateau occur mainly in these valleys, where forest establishment is restricted by drought. The soils found in these grasslands range from Brown to Black Chernozems; this variation in soil type is based on the degree of moisture effectiveness (estimated evapotranspiration minus precipitation). The Brown Chernozems occur at low elevations where evapotranspiration is highest, and have formed predominately on lacustrine silts and fluvial sands and gravels found in valley bottoms (Valentine 1978). These soils are characterized by a thin A horizon with an average carbon content of 1.3% (Fenger 1982). The Dark Brown and Black Chemozemic soils occur in subsequently drier and cooler areas, usually at higher elevations, where most of the parent material is glacial till composed of a variety of rock, including basalt lava, igneous granites, sedimentary shale, sandstone and limestone (Van Ryswyk and McLean 1989). Dark Brown and Black Chernozems are characterized by thicker and better expressed A horizons and increasing carbon content compared to the Brown Chemozoms. A common feature to most of the soils in this region is a silty aeolian veneer, on which most of the A horizons have formed (Fenger 1982). The gradient of moisture effectiveness found across elevations also has an effect on a number of soil physical and chemical properties. Some differences in soil properties with increasing elevation include increased cation exchange capacity (CEC), increased potassium (K), calcium (Ca), and magnesium (Mg), increased organic carbon, decreased C/N ratios, and decreased soil bulk density (Van Ryswyk et al. 1966). Ecology The bunchgrass (BG) zone in British Columbia occurs from valley bottoms up to approximately 700 - 1000 m in parts of the Okanagan valley, Similkameen River valley, Thompson River valley, Nicola River valley, and the middle Fraser and lower Chilcotin River valleys (Meidinger and Pojar 1991). This zone occurs mainly as isolated pockets within these deeply dissected valleys, but existed up to 1,300 to 1,800 m a.s.l. during the early Holocene (10,000 to 8,000 years ago). The bunchgrass zone represents the northern extent of the grassland and shrub-steppe that occupies the northern part of the intermountain region between the Rocky Mountains to the east and the Coast and Cascade Mountain to the west (Wikeem and Wikeem 2004). Pseudoroegneria spicata ((Pursh) A. Love) and Artemisia tridentata (Nutt.) are common throughout the region. These grasslands and shrub-steppe developed with little influence from grazing by large ungulates, and are much more sensitive to grazing impacts than those east of the Rocky Mountains (Mack and Thompson 1982). The British Columbia Ecological Classification system describes two Bunchgrass subzones that are found at two distinct elevational bands. Below 700 m, the lower grassland, or BGxh (Very Dry Hot subzone), is a shrub-steppe ecosystem dominated by P. spicata, A. tridentata and a well-developed biological soil crust community. Poa secunda (J. Presl), Ericameria nauseosa (Pall, ex Pursh) and Artemisia frigida (Willd.) are also common. At approximately 700 to 1000 m in elevation, the BGxw (Very Dry Warm subzone), or the middle grassland, is dominated by P. spicata, and characterized by the absence of A. tridentata. Other common species include P. Secunda, E. nauseosa, A. frigida, and Koeleria macraniha ((Ledeb.) Schult.). The upper grasslands, found 900-1000 m in elevation, are incorporated in the Interior Douglas-fir (IDF) zone, and are dominated by P. spicata, together with Festuca scabrella (Torr. ex Hook. var. major Vasey), in the central parts of the zone, and Stipa richardsonii ((Link) Barkworth) and Festuca saximontana (Rydb.) in the Chilcotin (Meidinger and Pojar 1991). Van Ryswyk et al. (1966) described these changes in vegetation zones as abrupt, and found that they were related to sharp differences in the ratio of precipitation to temperature. Biological Soil Crusts Biological soil crusts (BSCs - also referred to as cryptogamic, microbiotic, or biocrusts) are communities of lichens, bryophytes, free-living cyanobacteria and algae, micro­ fungi, and heterotrophic bacteria closely associated with the soil surface (Belnap 2003). BSC occur in arid and semi-arid ecosystems throughout the world, achieving greatest cover in areas with little vegetation. Much of the knowledge of BSCs in grasslands and shrub-steppe comes from studies in the western United States, while BSC communities in British Columbia have received relatively little attention (Marsh et al. 2006). A number of important ecological functions in arid and semi-arid environments are attributed to BSCs. Some of these functions include promoting soil stability and preventing soil erosion by wind and water (Neff et al. 2005, Malam Issa et al. 2007), improving soil water infiltration and/or storage (Belnap 2006), moderating fluctuations in soil climate (Veluci et al. 2006), influencing nutrient influx via dust entrapment (Williams et al., 2012) and contributing to nutrient cycling and soil fertility (Evans and Lange 2001, Elbert et al. 2012). BSCs have also been found to increase the uptake of certain trace minerals in nearby plants, creating higher quality forage for herbivores (Harper and Belnap 2001) and plants growing on BSCs 4 have been found to have increased height, phytomass and nitrogen content (Langhans et al. 2009). These ecological functions often depend on the species composition and physical structure of the BSC, which are predominately controlled by climate and disturbance history. The diversity of BSCs is generally positively correlated with ecosystem function, but this can depend on the function being studied (Bowker et al. 2010, Maestre et al. 2012). The contribution of BSCs to ecosystem functions is also generally highest in relatively undisturbed crusts with high biomass (Belnap 2003). Biological soil crusts around the world are often dominated by similar species, including the cyanobacteria Microcoleus vaginatus (Vaucher) Gomont, Nos toe commune Vaucher Ex Bomet and Flahault, Schizothrix calcicola (C. A. Agardh) Gomont; the lichens Collema tenax (Sw.)Ach., Fulgensia desertorum (Tomin) Poelt, Psora decipiens (Hedwig) Hofftn, Placidium squamulosum (Ach.) Breuss; and the mosses Syntrichia ruralis (Hedwig), Pterygoneurum ovatum Hedw. Dixon and Bryum spp. (Evans and Johansen 1999). The dominant species composition of BSCs is often determined by the stage of succession. BSC formation is initiated by non-heterocystous filamentous cyanobacteria, often Microcoleus spp. These cyanobacteria initiate the process of stabilization through a cyclic process of binding coarse soil particles and trapping fine atmospheric particles to form alternate layers of fine and coarse sediments (Williams et al. 2012). In addition to stabilizing the soil, this process improves the water holding capacity of the soil and prevents leaching of nutrients, creating a suitable habitat for other BSC organisms to colonize (Yeager et al. 2004). These organisms include free-living cyanobacteria Nostoc spp., Scytonema spp. and cyanolichens Collema spp. that are capable of fixing nitrogen (ie. converting atmospheric nitrogen (N2) into ammonium (NH4+)). Much of this fixed nitrogen is released into the soil, which can be the major source of 5 N in arid and semi-arid environments (Evans and Ehleringer 1993, Evans and Belnap 1999). Frequent or intense disturbances can revert and keep BSCs in earlier successional stages, and the rate of recovery to later successional stages has been found to vary by climate. Other factors that can affect the species composition of BSCs in a given region include soil chemistry, climate and vegetation (Ponzetti and McCune 2001, Root and McCune 2012). The BSC occurring in grasslands and shrub-steppe of interior British Columbia consists of lichens, bryophytes, free-living algae and cyanobacteria. The diversity of lichens and bryophytes along an elevation gradient in the Chilcotin region of central interior British Columbia is described in Marsh et al. (2006). Upper elevation sites have the highest number of species, followed by lower, then middle sites. These differences in species richness may be related to different disturbance histories rather than climatic variables. Grassland sites were also characterized by different BSC species than shrub-steppe sites. Differences in climate, species composition and age affect the microtopography of the BSC, which influence its appearance, as well as the movement and infiltration of materials such as dust, water and seeds (Belnap 2003). Early succession BSCs and those found in hot, hyper-arid deserts, are dominated by cyanobacteria, which create a veiy smooth surface structure. In regions with lower potential evapotranspiration, the dominant species shift to lichens and mosses, which give the BSC a rougher surface. These differences in surface relief can be explained by a greater affinity for dust trapping and accumulation by these organisms compared to cyanobacteria alone (Williams et al. 2012). Repeated expansion and contraction of the BSC surface caused by wet-dry cycles creates an increasingly complex surface topography over time by creating cracks and pores that enhance dust capture, and building pinnacles, towers and pedestals (Williams et al. 2012). The occurence of frost may also affect 6 the microtopography of BSCs, creating a pinnacled surface in areas dominated by cyanobacteria and a rolling surface in areas dominated by lichens and mosses. The grasslands and shrub-steppe of British Columbia are considered a cold, semi-arid environment, and the BSC surface is primarily rolling. This type of surface promotes trapping of seeds and nutrientrich dust particles, soil porosity, formation of soil aggregates, abundance of soil invertebrates and localized water infiltration and retention (Belnap 2003, 2006). BSC organisms span a range of trophic levels, which allows them to contribute to a wide selection of ecosystem functions. Also, unlike many plant communities, BSCs are able to maintain many of these functions during periods of drought. Despite this resilience, BSCs are highly sensitive to surface disturbance caused by grazers and human activity (Eldridge 1998). Such disturbances, particularly when they occur when the soil is dry, can destroy the cohesive structure of the BSC layer. Changes in fire cycles and introduction of annual grasses and other invasive species can also negatively impact BSC abundance, composition and structure (Belnap et al. 2006), which in turn affect the ability of BSCs to perform ecosystem services (Belnap 2003). Full recovery of BSC structure and function can take over a century in cooler regions, and multiple centuries in hotter regions. This combination of high ecological function and sensitivity to disturbance makes BSCs key candidates for studying the overall health of arid and semi-arid ecosystems (Belnap 1996, Ponzetti and McCune 2001). Carbon and Nitrogen Cycling in Grassland and Shrub-steppe Soils The biogeochemical cycling of N is often of interest in terrestrial ecosystems because it is a limiting nutrient for plant productivity (West 1991). Carbon inputs fuel soil food webs and provide the energy needed for many biogeochemical processes. Carbon in the form of soil organic carbon (SOC) also provides many benefits to soil structure. Soil carbon fluxes are 7 important for measuring the ability of different ecosystems to take up and sequester atmospheric CO2 . Grassland soils may be of particular importance for carbon sequestration due to their ability to store large amounts of SOC in the A horizons (Pennock et al. 2011). BSCs have been found to be major contributors to both carbon and nitrogen cycling in arid and semi-arid environments throughout the world (Elbert et al. 2012). In North America, BSCs are responsible for up to 10% of the total fixed carbon and 80% of the total biologically fixed nitrogen in terrestrial systems (Elbert et al. 2012). Fluxes o f Soil C and N The primary inputs of soil carbon in grasslands and shrub-steppe vary with climate, seasonality, and the dominant vegetation. In more mesic and higher elevation environments where grasses dominate, carbon input into the soil occurs via the breakdown of below-ground tissues of grasses, thus creating a soil profile with little to no organic layer and a fairly deep A horizon enriched in SOC (Pennock et al. 2011). In shrub-steppe ecosystems, soil carbon is also added through the breakdown of leaf litter on the surface of the soil. The relative contribution of BSCs to soil carbon increases in more xeric environments as plant cover decreases. In areas that are too hot and dry to support a dense cover of grasses and shrubs, BSCs become the dominant cover and the primary contributor of fixed carbon (Belnap et al. 2001). At times when plants cannot, BSCs are able to fix carbon due to their ability to rapidly regain physiological function upon hydration after desiccation or freezing. The carbon gain via net photosynthesis by BSC organisms is necessary for fueling nitrogen fixation in BSCs. Carbon fixed by BSC organisms can enter the soil via leaching or decaying processes (Lange 2001). The soil carbon released by BSCs is an important step in the process of BSC development and structure, and is also utilized by many heterotrophic organisms in the soil. 8 Losses of soil carbon can occur when respiration exceeds carbon gain via photosynthesis. Soil carbon can also be lost by burning or directly via soil erosion. Increased temperatures tend to lead to losses in soil carbon due to higher respiration rates and a decline in photosynthetic activity if temperatures exceed the optimum for photosynthesis. In arid and semi-arid environments, nitrogen is the most important factor limiting biological activity after carbon and moisture availability (West 1991).The primary inputs of nitrogen into soils are via atmospheric deposition, biological nitrogen fixation, and decomposition of plant and animal residues. N-fixation by BSCs has been found to be the dominant source of N in a number of arid and semi-arid ecosystems throughout the world (Evans and Lange 2001). In B.C. grasslands and shrub-steppe, plants capable of symbiotic Nfixation are not common, thus it is expected that BSCs are also a major source of N in this system. As much as 70% of the nitrogen fixed by BSCs is released to the surrounding soil, mainly in the form of NCb* and small quantities of NH4 +, amides, peptides and free amino acids, becoming available to plants and soil microbes (Belnap 2001). Nitrogen that enters the soil can either be taken up by plants, leached down, lost by soil erosion, or returned to the atmosphere via nitrification, denitrification and ammonia volatilization. Gaseous losses appear to be the most prevalent in BSC associated soils, particularly the production of N 2 , N 2 O and NO2 during nitrification and denitrification, which can account for as much as 80% of fixed N. BSCs can provide ideal environments for these processes following precipitation events due to the high availability of labile nitrogen and carbon caused by cell leakage (Evans and Johansen 1999). Conversely, BSC associated soils in the Colorado plateau have been found to have negligible rates of denitrification due to extremely minimal presence of denitrifiers (Johnson et al. 2007). 9 The natural abundance of the rare stable nitrogen isotope, 15N, can be used to study nitrogen dynamics in an ecosystem (Robinson 2001). This approach takes advantage of the fact that the isotopic composition (81SN) of N in the atmosphere is 0%o (per mil, or part per thousand), and organisms and soil that receive most of their N from fixation will have 815N near 0%o. Measurements of 8I5N have been used in this way to track the different pathways of N-input into a desert ecosystem (Russow et al. 2004). Chemical reactions involving N result in fractionation, as I5N is more energetically expensive to break or form chemical bonds with than 14N. Thus soils where more N losses occur over time than N inputs will have a higher S15N, which has been found on disturbed soils in arid and semi-arid envrionments (Evans and Ehleringer 1993, Frank and Evans 1997, Evans and Belnap 1999, Aranibar et al. 2003).Values of ,5N natural abundance that are within the range of -2%o to 2%o have been associated with significant atmospheric inputs ofN (Evans and Lange 2001). Distribution o f Soil C and N The spatial distribution of nutrient availability and cycling in semi-arid grasslands and shrub-steppe can be influenced by several interacting factors including landscape position (Schimel et al. 1985, Verchot et al. 2002), vegetation type, microclimate and the physical properties of the soil (Hook and Burke 2000). These interacting factors can be quite complex and are drawn from studies in many different environments. Here I review how these factors relate to the grasslands and shrub-steppe associated with biological soil crusts in our study area. Aird and semi-arid envrionments that are largely limited by moisture availability tend to have a discontinuous aboveground cover of vegetation. This discontinuous cover has been found to lead to the formation of resource islands, where nutrients and soil material 10 accumulate around plants, creating a heterogeneous distribution of soil nutrients. Shrubs and grasses assist in resource island formation by the accumulation of litter from their above and belowground tissues, and by acting as deposition sites for physical material redistributed by erosion from plant interspaces. The resource island effect can be greatly counteracted by the presence of well-developed BSCs (Housman et al. 2007). Biological soil crusts containing later successional species of lichen and cyanobacteria allow a more homogenous distribution of nutrients by directly fixing C and N, preventing soil loss and subsequent nutrient depletion in the interspaces, and by creating habitat for soil fauna. The presence of BSCs can also influence soil texture by trapping and holding particles in the silt/clay fractions through a combination of surface roughness and the sticky cyanobacterial sheath (Belnap et al. 2001, Williams et al. 2012). The negatively-charged silt/clay particles bind to plant essential nutrients, increasing their availability relative to uncrusted soil. The presence of grazers and the timing and density of grazing can also have a strong effect on biogeochemical cycles. In the inter-mountain west, shrub-steppe is often characterized by a clumped distribution of shrubs and grasses and a well-developed cover of BSCs in the interspaces between plants (Mack and Thompson 1982). The evolutionary history of this shrub-steppe includes very limited grazing by smaller ungulates (i.e. deer, sheep), mostly during the winter. Livestock grazing in these ecosystems generally reduces C and N input through changes in the species composition, cover and physiological activity of BSCs. The decline in BSC cover also leads to increased losses of soil C, N and other nutrients due to an increase in soil erosion by wind (Neff et al. 2005) and a decline in the physiological activity of BSC organisms coupled with increased gaseous losses of N (Evans and Belnap 1999). 11 Biological N2 Fixation The biological fixation of nitrogen is facilitated by the nitrogenase enzyme complex, which is only found in certain prokaryotes. In biological soil crusts, most of the nitrogen fixing organisms are free-living and lichenized cyanobacteria, including the genera Nostoc and Scytonema (Belnap 2001). Nitrogenase is highly sensitive to oxygen and N-fixing cyanobacteria must contain the enzyme within specialized cells called heterocysts that shelter the reaction from O2 . Other N-fixing associations in BSCs include cyanobacteria living epiphytically on mosses and chlorolichens (Belnap 2001). The rate of nitrogen fixation can be measured directly by using I5N2, or indirectly using acetylene reduction analysis (ARA). ARA measures the rate at which the nitrogen fixing enzyme nitrogenase reduces acetylene (C2H2) gas to ethylene (C2H4 ), and provides an index of nitrogenase activity (Stewart et al. 1967). ARA is preferred rather than measures of direct N2 input due to its ease of use and low cost. ARA provides a good measure of relative nitrogenase activity, but must be calibrated to rates of direct 15N2 incorporation in order to accurately estimate actual N-inputs (Evans and Johansen 1999, Belnap 2002). The theoretical conversion ratio of ethylene production to N2 incorporation is 3:1, but has been found to vary from 0.02 to as high as 56 (Belnap 2001). Conversion ratios can vary by species associations, incubation temperatures and seasonality (Liengen 1999a). Factors influencing N-fixation rates It is important to understand the primary factors that influence rates of nitrogen fixation in BSCs, as these rates have been found to vary both spatially and temporally. A greater understanding of these factors can help explain variation and more accurately estimate N inputs across the landscape. The dominant organisms present in BSCs influence the rates of 12 nitrogen fixation at optimal conditions (Belnap 2002). Light cyanobacterial crusts dominated by Microcoleus, have the lowest rates of nitrogen fixation. Dark cyanobacteria crusts dominated by Nostoc and Scytonema have intermediate rates of nitrogen fixation, while BSCs dominated by Collema cyanolichens have the highest rates. The increase in nitrogenase activity found between early and late successional stages of BSCs can be explained by an increase in biomass of Nostoc and Scytonema (Housman et al. 2006). Nitrogenase activity in biological soil crusts has been found to be highly dependent on a number of abiotic factors, including temperature and moisture levels, and availability of C and N (Evans and Johansen 1999, Belnap 2001, 2002, Hartley and Schlesinger 2002). These controlling factors are hierarchical, with moisture being the ultimate determinant of nitrogenase activity. BSCs are adapted to long periods of drought, and are inactive when dry, but able to quickly resume physiological activity upon hydration (Evans and Johansen 1999). The effects of moisture content on nitrogenase activity is highly dependent on species, habitat and pre-experimental conditions (Belnap 2002). Long periods of desiccation can increase the recovery time for nitrogenase activity. Prolonged hydration may also reduce N-fixation rates, possibly due to losses of glucose that eventually lead to a depletion of energy reserves. Alternating periods of wet and dry soils may lead to greater overall rates of nitrogen fixation than prolonged periods of hydration (Belnap 2002). Different species also have different moisture requirements for activation of physiological activity (moisture compensation point), with cyanolichens having much higher moisture compensation points than cyanobacteria (Lange 2001). The secondary factor controlling nitrogenase activity is the amount and accessibility of photosynthetic products (Belnap 2002). Nitrogen fixation is very energetically expensive; any 13 factors that affect recent and current photosynthetic rates will also affect potential nitrogenase activity. Conditions that result in high respiration rates relative to photosynthesis can deplete available carbon stores and increase the recovery time for nitrogenase activity (Coxson and Kershaw 1983a). Moderate temperatures and long periods of moisture in the crust allow maximum nitrogenase activity due mostly to high carbon stores from prolonged photosynthetic activity (Belnap 2002). Some carbon stores are initially lost upon rehydration and there is a lag time for photosynthesis to catch up with respiratory carbon losses. As light intensity can affect rates of photosynthetic activity, it in turn can affect rates of nitrogenase activity. The effect of light intensity on nitrogenase activity may depend on species and habitat, as reported levels of light saturation vary. In general, most BSCs reach light saturation at relatively low light levels, and do not show any decline with increasing light (Belnap 2001). Optimal temperatures for nitrogenase activity are between 20-30°C for BSCs from temperate, tropical and polar regions (Coxson and Kershaw 1983b, Liengen and Olsen 1997, Belnap 2002). Regional differences in temperature optima for physiological activity may reflect the timing and temperature conditions of moisture availability. Lower temperature optima for N-fixation have been reported for polar environments (Liengen 1999b, Stewart et al. 201 lb) where high soil moisture coincides with cooler spring and summer temperatures, and higher optima have been reported for temperate regions in which most of the precipitation falls in the summer and coincides with higher temperatures (Zhao et al. 2010). Temperatures above the optimum tend to lead to a rapid decline in nitrogenase activity. The minimum temperature for nitrogenase activity for most BSCs studied is 0°C, however, nitrogenase activity has been recorded as low as -7.6°C (Belnap 2001). Recovery of nitrogenase activity 14 after freezing depends on conditions before and after freezing, and a lag time occurs after freezing. Light exposure before freezing reduces recovery time. Surface disturbances have been found to inhibit nitrogen fixation even without a corresponding loss of BSC biomass. In light cyanobacterial (Microcoleus dominated) crusts, this may occur due to a disruption in soil aggregates that results in a loss of anaerobic micro­ sites for non-heterocystic N-fixation to occur (Belnap 1996). N-fixation by dark cyanobacterial (Scytonema and Nostoc dominated) and lichen dominated crusts may be less sensitive to surface disturbance because they are not as reliant on aggregation for nitrogenase activity. Surface disturbances do disrupt N-fixation over the long-term in these crusts by promoting burial and wind erosion of the existing BSC surface. In the longer term, grazing under even very light densities was found to reduce nitrogenase activity of Collema and dark cyanobacteria in BSCs in a Mongolian steppe (Liu et al. 2009). The recovery of BSC function following grazing can take decades, as depressed nitrogenase activity was found in BSCs 30 years after the cessation of grazing. Heavier grazing intensities were found to reduce overall N input through a combination of decreased abundance of N-fixing lichens and cyanobacteria, and decreased nitrogenase activity of these organisms (Ponzetti and McCune 2001, Housman et al. 2006). Biological Soil Crusts and Climate Change in Grasslands and Shrub-steppe Across Canada, temperatures have increased by 0.5°C to 1.5°C over the last century, while precipitation has also increased, particularly during the winter months (Zhang et al. 2000). Mean annual temperatures in western North America are expected to increase by another 2-6°C over the next 100 years, accompanied by a general trend of increased annual 15 precipitation in the north (IPCC W GI2007). Seasonal precipitation patterns are also expected to change, with less precipitation in the summer months, and more in the winter months. Historical weather data show that mean annual temperatures in the Cariboo-Chilcotin region have increased by approximately 1°C over the last century, and the rate of warming has increased over the latter part of the century (Dawson et al. 2008). Mean winter temperatures are expected to be the most impacted by climate change in this region, with a projected increase of 2.5°C to 3.5°C over the next 50 years. Patterns of precipitation are expected to change with predictions of increased in mean winter precipitation, but less snow accumulation and earlier snowmelt, as well as decreased summer precipitation and increased evaporative demand (Dawson et al. 2008, Spittlehouse 2008). Potential species ranges may expand or be reduced northward and upward in elevation, and grasslands in particular may expand as climate becomes less favorable for tree regeneration. The potential range of the bunchgrass zone has been predicted to expand northward and upwards in elevation by over 400% in area by 2055, while the hottest and driest parts of the bunchgrass zone will have climates more similar to shrub-steppe in the northwest United States (Hamann and Wang 2006). The actual response of ecosystems to these shifts in climate may be quite complex, particularly in a region as topographically and climatically variable as British Columbia (Kimmins and Lavender 1992). Changes in climate are also expected to result in increased frequency of extreme events such as drought, heat waves and heavy precipitation (Walker and Sydneysmith 2008). Sustainable management of ecosystems must include an understanding o f how systems respond to change (Nitschke and Innes 2008). Grasslands and shrub-steppe in B.C. are managed for multiple values, including rangeland and wildlife habitat. Climate change poses 16 risks to these values, such as increased risk of fires, and reduced ability for regeneration post­ disturbance. As BSC are an important component of the bunchgrass zone, understanding their response to climate change is needed for assessing potentail future risks and vulnerabilities. The bunchgrass zones of B.C. represent the northern range of shrub-steppe and grassland of intermountain western North America and climate change is expected to have the greatest impact at higher latitudes and altitudes, thus it is especially important to increase the knowledge of climate vulnerabilities in this region. Increased temperatures are expected to lead to declines in soil carbon and C:N ratios due to increases in microbial activity and decomposition (Rosenzweig and Hillel 2000). Both nitrification and denitrification rates may also increase, leading to greater losses of soil nitrogen. Decreased soil moisture may also lead to declines in primary productivity and nitrogen fixation, leading to further net losses of soil C and N. A reciprocal transplant experiment in a Washington shrub-steppe found that a shift to a wanner and drier climate reduced total soil carbon by 32% and nitrogen by 40% over a period of five years (Link et al. 2003). The majority of these losses were from the particulate organic matter fraction, suggesting that climate change may also negatively impact soil structure and aggregate stability, further accelerating soil nutrient losses. The effects of increased temperature on biological soil crusts is not clear in the literature. The presence and abundance of Collema has been found to shift dynamically between years, and was negatively correlated with summer maximum temperatures and spring minimum temperatures (Belnap et al. 2006). Conversely, an experimental 2-3°C increase in soil temperature over a period of 2 years found little change in BSC cover, composition and function with warming alone (Johnson et al. 2012, Zelikova et al. 2012). 17 Warming in combination with increased precipitation frequency led to significant declines in cyanobacterial abundance and biomass (Johnson et al. 2012) and a decline in photosynthetic performance of lichens, particularly Collema (Zelikova et al. 2012). Belnap et al. (2004) found that a shift to shorter, more frequent precipitation events led to a decline in photosynthetic performance, nitrogenase activity and the ability to produce protective pigments. Shorter hydration times can cause a net carbon loss in BSC organisms as physiological activity takes time to overcome the losses caused by re-wetting. Prolonged carbon losses lead to an inability to produce protective pigments, which can lead to increased mortality of BSC organisms. Lichens and dark cyanobacterial BSCs are more sensitive to both carbon losses from short precipitation events as well as damage by solar insolation than are light Microcoleus dominated BSCs. This suggests that any changes in climate that lead to shorter periods of BSC hydration can cause a shift in BSC organisms from late successional lichen and dark cyanobacteria to more early successional Microcoleus (Belnap et al. 2004). One effect of climate change that may be particularly important in B.C. grasslands and shrub-steppe are changes that occur during winter months. Reduced snow cover as a result of higher winter temperatures was found to increase the photosynthetic activity of Collema (Zelikova et al. 2012). Physiological activity of soil cyanobacteria can resume rapidly in the day during freeze-thaw cycles if sufficient light is available for photosynthetic activity (Coxson and Kershaw 1983b). Thus, a decline in snow-cover in combination with increased precipitation and warmer temperatures in the winter months may lead to an increase in BSC activity and abundance. What is not known is whether these gains in the winter period could offset the losses caused by a hotter and drier summer period. 18 RESEARCH OBJECTIVES The purpose of this research is to develop a more comprehensive understanding o f the relationship between climate and nitrogen inputs by biological soil crusts in British Columbia grasslands and shrub-steppe. In chapter one, we will investigate variation in nitrogen and carbon of BSCs and associated soils in relationship to gradients in temperature and moisture availability that occur across an elevation gradient in Lac du Bois Grasslands, B.C. (see objective 1). In chapter two, we will examine patterns o f nitrogen fixation activity in relation to changes in moisture, light and temperature availability, and model nitrogen fixation rates throughout the year at the microclimate scale in BSCs from Farwell Canyon, B.C (see objective 2). Objective 1: a) Compare the availability of nitrogen, carbon, and the isotopic composition of nitrogen in the BSC to that of the underlying soil in order to determine the importance of BSCs to soil nitrogen and carbon inputs. b) Examine spatial patterns of changes in nitrogen and carbon content in the BSC and underlying soil across a gradient of temperature and moisture availability related to elevation. Objective 2: a) Evaluate the response of ^.fixation by biological soil crusts from a typical shrubsteppe community to changes in moisture, temperature and light availability. 19 b) Model N2-fixation rates throughout the year based on the above response using measurements of BSC- level moisture, temperature and light, and use these rates to estimate landscape-level N inputs. c) Evaluate the potential effects of future climate change on N-inputs by BSCs in British Columbia shrub-steppe. 20 CHAPTER 1: NUTRIENT COMPOSITION OF BIOLOGICAL SOIL CRUSTS AND ASSOCIATED SOILS ACROSS AN ELEVATION GRADIENT IN LAC DU BOIS GRASSLANDS, BRITISH COLUMBIA INTRODUCTION Grassland and shrub-steppe ecosystems in British Columbia occupy less than 1% of the province, but are home to an exceptional diversity of flora and fauna, including over 30% of threatened and/or endangered species (GCC 2004). These ecosystems have greatly decreased over the last century and are currently threatened by forest encroachment, agricultural and urban expansion, and overuse by livestock grazing and recreational activities. Due to their importance and threatened status, there is a growing need to better understand the impacts of climate change on grasslands and shrub-steppe in B.C. Temperature and precipitation in Western Canada have significantly increased over the last century, particularly in the winter and spring (Zhang et al. 2000, Vincent and Mekis 2006). Mean annual temperatures in western North America are expected to increase by another 26°C over the next 100 years, accompanied by a trend of increased annual precipitation in the north (IPCC 2007). Seasonal precipitation patterns are also expected to change, with less precipitation in the summer months, and more in the winter months. Potential impacts of climate change on grassland and shrub-steppe ecosystems in B.C. range from expansion into currently forested areas, changes in soil nutrient cycling leading to declines in soil carbon and nitrogen, and environments that are more favorable for establishment of invasive exotic grasses (Spittlehouse 2008). Understanding the potential ecological response and vulnerability to climate change can help remove uncertainty when managing ecosystems (Nitschke and Innes 2008) 21 An important component of many grassland and shrub-steppe ecosystems are biological soil crusts (BSCs) - communities of lichens, mosses, free-living cyanobacteria and algae intimately associated with the soil surface. BSCs are important sources of soil carbon and nitrogen in many arid and semi-arid ecosystems (Evans and Johansen 1999, Elbert et al. 2012) and also play roles in improving soil stability and structure, preventing erosion by wind and water, and improving soil water infiltration and retention (Belnap 2003). BSC communities are useful model systems for studying soil ecological processes due to their multiple ecosystem roles and high sensitivity to disturbance (Bowker et al. 2010). BSCs are important components of grassland and shrub-steppe of Western Canada, but relatively little research has been conducted on BSC communities in this area (Marsh et al. 2006). The natural abundance of 1SN (81SN) is a useful tool for assessing ecosystem nitrogen dynamics, and has been used in a number of ecosystem studies with BSCs (Evans and Ehleringer 1993, Evans and Belnap 1999, Aranibar et al. 2003, Russow et al. 2004). Systems that rely solely on atmospheric fixation for nitrogen will have 815N values close to that of atmospheric (<0%o); values of 8iSN ranging from -2 to 2%o are indicative of significant input via N-fixation (Evans and Lange 2001). The physiography of Lac du Bois grasslands, located near Kamloops, B.C., where grass and shrub communities occur across an elevation range of approximately 300-1200 m, provides a unique opportunity to study the contribution of BSCs to soil nutrient cycling across a climatic gradient of temperature and precipitation. The use of elevation gradients as a proxy for climate change has been applied in a number ecological studies, including several grassland systems (Dahlgren et al. 1997, Smith et al. 2002, Link et al. 2003). The purpose o f this study was to (1) compare the availability of nitrogen, carbon, and the isotopic 22 composition of nitrogen in the BSC to that of the underlying soil in order to determine the importance of BSCs to soil nitrogen and carbon inputs; and (2) examine spatial patterns of changes in nitrogen and carbon content in the BSC and underlying soil across a gradient of temperature and moisture availability related to elevation. METHODS Study Site The study area is located 6 km northwest of Kamloops, British Columbia, Canada, within Lac Du Bois Grassland Provincial Park (50.75° N, 120.4° W). The park is approximately 15,000 hectares in size and supports shrub-steppe, grassland and interior Douglas-fir forest on gently sloping, south-facing terrain. The climate in the area is semi-arid and mild continental, with a mean annual temperature of 8.9°C and mean annual precipitation of 279 mm. Average monthly temperatures at the Kamloops airport (345 m a.s.l.) range from -4.2°C in January to 21°C in July (Environment Canada 2012). Most of the summer precipitation falls during thunderstorms throughout June and August, and most of the winter precipitation falls as snow in December and January (Tisdale 1947). The driest months are March and April. Climate becomes cooler and wetter with increasing elevation (Table 1.1). The study site can be classified in three distinct vegetation zones, which are related to sharp differences in the ratio of precipitation to temperature with elevation (Van Ryswyk et al. 1966). Below 700 m, the BGxh (Very Dry Hot subzone), is a shrub-steppe dominated by P. spicata, A. tridentata and a well-developed biological soil crust. Although this zone is a shrub-steppe, it is commonly referred to as the lower grassland, as described in Van Ryswyk 23 et al. (1966). At approximately 700 to 1000 m in elevation, the BGxw (Very Dry Warm subzone), or the middle grassland, is dominated by P. spicata, and characterized by the absence of A. tridentata. Other common species include P. Secunda, E. nauseosa, A. frigida, and K. macrantha. The upper grasslands, found 900-1000 m in elevation, are incorporated in the Interior Douglas-fir Very Dry Hot variant (IDFxh2), and are by dominated P. spicata and F. scabrella (Meidinger and Pojar 1991). The respective soils associated with these grasslands are Brown, Dark Brown, and Black Chernozems that developed on parent material composed of glacial till. A common feature to most of the soils in this region is a silty aeolian veneer, on which most of the Ah horizons have formed (Fenger 1982). Table 1.1 Summary of climate normals at Kamloops Airport (Environment Canada 2012) and climate estimates at different elevations in Lac du bois Grasslands, British Columbia1". Location Elev. Temperature (°C) Site (Period) Precipitation (mm) (m) Mean Mean Mean Mean Precip. as (lat., long.) annual Jan. July annual snow (cm) 345 8.9 -4.2 21 279 Kamloops 50°42'08.00" N, 75.5 120°26'31.00" W Airport (1971-2000) 380 7.6 ^.8 19.2 291 54 Lac du Bois 50°43'03.07" N, 120o26'46.72"W (1971-2000) 50°45T6.09"N, 722 6.5 -5.4 18 369 78 120o27'16.56" W 50°47'22.56" N, 16.4 392 917 5.1 -6.3 103 120°26'49.27" W •Temperature and precipitation normals for 1971 - 2000 from ClimateWNA modelling using Wang et al. (2012; see http://www.genetics.forestry.ubc.ca/cfcg/ClimateWNA/ClimateWNA.html). 24 Figure 1.1 Views of the upper elevation grasslands (top left), lower elevation shrub-steppe (bottom left) and the representative biological soil crust (right) at the Lac du Bois Grasslands, British Columbia (Photos - P. Sanborn). Field Methods A digital elevation model of the park (C. Bulmer, B.C. Ministry of Forests and Range, Research Branch) was used to randomly select 50 sites along south-east to south-west facing slopes within the park boundaries. Sites were limited to grass dominated biogeoclimatic zones (BGxh, BGxw, IDFxh2) and slopes less than 30%. Sampling was conducted over a period of two days in June 2010. Sites ranged from an elevation of 380 m to 917 m. A 25 sample of the BSC and 10 cm of underlying mineral soil, consisting of the Ah horizon, were collected from each site, and the elevation, slope, aspect, dominant vegetation and approximate crust cover were also recorded. At each site, BSCs exhibiting the least amount of surface disturbance were selected from interspaces between vegetation. Samples were air-dried, sieved (<2 mm) and sent for analysis of total nitrogen, total carbon and mineralizable nitrogen (Ministry of Forests and Range, Analytical Chemistry Section, Victoria, B.C., Canada). A small amount (~30 g) from each sample was finely ground with a ball mill and sent for analysis of 15N natural abundance (Stable Isotope Facilities, University of Saskatchewan, Saskatoon, SK, Canada). Data Analysis Differences in chemical parameters between the crust and underlying soil were assessed using a paired t-test. The sampling points were grouped based on the divisions between lower, middle and upper grasslands described in Van Ryswyk et al. (1966). A twoway ANOVA was conducted to detect significant (p<0.05) differences in measured parameters across elevation and between the BSC and soil. Pair-wise multiple comparisons were conducted using the Holm-Sidak method. Statistical analyses were conducted in SigmaPlot 11.0 (Systat Software, Inc., San Jose CA, USA). RESULTS The BSC layer had significantly (p<0.05) higher total carbon, total nitrogen and mineralizable nitrogen than the underlying (l-10cm) mineral soil (Table 1.2). The 8I5N natural abundance in the BSC ranged from -0.38 to 4.2596o and was significantly lower compared to the underlying soil. Total carbon and total nitrogen were approximately twofold 26 higher, while mineralizable nitrogen was almost 6 times higher in the BSC relative to the mineral soil. Table 1.2: Selected Chemical Properties of Biological Soil Crusts and underlying mineral soils at Lac du Bois Grasslands, British Columbia. The mean differences reported are significantly different based on a paired two-sample t-test (P<0.05) Diff P Parameter Soil BSC (±SE) mean (±SE) mean (±SE) mean (0.70) 2.23 2.73 (0.332) <0.0001 Total C (%) 4.97 (0.15) (0.027) 0.186 (0.018) <0.0001 Total N(%) 0.361 (0.012) 0.175 (0.240) <0.0001 C:N 13.25 (0.27) 1.32 11.93 (0.13) Mineralizable N (mg/kg) 266.6 (18.6) 47.55 219.04 (16.22) <0.0001 (4.03) (0.16) 2.91 (1.65) 5ISN Natural abundance (%o) 2.24 5.15 (0.20) <0.0001 Total carbon, nitrogen and mineralizable nitrogen were also found to increase significantly (P<0.05) with elevation in both the BSC and the underlying mineral soil (Figure 1.2). The C:N ratios were consistent across elevations, and were significantly lower in the soil compared to the crust, except at the lowest elevations (Figure 1.2). The 5,5N natural abundance was not found to be related to elevation in either the BSC or the soil (Figure 1.2). The proportion of mineralizable N within the total N pool was significantly higher (P<0.001) in the BSC compared to the soil across all elevations, and also increased significantly (PO.OOl) with decreasing elevation in the BSC. 27 Soil (1-10 cm) N 200 2 100 <600 m 600-800 m 800-1000 m Elevation Figure 1.2: Total Carbon, Total Nitrogen and Mineralizable Nitrogen in the biological soil crusts (BSC) and underlying soil (l-10cm) across an elevation gradient in Lac du Bois Grasslands, British Columbia. Error bars represent standard error and letters denote significant differences (p<0.05) across soil type and elevation class. 28 Soil (1-10 cm) <600 m 600-800 m 800-1000 m Elevation Figure 1.3: Ratio of Total C:N and I5N natural abundance in the biological soil crusts (BSC) and underlying soil (l-10cm) across an elevation gradient in Lac du Bois Grasslands, British Columbia. Error bars represent standard error and letters denote significant differences (p<0.05) across soil type and elevation class. DISCUSSION The average total C, total N and mineralizable N in both the crust and underlying soil were all within the range of that reported for Chilcotin grassland and shrub-steppe ecosystems some 200 km north of the present study site (Marsh et al. 2006). The twofold enrichment of nitrogen in the BSC suggests that that the cyanobacteria and cyanolichen components contribute to soil nitrogen content. Many BSC species have been found to release N29 compounds, particularly in the form of nitrate and ammonia into the surrounding soil (Belnap et al. 2001). In addition to their ability to fix nitrogen, BSCs may also be able to capture and immobilize exogenous nitrogen that could otherwise be lost from the ecosystem (Hawkes 2003). Similarly, BSC organisms can contribute to soil carbon through the release of extracellular carbon compounds, which can represent up to 50% of the total fixed carbon (Belnap et al. 2001). The much higher concentrations of mineralizable nitrogen in the BSC relative to the underlying soil also suggests that substantial release of labile nitrogen is occurring in these ecosystems. Furthermore, the mineralizable N comprised 8.1% of the total N pool in the BSC compared to only 2.4% in the underlying soil. This suggests that a higher proportion of nitrogen within biological crusts is in an active form that is readily available for plant and microbial uptake. Also, while BSCs at the lower elevation grasslands had smaller nitrogen pools than at higher elevation grasslands, they had the highest proportion of mineralizable nitrogen within that pool (~10.8%) indicating that the relative proportion of biologically active nitrogen increases with decreasing elevation. This change in availability only occurs in the BSC, not in the underlying soil. Values of 815N ranging from -2 to 2%o are indicative of significant input via Nfixation (Evans and Lange 2001). The average 5,SN value for the BSCs in Lac du Bois Grasslands (2.2496o) is slightly higher than this range, similar to that reported for BSCs in the Chilcotin region of British Columbia (Marsh et al. 2006) and for disturbed cyanobacterial crusts in Utah (Evans and Belnap 1999). This suggests that N-input via fixation is important, but may not be the only source of N for BSCs in this ecosystem. The higher 8,SN in some sites may also be due to higher rates of N losses compared to inputs via fixation. Processes 30 such as ammonia volatilization and denitrification are subject to fractionation, meaning that ,4N is favoured over l5N during the transformation. This results in an enrichment of 15N in the remaining N-pool, and higher than expected 5,5N. The increase in soil nitrogen and carbon with elevation is likely related to climatic gradients of temperature and precipitation (Van Ryswyk et al. 1966, Smith et al. 2002). Lower elevation sites have lower annual precipitation and higher average temperature, resulting in overall higher potential evapotranspiration. Lower soil carbon at lower elevations can be attributed to increased soil respiration and soil carbon turnover due to higher temperature, or to decreased allocation to below-ground biomass in plants (Link et al. 2003). Lower soil N at lower elevations may be due to higher rates of gaseous losses caused by precipitation on dry soil, as well as a higher N turnover in the litter (Smith et al. 2002). Lower crust and soil N may also be due to lower rates of N-fixation in the BSC as higher temperatures can lead to the decline of cyanolichens and heterocystous cyanobacteria (Belnap et al. 2004, 2006). The 51SN showed large variability in both the BSC (-0.38 to 4.25%o) and soil (0.20 to 7.23%o), but we were unable to explain this variability using elevation or any of the other measured variables. Factors unrelated to climate may affect the I5N natural abundance of the BSC and soil. Cattle are grazed at a low density throughout Lac du Bois Grassland park. The presence of grazers may increase soil 1SN abundance by facilitating N-losses via volatilization and nitrification in urine and dung patches (Frank and Evans 1997). Furthermore, soil trampling by grazers can stimulate N-mineralization, increasing losses of inorganic N by leaching, ammonia volatilization, and/or denitrification. Another factor that may affect 8,SN in the BSCs among sites is the influence of plant litter in areas with dense grass or shrub cover. BSCs growing under shrubs have been found to higher 815N due to an increased incorporation of N from leaf litter rather than from N-fixation (Billings et al. 2003). In order to better understand this spatial variation in 8ISN future studies of BSC nutrient cycling in this area should take into account present and historical disturbance, as well as the influence of surrounding vegetation. Our results suggest that biological soil crusts contribute to soil nitrogen and carbon across all elevations in this ecosystem. While soil C and N within BSCs is greater at higher elevations, their overall contribution to soil nutrients is much lower, due to lower crust cover relative to vegetation. A recent analysis of vegetation cover at Lac du Bois grasslands found that soil lichen cover declined from an average of 25.1% (±17.51) at lower elevations to 2.18% (±2.81) in the upper grasslands (Lee 2011). Cyanobacterial crusts were not measured, but bare soil cover (part of which may be cyanobacterial crusts) declined with increasing elevation from 17.59% (±7.87) to 0.93% (±1.19). It is also important to note that although our upper elevation sample sites were selected within the biogeoclimatic zone representing the upper grasslands (IDFxh), we only selected sites with a southern aspect, which are warmer and drier than sites representative of this zone. Thus the results of this study may not fully apply to conditions typical of the upper grasslands. Current climate models predict an increase in temperature in this region of at least 2°C over the next 70 years. This increase in temperature would lead to an increase in evaporative demand as summer precipitation is also expected to decrease (Spittlehouse 2008). The difference in mean annual temperature between our lowest and highest sites is 2.5°C. 32 This suggests that the soils, vegetation and BSC communities in the upper sites could eventually resemble those in the lower sites. Similar to studies o f other elevation transects in grasslands and shrub-steppe, we may expect both soil C and N to decrease as respiration and denitrification increase and productivity decreases (Smith et al. 2002, Link et al. 2003). At the lower elevations, species composition may shift to a greater abundance cyanobacteriadominated BSCs, as lichens, particularly the cyanolichen Collema have been found to be the most sensitive to increases in temperature and declines in precipitation (Belnap et al. 2004, Belnap 2006). 33 CHAPTER 2: SEASONAL PATTERNS OF NITROGEN FIXATION IN BIOLOGICAL SOIL CRUSTS FROM BRITISH COLUMBIA’S CHILCOTIN REGION. INTRODUCTION Biological soil crusts (BSC) are communities of lichens, mosses, free-living cyanobacteria and algae found intimately associated with the soil surface in arid and semiarid ecosystems throughout the world. BSCs have been found to improve soil stability and structure and prevent erosion by wind and water (Neff et al. 2005, Malam Issa et al. 2007), improve soil water infiltration and retention (Belnap 2006) and are important contributors to soil carbon and nitrogen cycles (Evans and Johansen 1999, Elbert et al. 2012). Certain cyanobacteria (i.e. Nostoc, Scytonema) and lichens (i.e. Collema spp.) found in BSC communities are also able to fix atmospheric nitrogen. Much of this fixed nitrogen is released into the surrounding soil and can be the dominant source of nitrogen in many arid and semi-arid ecosystems (Evans and Ehleringer 1993, Evans and Belnap 1999). Seasonal differences in nitrogen fixation by BSCs can be attributed to differences in temperature, moisture and light availability. The most important factor controlling nitrogen fixation is moisture availability, as physiological activity ceases when moisture is not available. Upon hydration, most BSC organisms are able to rapidly resume physiological function. The rate of recovery of nitrogen fixation is largely dependent on the availability of carbon stores, which are related to the environmental conditions prior to desiccation as well as the conditions dining the recovery period (Coxson and Kershaw 1983a). Rates of nitrogenase activity have been found to be positively related to light intensity, but have been found to achieve maximum rates at relatively low light levels (Belnap 2001). After moisture and 34 carbon availability, the next most important factor controlling rates of nitrogenase activity is temperature. Optimum nitrogenase activity generally occurs between 20 and 30°C, and will cease when temperatures are too low or too high (Evans and Johansen 1999). Most of the BSC studies in North America have focused on the shrub-steppe and grasslands of the Columbia and inter-mountain basins. The grasslands and shrub-steppe of British Columbia represent the northern extent of those ecosystems further south and show many similarities in community composition of BSCs (Marsh et al. 2006). There are however, some differences in climate and precipitation patterns within these regions (Tisdale 1947). Thus, BSCs in British Columbia may have different estimates and seasonal patterns of nitrogen fixation than those already established for North America. The acetylene reduction assay (ARA) is a commonly used method for measuring nitrogen fixation potential due to its relatively low cost and ease of use compared to the more direct measurement using 15N incorporation (Evans and Johansen 1999, Belnap 2001). It works on the principle that nitrogenase will readily accept acetylene and convert it to ethylene. The theoretical conversion of 3 mol of acetylene assimilation to 1 mol of N 2 incorporation is often used to estimate nitrogen inputs from ARA. The true conversion ratio however, has been found to vary among species and environmental conditions, and conversion factors specific to the study should be determined by comparing ARA rates to rates of 15N incorporation under the same conditions (Stewart et al. 1967). The purpose of this study was to measure the response of nitrogen fixation activity by BSCs to changes in temperature, light and moisture. This response was then used to estimate nitrogen inputs by BSC communities based on a two-year dataset of BSC microclimate in a 35 shrub-steppe in the Chilcotin region of British Columbia, Canada. Estimates of nitrogen inputs can be used to determine the relative importance of BSCs to nitrogen cycling in these ecosystems and help develop management strategies under a changing climate regime. METHODS Study Site The main study site was located at Farwell Canyon (51°49'27.35" N, 122°32'46.00" W), a shrub-steppe located along the Chilcotin River approximately 50 km southwest of Williams Lake, British Columbia. This site was located within the very hot dry subzone of the British Columbia Ecological Classification bunchgrass zone (BGxh) (Steen and Coupe 1997). The climate in this part of the bunchgrass zone is characterized by warm to hot, dry summers and moderately cold winters with most of the precipitation occurring in December to January and in June (Table 2.1). The nearest long-term climate data set is from the Williams Lake airport, 40 km northeast of the study site. Climate B.C. modelling (Wang et al. 2012) suggests that the Farwell Canyon site is drier (348 mm) than Williams Lake (450 mm) and warmer (Table 2.2). Table 2.1. Climatic characteristics of the very hot dry bunchgrass biogeoclimatic subzone (BGxh3) for the study site (Steen and Coupd 1997). Precipitation (mm) Temperature (°C) Mean annual Wannest month Coldest Month Frost free days 5.9 19 -10.6 182 Mean annual Mean summer Mean winter Precipitation as snow (cm) 330 177 153 71 Study site topography consists of a series of terraces leading down towards the river, dissected by steep valley slopes. Soils at the site are Brown Chernozems that have formed on 36 sandy fluvial and glaciofluvial parent materials covered with an aeolian veneer (10-50 cm thick) of silt and fine sand (Steen and Coupe 1997). The area selected for BSC sampling and microclimate measurements was on a lower terrace (515m elevation) with a slight southfacing aspect. Although the study site was located on privately owned rangeland, the area has seen relatively little physical disturbance, as most grazing by cattle occurs on the upper terraces. Table 2.2. Summary of climate normals at Williams Lake Airport (from Environment Canada 2012) and at Farwell Canyon, British Columbia (estimated from Climate B.C. modelling using Wang et al. (2012). Site (Period) Location (lat., long.) Elev. (m) Farwell Canyon (1971-2000) 51°49'27.35" N, 122°32,46.00" W 515 Temperature (°C) Precipitation (mm) Mean Mean Mean Mean Precip. as annual winter summer annual snow (cm) 5.2 -1.0 348 14.0 72 Williams Lake (1971-2000) (2009-2010) (2010-2011) 52°10'59.00" N, 122°03'15.00 W 940 4.2 -1.9 12.7 450 193 4.9 3.7 -0.6 -2.5 12.6 12.3 348 526 117 162 The site vegetation was dominated by A. tridentata, P. spicata, and a well-developed dark-colored biological soil crust community on open soil surfaces between vegetation patches. Samples were randomly collected from BSC communities that were representative of the study site. Microclimate Measurements Soil surface microclimate of dark BSC communities at the Farwell Canyon site was monitored from November 2009 to October 2011. Sensors were placed in two adjacent BSC 37 patches on the lower terrace. Each patch was ca. 4 m2 in area, located in gaps between matured, tridentata plants. Crust-level parameters measured were temperature, moisture and light intensity. Figure 2.1 View of the vegetation and landscape representative of the study site, at Farwell Canyon, British Columbia. BSC surface temperature (°C) was measured with fme-wire (Type T, copperconstantan) thermocouples (n=6) connected to a multiplexer (AM25T, Campbell Scientific Inc). Additional fine-wire (unshielded) thermocouples (n-2) were placed approximately 1 m above ground to measure ambient air temperature. Light intensity, as incident photosynthetically active radiation (pmol PAR m 'V ), was measured with LI-COR Quantum 38 sensors (n=4) (Li-Cor, Lincoln NE, USA) placed on the crust surface. Volumetric soil moisture was measured with ECH2 O EC5 moisture sensors («=4) (Decagon Devices, Pullman WA, USA), buried 2 cm below the BSC surface. 50 40 30 20 10 0 0.0 0.2 0.4 0.6 0.8 1.0 Relative Impedence Response Figure 2.2. Calibration curve for mean relative impedance response (n=8) and gravimetric moisture content (MC) of field collected BSC (w=8-12). Error bars represent standard error. BSC hydration status was determined using impedance clips (n=8) inserted 2 mm apart at the crust surface (after Coxson 1991). Values from the impedance clips were standardized to a relative response of 100% during known wetting events. This was done every few months as the impedance response was found to diminish over time. The mean relative impedance response across all the sensors was calibrated against the gravimetric moisture (w/w) content of crust samples collected during different periods throughout the two years (Figure 2.2). An exponential curve fitted against this plot (f=2.7797*exp(2.7797*x); R2=0.9551) was then used to convert the impedance values to % moisture of the BSC. The impedance response method does not provide accurate 39 measurements at sub-zero temperatures. We therefore made the assumption that the moisture content of BSC remained constant after temperatures fell below 0°C until the next thaw period. This assumption may be accurate for crusts under a snowpack; however, exposed crusts often continue to dry from sublimation at sub-zero temperatures, an error that is not corrected in the present data set. Measurements for each sensor were taken every 30 seconds and summaries of average, maximum and minimum measurements were recorded at 10 minute intervals using a Campbell Scientific CR23X datalogger. The microclimate dataset is continuous with the exception of the period March 10-April 19,2010. A B.C. Ministry of Transportation weather station at Riske Creek (11 km northeast of study site at 1111 m a.s.l.) recorded two days of precipitation with a total of 3.2 mm during this period. An additional datalogger (Campbell Scientific CR10 and CR10X) was set up in a secondary area of south-facing BSC communities using the same types of sensors from Februaiy to May 2011, providing additional spot observations during the snow melt period. Acetylene Reduction Analysis Measurements of nitrogenase activity were conducted using the acetylene reduction assay (Stewart et al. 1967). BSC samples were collected as disks 2.5 cm in diameter and 1 cm deep. These were stored dry in growth chambers; summer collections were held at 10/15°C with an 18/6 hour light-dark photoperiod, while winter collections were held at 5/10°C with an 8/16 hour light-dark photoperiod. The size of the samples approximated that of mounds that characterize the BSC surface in this ecosystem, which generally range from 2 to 10 cm in diameter. Samples were hydrated to ~30% moisture content (w/w) for 24 hours 40 prior to ARA treatment. ARA incubations were conducted in 100 mL glass cuvettes equipped with rubber stoppers and sealed to glass plates with high-vacuum grease. The air within the cuvettes was replaced with acetylene gas equal to 10% of the chamber volume. The acetylene was generated on-site with CaC2 and water. A second control sample containing a BSC sample was fitted with a fine-wire thermocouple to monitor BSC surface temperature within cuvettes. Samples were incubated for 4-6 hours with 10 replicates for each set of incubations. A subsample (0.5 mL) of the air within the cuvettes was injected into a gas chromatograph (SRI 8610A, Wennick Scientific Corporation, Ottawa, ON, Canada) fitted with a Porapak column (Alltech Canada, Guelph, ON, Canada) and a flame ionization detector, to measure the concentration of ethylene produced. A calibration gas containing 101 pi l 1ethylene (Linde Canada Ltd., Edmonton, Alta) was used to calibrate the gas chromatograph for each incubation. Control cuvettes containing (1) BSCs without acetylene and (2) bare soil (no BSC) injected with acetylene, did not show any production of ethylene. In order to determine the effect of thallus moisture content on nitrogenase activity, summer collected BSC were hydrated to 30% moisture content (w/w) and pre-treated for 48 hours, after which time ARA incubations were conducted at 21°C and 200 pmol PAR m 'V 1, allowing the crusts to dry down to 20,15, 10, 7.5, and 5% MC (w/w) between sequential incubations. In order to determine the effect of length of prior desiccation exposure, three sets (n=10) of summer collected BSC samples were collected and pre-treated as described above for 48 hours, then incubated in order to determine baseline ARA rates. Each set of samples was held dry for a period of 1 week, 3 weeks, or 6 weeks. The samples were then hydrated and incubated after 2,24,48, 72 and 96 hours since hydration. 41 ARA incubations were conducted in situ at the study site during several precipitation and snowmelt episodes over the course of the study. Crusts were collected in the same manner as described above and immediately incubated for 4-6 hours with 10% (v/v) of acetylene, with cuvettes placed in-situ on the soil surface. Temperature within the cuvettes was kept within 3°C of ambient crust temperatures using ice packed around the cuvettes to cool the incubation chambers when necessary. A sample of the air within the incubation chambers was collected using vacutainers (Becton-Dickson, Franklin Lakes, NJ) and were brought back to the lab for analysis with the gas chromatograph. Reference containers were tested for ethylene contamination. In order to determine the effect of temperature and light intensity on nitrogenase activity incubations were conducted under the following temperature and light combinations: 14,21, 28°C and 100, 300, 600,1000 pmol PAR m ' V 1 for BSCs collected in the summer. Winter-collected BSCs were incubated at 0, 7, 14, 21°C and 100, 300,600 pmol PAR m ' V 1. Dark incubations (0 pmol PAR m ' V 1) were conducted at 10°C for both winter and summer collected BSCs. Rates of nitrogenase activity were calculated as micromoles of ethylene reduced per hour per m2 based on the incubation time and the area of the crust samples. lsN-Incubations Summer collected BSC samples were collected, stored and pre-incubated in the same manner as for the ARA incubations. An incubation time of 48 hours was used in order to ensure detection of 15N enrichment. Two sets of 10 replicates were incubated at 14°C and 21°C respectively, under 200 pmol PAR m ' V 1. The air within the cuvettes (10% v/v) was replaced with 10% (v/v) 15N2 gas (Cambridge Isotope Laboratories Inc., Andover, MA, USA, 15N2 , 98%+). Each cuvette was also injected with 5% (v/v) CO2 in order to minimize CO2 42 limitation during the long incubation period. After each incubation, samples were immediately dried at 105°C, ground in a ball mill and sent for 15N and total N analysis (Stable Isotope Facilities, University of Saskatchewan, Saskatoon, SK, Canada). Control samples (n=1 0 ) were incubated at the same time but with C2 H2 were used to determine the natural abundance ,5N and the acetylene reduction rate. The rate o f nitrogen uptake was calculated as per Liengen (1999a) using the equation: Y = ( atom % 15J W e„ \ V 100 J / to ta l Nsampie x 109\ / \ t x 28 / 100 \ \% 1 5 n a ir ) ^ ' where Y (nmol N-gdw"l-h_1) is the rate of N 2 uptake, atom% 15N«cess is the difference between atom% 1^sample and atom% 1^control, total N is the total amount of nitrogen in the sample (g- 1 0 0 gdw-1), t is the incubation time, 28 is the molecular weight of N 2 (g/mol), and %15N air is the percentage of 1SN out of the total amount of N gas in each incubation chamber (12.81%). Conversion ratios for ,SN uptake to ARA were calculated from the pooled averages of the samples incubated with 15N and the samples incubated with C 2 H2 for each temperature (Table 2.3). Table 2.3. Summary of conversion ratio for ethylene production to 15N-incorporation (±SE) by biological soil crusts incubated for 48 hours at 21°C and 200 pmol PAR m 'V 1. n=10. Total N (fi/gdw) 0.126 (0.011) Atom %,SN excess 0.02703 (0.0042) ,5N2 fixed (nmol/cm2/h) 124.64 (21.45) C2H4 Produced (nmol/cmz/h) 7.38 (0.238) C1H4/N2 0.059 Modeling BSC N-fixation activity under field conditions The data obtained from the ARA temperature x light matrices for winter and summer collected BSCs were fitted to a 3-dimensional Gaussian curve in SigmaPlot 11.0. Individual 43 curves were obtained for each of the winter and summer periods (R2=0.9648, p<0.0001; R2=0.9634, p<0.0001). These curves were subsequently used to estimate optimal ARA in field BSC communities based on measured field microclimate conditions within 10 minute blocks in the dataset. Model assumptions include saturation of ARA above 600 pmol PAR m 'V 1in winter and above 1000 pmol PAR-m'2-s'1in summer and continued ARA activity during periods of darkness under snow cover. ARA activity was set at zero when temperatures were below -5°C. The dates of May 1st to September 30th were used to differentiate between winter and summer in the data, which roughly corresponded to the frost-free period in the dataset. The calculated ARA estimates (which assumed optimal BSC moisture content and optimal pre-treatment hydration conditions) were corrected for actual moisture content and prior desiccation exposure using polynomial curves fitted to relative ARA responses for these variables. Crusts were assumed to be physiologically active at moisture contents above 7.5% (w/w), corresponding to the lowest moisture content at which ARA activity was detectable during laboratory experiments. Visual estimates of percent cover of dark BSCs at the Farwell Canyon study site were taken within 30 cm x 30 cm quadrats placed at 5 m intervals along N-S oriented 50 m transect lines. Lines were placed 10 m apart, with four groups of three lines dispersed within an approximately 1 km2 area around the study site. 44 RESULTS Farwell Canyon BSC Communities BSC communities on the lower Chilcotin River valley terraces at Farwell Canyon showed little evidence of past grazing; crust communities were largely intact, with mean cover of dark crusts averaging 52.7% (+ or - 2.73% S.E., n=120). The BSC communities contained several rare species, including Gypsophila macrophylla, and what appears to be a new species of Endocarpon (C. Bjork, Personal Communication). Collema coccophorum and C. tenax were common nitrogen fixing lichen constituents of the dark BSC communities, while Nostoc sp. and Scytonema cf. intertextum were abundant free-living constituents (Table 2.4). 45 Table 2.4. Species identified in BSC samples from Farwell Canyon study site. Species marked with an "X" include those found in the dark BSC communities sampled for ARA incubations. Species Comments Cyanobacteria Nostoc sp. * x Scytonema cf. intertextum* x Chroococcidiopsis sp. x Lichens Acarospora schleicheri Caloplaca atroalba Caloplaca cerina Caloplaca tominii x Candelariella aggregate x Cladonia pyxidata x rare species Collema coccophorum* x Collema tenax* x Diploschistes muscorum Endocarpon sp. nov. x rare species Endocarpon pusillum x Fulgensia bracteata x rare species Gypsoplaca macrophylla rare species Lecanora hagenii group Megaspora verrucosa? x Peltigera ponojensis* Physconia muscigena Placidium lachneum x Placidium lacinulatum x Placidium squamulosum Polychidium muscicola* Psora decipiens Psora globifera Psora Montana Rinodina terrestris x Toninia sedifolia Caloplaca ammiospila? ^Indicates N-fixing species 46 Acetylene Reduction Assays The temperature-light response of ARA under optimal hydration shows an increase in ARA activity with increasing temperature and light for both winter- and summer-collected BSC (Figure 2.3). Winter-collected crusts showed lower ARA activity than summer collected crusts at similar temperature and light conditions with a maximum rate of 116 pmol C2 H2 -m'2 h'1 at 21°C and 600 pmol PAR m ’V 1 compared to 173 pmol C2 H2 -m'2 h'1 at the same conditions for summer collected crusts. There was little difference in ARA rates between BSC incubated at 21 and 28°C. ARA activity increased with increasing light intensity, though this pattern appeared less pronounced at lower temperatures. ARA rates declined slightly between 30% and 15% MC and then exhibited a marked drop below 15% MC (Figure 2.3). The recovery curve of ARA activity over time after hydration showed that peak ARA activity occurs after BSCs are held hydrated for between 24 and 48 hours (Figure 2.4). ARA rates were higher for summer collected material when incubated at 21 °C, but were not significantly different (t-test; p>0.05) when incubated at 14°C. There were no significant differences (ANOVA, P>0.05) between the recovery curves of BSCs held dry for different time periods. BSC Microclimate Data A comparison of temperature and precipitation recorded at the Williams Lake Airport (Environment Canada 2012) during the study period with the 30 year average (1971-2000) reveals that 2009-2010 was a dry year compared to normal, while 2010-2011 was a wet year, particularly in the summer. Winter temperatures were warmer than normal in 2009-2010 and cooler than normal in 2010-2011, while summer temperatures were fairly close to normal for both years (Table 2.2). 47 A. Writer B. Summer 180 250 140 200 • C 120 150 100 3 > 801 $ 401 1000 800 20 800 'C\ 400 200 -10 Figure 2.3. Temperature and Light Response of acetylene reduction assay (ARA) fitted with multivariate curves by BSCs collected in (A) March and (B) July, n =10 for each data point. 120 R = 0.98 R2 = 0.9492 100 80 60 40 20 0 10 20 0 30 Moisture Content (% w/w) 20 40 60 80 100 120 140 Time Since Hydration (hours) Figure 2.4. (A) Relative ARA response to moisture content for BSCs pre-treated at optimum conditions and incubated at 21°C and 200 pmol PAR m 'V 1 (B) Relative ARA response over time starting at 2 hours after hydration to 30% moisture content. 48 V 500 | 200 Feb May Aug Nov Figure 2.5. Summary of BSC microclimate from 11 November 2009 to 1 October 2011 at Farwell Canyon, British Columbia. As snow-melt pockets develop during the late winter period (Figure 2.6) extreme diel temperature fluctuations can be experienced by hydrated BSCs. A representative data set from snow melt pockets under late-winter conditions (Figure 2.7) shows daily freeze-thaw cycles, with night-time BSC temperatures reaching -20°C, followed by crust surface temperatures of over 10°C under full sun conditions during the following midday period. 49 Figure 2.6. View of biological soil crusts melting out from under snow cover in early March at the Farwell Canyon study site. The width of the crust segment in the center of the image is ca. 9 cm. Hydration episodes at the BSC surface due to rainfall were on average 18 hours long, with a maximum hydration episode of 115 hours. Time between wetting events lasted an average of 2 days with a maximum of 16 days. The pattern of hydration events varied between seasons and between the two study years. Hydrated BSCs in the summer spent a majority of time in the light (65% in 2010 and 56% in 2011), while in the winter, hydrated BSCs were in the dark for the majority of time (70% in 2009-2010 and 56% in 2010-2011) (Figure 2.8). Thallus temperatures for hydrated BSCs were predominately between -10 and 0°C in the winter and between 5 and 15°C in the summer. 50 Predicted BSC N-fixation Although BSC communities experience extreme daily temperature fluctuations during the late winter period, model values for ARA on a south-facing aspect in late February and early March (Figure 2.7) predict substantial recovery during midday thaw periods. Air BSC -10 -20 1500 1000 500 12 24 12 24 12 24 12 24 12 24 12 24 Hours 27 28 1 February 2 3 4 March Figure 2.7. BSC microclimate on south-facing aspect from 27 Feb. to 4 March 2011. From top: predicted rates of acetylene reduction (ARA) from microclimate model, air and BSC surface temperature, and incident photosynthetically active radiation (PAR) (pmol PAR m-2s-l ) at the BSC surface. 51 Verification of late-winter ARA recovery was obtained in mid-March at the Farwell Canyon site (Figure 2.9). Field measurements of ARA activity showed peak rates in the midaftemoon and low but measurable rates overnight (Figure 2.9). ARA rates measured in the field were similar both days, even though temperature and light levels were lower during the second day of incubations. This resulted in predicted ARA rates being lower than actual rates during the second day of incubations. C. Sum m er Y ear 1 A. W inter Year 1* Dark Light ■ ■ 800 600 400 TJ & 200 -5 o X D. Sum m er Y ear 2 3eo X F. Sum m er Y ear 1 H. Sum m er Y ear 2 G. W inter Y ear 2 Temperature Class (°C) Figure 2.8.(A-D) Time (in hours) spent by hydrated (> 7.5% MC w/w) BSCs and (E-H) the estimated amount of C2H2fixed (in mmol of C2H2-m'2) by hydrated BSCs at temperatures plotted in 5°C class intervals for winter (October to April) and summer (May to September), for the periods November 2009- September 2010 (Year 1) and October 2010 to September 2011 (Year 2). Calculations exclude periods of missing data (from 1 Nov - 11 Nov Year 1 and 9 Mar - 13 Apr Year 2). 52 - BSC Air -10 EC 400 O 200 - 20 - co 100 • - ARA - Predicted ARA - Field 40 - 20 Mar 16 6 12 6 Mar 17 6 12 6 Mar 18 6 12 6 M ar 19 6 12 6 M ar 20 Figure 2.9. Predicted ARA rates based on BSC microclimate measurements during a spring snowmelt event at Farwell Canyon in March 2011. The symbols represent ARA rates measured in situ with standard error (Y-error bars) and incubation time (X-error bars). Field measurements of ARA during spring-time rainfall events also found that predicted ARA was sometimes lower than actual rates during field measurements (Figure 2.10). Field ARA measured 30 hours after wetting (on May 17) was higher than predicted by 53 the model, as were rates measured the following night (Figure 2.10). During the following period of full sun exposure as crusts started to dry ARA rates were variable, but were close to predicted rates. 40 - BSC Air 30 ■ 20 - > V v -r " ' v - 'v 1200 CL 800 ■ 400 ■ g> 25 20 15 - 10 - ARA-Field ARA - Predicted 200 - May 16 6 12 6 May 17 6 12 6 May 18 6 12 6 M ay 19 6 12 6 M ay 20 Figure 2.10. Predicted ARA rates based on BSC microclimate measurements during a rain event at Farwell Canyon in May 2011. The symbols represent ARA rates measured in situ with standard error (Y-error bars) and incubation time (X-error bars). 54 BSC Air 30 - 1600 VC 1200 - 2: 400 • ^ 300 - o 100 - _g> Jul 10 • A R A -F ield ARA - Predicted 8 12 6 12 8 1 , Ju l 12 e 12 8 Jul 13 8 12 8 Jul 14 Figure 2.11. Predicted ARA rates based on BSC microclimate measurements during a rain event at Farwell Canyon in July 2011. The symbols represent ARA rates measured in situ with standard error (Y-error bars) and incubation time (X-error bars). Observed ARA activity during a rain event in July shows a pattern of physiological recovery after wetting (Figure 2.11). ARA rates of 14 pm olm 2 h'1were measured in the mid-morning, shortly after wetting, and rose to 116 pmol m2 h'’ in the late afternoon, approximately 12 hours after initial hydration. Maximum ARA activity was reached during the following morning, with average rates of nearly 200 pmol m2 hT. The model 55 underestimated ARA rates during this recovery period, but fitted more closely with the field data after 24 hours from initial hydration. A drop in ARA rates during periods of high solar insolation immediately prior to drying was observed was observed during both summer field incubation sets (Figure 2.10 and Figure 2.11). Based on the model, the majority of N-fixation activity in the summer occurred under light conditions between the temperatures of 10 and 20°C (Figure 2.12). Only 15% of fixation activity during the summer occurred in the dark, although hydrated BSCs were in the dark approximately 40% of the time. The distribution of summer ARA activity across temperature and light classes was similar for both years, but Year 2 had higher ARA activity overall. A greater amount of ARA activity occurred in the light and at higher temperatures in Year 2 compared to Year 1. 800 ■Q 600 - >> 400 - 2 200 - - O o c n Uf (V w >. c W 9 - IS g —3 at a. ft 5^ > o w .® “ oW f wf cz jo *= t nw s ® ^ >, c Q- ® 3 2010-2011 2009 - 2010 Figure 2.12. Monthly summary of BSC hydration time and estimated ARA activity at a shrub-steppe and shrub-steppe in Farwell Canyon, B.C. Asterisks denote time periods with missing data. 56 The monthly patterns of ARA activity and BSC hydration differed between the two study years (Figure 2.8). February had the highest monthly ARA activity in Year 1, followed by January and March. In Year 2, July had the highest monthly ARA activity, followed by March and May. Predicted landscape-level nitrogen inputs based on a 3:1 C 2 H2 :N2 conversion ratio, were between 0.7 and 0.8 kg N-ha'1(36 to 40 kg N-ha'1with a 0.06 conversion ratio) in the winter period (Table 2.5). Predicted nitrogen inputs in the summer were between 0.57 and 0.96 kg N-ha'1(28 to 48 kg N-ha'1 with a 0.06 conversion ratio). Table 2.5. Summary of estimated seasonal N-inputs per hectare by BSCs in a shrub-steppe in Farwell Canyon, B.C. The estimates are based on conversion from modeled C2H2 inputs using both an experimentally derived C2H2:N2 conversion ratio of 0.06 and a commonly used conversion ratio of 3. Landscape level N-input estimates were determined using a BSC cover estimate of 52.7%. Estimated N-input Estimated N-input Est. C2H2 Time (0.06:1 C2H2:N2ratio) (3:1 C2H2:N2 ratio) Hydrated fixed (mmol-m'2) (mg-m'2) (kg-ha'1) (days) (kg-ha'1) (mg-m'2) Year 1 (2009-2010) Winter* 119.2 152 0.80 40 16.3 7621 Summer 22.5 11.5 0.57 108 5381 28 Year 2 (2010-2011) 14.8 Winter 142.5 138 0.73 6981 36 9087 Summer 29.6 19.5 182 0.96 48 DISCUSSION Response of nitrogenase activity to microclimate and seasonal patterns The general temperature, light and moisture response curves for nitrogenase activity are similar to those of other BSCs and terrestrial cyanobacteria (Coxson and Kershaw 1983a, Marsh et al. 2006, Su et al. 2011). Nitrogenase activity decreased with declining moisture below 15% MC (w/w) and showed little response at higher moisture contents. It is difficult to compare moisture response to other studies as different measures of moisture contents are 57 often used (Belnap 2001), but the shape of this response is similar to that reported for other BSCs and terrestrial cyanobacteria. Cyanobacteria-dominated BSCs from the hilly Loess Plateau in the northern Shaanxi province of China exhibited little change in nitrogenase activity at field water capacities above 40%, which corresponded to 12-19% (w/w) depending on the age of the crust (Zhao et al. 2010). Recovery of nitrogenase activity can vary based on prior hydration conditions (Coxson and Kershaw 1983a, Belnap 2002). Full recovery of nitrogenase activity occurred between 24 and 48 hours of continuous hydration. Our field measurements suggest the recovery curve used in the model may overestimate actual recovery time, as field ARA measurements within several hours after hydration were typically higher than predicted rates. The optimal thallus temperature for nitrogenase activity in this study was between 21 and 28°C. This is consistent with the optimal temperature for BSCs collected from other cold-temperate environments (Liengen 1999b, Belnap 2002, Marsh et al. 2006). Nitrogenase activity has been shown to sharply decline at temperatures above the optimum (Belnap 2002), but hydrated BSCs at our study site rarely experienced temperatures above 30°C. Most of the literature reports nitrogenase activity to cease below 0°C, but nitrogenase activity has been reported at temperatures as low as -7.5°C (Belnap 2001). Our laboratory incubations detected low but measureable rates of nitrogenase activity at a thallus temperature of 0°C, and our field data showed overnight ARA rates at temperatures down to -3°C fit closely to modelled rates. In order to provide a conservative estimate of rates of winter nitrogen fixation, our calculations assumed that nitrogenase activity ceased at temperatures below -5°C. This temperature cut-off also minimizes bias in our ARA estimates during long periods when BSCs were in the dark under snow pack at temperatures below 58 5°C (as we did not correct for the decline in nitrogen fixation during prolonged dark periods under snowpack). It is reasonable to assume that very little ARA activity would occur during long-term burial under snow pack, as was reported by Crittenden and Kershaw (1979) in Stereocaulon mats. Our microclimate measurements show repeated freeze-thaw cycles in the late winter and early spring. Temperatures for hydrated BSCs during these periods could range from 10°C at night to over 20°C in the day. The response of nitrogenase activity to freeze-thaw events in Nostoc commune was dependent on the temperature and light conditions during the diurnal cycles (Coxson and Kershaw 1983b). Nitrogenase can take several days to recover fully after long periods of inactivation during freezing, but recovery is rapid during freezethaw if carbon stores can be replenished by photosynthesis during the day (Coxson and Kershaw 1983b). We did not incorporate recovery time between freeze-thaw events, and our late-winter field incubation results suggest that the model fit closely and may even underestimate nitrogenase activity after freezing. In general, soil cyanobacteria and cyanolichens have been found to reach light saturation at low light levels (Belnap 2001). Light saturation for BSCs in this study was between 600 and 1000 pmol PAR m ' V for crusts collected in the summer and between 300 and 600 pmol PAR m 'V in the winter. Light saturation also appeared to be temperature dependent, with rates stabilizing at lower light levels under lower temperatures. Reported light saturation for BSC organisms is variable; nitrogenase activity in N. commune reached saturations levels at above 900 pmol PAR m ' V (Coxson and Kershaw 1983b), cyanobacteria from the low arctic showed little to no response to light intensity (Stewart et al. 2011b). 59 BSCs can continue to fix nitrogen in darkness, depending on the amount of available carbon stores and the rate of respiration. There were 30 consecutive days in the winter of 2009 and 74 days in the winter of 2010 where the light levels did not exceed 50 pmol PAR m 'V due to snow cover. BSC temperatures during these periods did not exceed -0.2°C. Dark ARA incubations conducted at 1°C showed continued low rates of nitrogenase activity over the course of 48 hours. It is unclear how long nitrogenase activity can persist during extended periods of low light and low temperatures under snow cover. Potential landscape-level nitrogen inputs The landscape level estimates for nitrogen input in the Chilcotin shrub-steppe ecosystem varied dramatically, depending on which ratio was used to convert from acetylene reduction estimates to actual nitrogen fixed. If we used the ARA/Nis conversion ratio obtained in this study our estimate of landscape level nitrogen fixation reach approximately 75 kg/ha/year. These estimates for annual nitrogen BSC fixation are high compared to most recent studies. They are however, within the range of 10-100 kg/ha/year reported by West and Skujins (1977) for cold desert ecosystems. There are several factors that may result in this high estimate, as discussed below. One factor responsible for these high estimates of nitrogen is our experimentally derived conversion ratio of 0.06. This conversion ratio is well below the theoretical factor of 3, often used in studies which do not calibrate ARA to ,5N2 incorporation. Conversion ratios reported in the literature for terrestrial habitats have been found to vary widely. Our conversion ratio was determined using similar methodology to Liengen (1999a), who reported conversion ratios ranging from 0.022 to 0.073 for Arctic cyanobacterial crusts. Belnap (2002) also used Liengen’s (1999a) conversion ratio of 0.062 and suggested that a 3:1 60 ratio greatly underestimates N inputs in most BSC studies. Differences in conversion ratios may be due to a suppression of acetylene reduction, the favoring of l5N2 incorporation, or a combination of both (Liengen 1999a). Suppression of ARA has been found to occur during extended incubations. ARA rates during the side-by-side incubations in our study were similar to those calculated during shorter incubation times under the same temperature and light conditions, thus this is not likely to be a contributing factor in our study. Another possibility, as suggested by Liengen (1999a), is the differential diffusion of acetylene, ethylene and dinitrogen through the soil and polysaccharide gels surrounding the BSCs. The exchange of smaller dinitrogen molecules may be favoured over the larger acetylene and ethylene molecules, resulting in a lower conversion ratio. ARA rates vary widely among different studies, over a range of 5 orders of magnitude (Su et al. 2011). Many factors can affect these differences in rates, including pre­ treatment conditions (Coxson and Kershaw 1983a, Belnap 2002), species composition (Belnap 2002, Stewart et al. 2011a, Su et al. 2011), successional stage (Housman et al. 2006) and disturbance history (Belnap 1996). The suppression of acetylene reduction due to differential diffusion or other effects may also play a role, further stressing the need for sitespecific conversion ratios. The optimum ARA rates observed in this study were similar to the maximum rates observed in Collema dominated BSCs from a cold desert in Moab, Utah (Belnap 2002). There are several other factors that may affect our landscape-level estimate of nitrogen input. Our estimate of BSC cover was based only on surveys of gently sloped, south-facing terraces within the study area. BSC cover would be expected to decrease with increasing slope. Aspect can strongly affect nitrogenase activity, even at the microscale, due 61 to differences in surface temperature and drying times (Davidson et al. 2002). Species composition can also affect potential nitrogenase activity of BSCs. Chlorolichen dominated BSCs will have lower rates of nitrogenase activity than cyanobacteria dominated dark BSCs (Wu et al. 2009, Su et al. 2011). Our estimates of cover and nitrogenase activity were based on material collected from relatively undisturbed shrub-steppe. A number of studies have noted declines in both BSC cover and nitrogenase activity with grazing and other human activity (Evans and Ehleringer 1993, Belnap 1996, Evans and Belnap 1999, Housman et al. 2006). The nitrogen fixation potential in areas more heavily impacted by surface disturbance would likely be lower than what is reported here. Seasonality of grazing may also be an important factor to consider. Memmott et al. (1998), for instance, found that BSC cover declined from 27% in control sites in cold desert ecosystems of the US Great Basin to 14% in summer grazed paddocks and 10% in spring grazed paddocks. These are values are similar to what we have observed anecdotally in more heavily grazed Chilcotin shrub-steppe. Seasonal patterns of nitrogen inputs by Chilcotin BSC communities Traditional consideration of rangeland productivity focuses primarily on the spring and summer growing seasons, when soils are unfrozen and water is available for plant growth (Webb et al. 1978, Morton 1979). However, BSC communities do not face a similar constraint. They can be physiologically active at any time of year, whenever water availability and temperatures allow. In our 2009-2010 field season, which was characterized by a relatively dry summer period, we found that high rates of BSC physiological activity occurred during the late winter and early spring period, when snow melt episodes were triggered by intense solar insolation. In our second field season the highest rates of ARA activity occurred during July wetting events, although late winter (March) melt episodes 62 were again highly significant. The results of the model indicate that differences in seasonal precipitation patterns may have profound effects of the overall potential fixed nitrogen and the composition of BSC communities, a point emphasized by Belnap et al. (2004), who found that Collema dominated crusts showed a large decline after treatments simulating greater summer precipitation frequency. Nitrogen released by BSC communities during summer rainfall events is likely available to higher plant communities, however it is unclear what is the fate of fixed nitrogen released to a shallow active layer above frozen ground in late winter. In many respects the seasonal dynamics of late winter nitrogen cycling in Chilcotin shrub-steppe may be closer to that of BSC communities in polar ecosystems, where released nitrogen can be entrained in a shallow active layer (Stewart et al. 201 la, Niederberger et al. 2012). Marsh et al. (2006) documented enrichment in total and mineralizable N in soils immediately under Chilcotin BSC communities, suggesting that BSCs may play a major role in enriching these soils. Marsh et al. (2006) also found that 15N natural abundance values in soils under BSCs were much closer to atmospheric values than to soil derived values, again supporting the hypothesis of BSC-derived nitrogen in Chilcotin soils. However, further estimates of nitrogen uptake by plants, nitrogen volatilization, denitrification and leaching, are needed for this area, particularly during the critical late-winter and early-spring period. 63 SUMMARY AND CONCLUSIONS Biological soil crusts have been found to play significant roles in soil nitrogen cycling in arid and semi-arid throughout the world, and British Columbia grasslands and shrub-steppe appear to be no exception. Nitrogen inputs by BSCs in this ecosystem are also sensitive to changes in temperature and precipitation patterns, stressing the need to include these communities in any climate change planning and management of grasslands and shrubsteppe in this province. At Lac du Bois Grasslands, we compared the nitrogen and carbon content o f BSCs and the underlying soil across an elevation gradient corresponding with increasing precipitation and decreased temperature. BSCs had a higher nitrogen and carbon content with increasing elevation and consistently higher carbon and nitrogen than the underlying soil. Furthermore, the average nitrogen isotopic composition of 15N in the BSCs was 2.24%o, indicating that crusts are net importers of atmospheric nitrogen, and likely provide available nitrogen to nearby plants. The decline in carbon and nitrogen content at lower elevations may be due to lower rates of productivity and higher rates of N-losses. This suggests that a shift to a climate with higher potential evapotranspiration will lead to a decline in nitrogen and carbon content in soils occupied by BSCs. In the Chilcotin grasslands, we determined the response of nitrogen fixation potential in BSCs from a lower elevation grassland to changes in moisture, temperature and light availability. Based on this response, we were able to model nitrogen fixation potential over a two year period and determine an annual landscape estimate of nitrogen input for this 64 ecosystem. This landscape estimate varied substantially depending on the conversion ratio used to convert rates of acetylene converted to rates of nitrogen fixed. Using a theoretical ratio of 3:1, estimates of average annual BSC nitrogen fixation were 1.7 kgN h a '\ However, using a conversion ratio of 0.06, obtained from side-by-side measurements o f acetylene reduction and ,5N uptake, average landscape level estimates of BSC nitrogen fixation are 83 kg N ha'1. Our estimates of nitrogen fixation also varied between years due to differences in the magnitude of summer precipitation. The results from the Lac du Bois grasslands were very similar to those previously reported in the Chilcotin region. This suggests that soil nutrient inputs by BSCs are similar in both of these areas, and may be consistent throughout similar grassland and shrub-steppe ecosystems in British Columbia. Our estimates for nitrogen fixation rates could also potentially be applied to similar BSC communities throughout British Columbia. These two studies both provide insight into the potential effects of climate change on BSCs, but at different scales. Our study of an elevational transect at the Lac du Bois grasslands suggests some potential effects of long-term shifts towards a hotter and drier climate, while our study of annual N-fixation rates in the Chilcotin shows differences in Nfixation potential under different patterns of temperature and precipitation. The results of both studies are important to the management of grasslands and shrub-steppe for climate change as our future climate is predicted to have both long term directional shifts as well as more short-term seasonal shifts such as timing of precipitation and snowmelt. The results presented in this thesis open up a variety of opportunities for further study. While we were able to quantify nitrogen inputs for lower elevation grasslands, the fate 65 of this fixed nitrogen is still largely unknown. Estimates of denitrification, ammonia volatilization, and plant uptake are needed to better understand the significance of BSCs to overall N-cycling in these ecosystems. Estimates of N-fixation rates for different BSC community types, including those found in higher elevation grasslands, can also help refine our model. Furthermore, our results suggest that BSCs contribute to soil carbon in grasslands, but the significance of BSCs to carbon cycling in this region is also yet to be quantified. Given the growing interest in soil carbon sequestration as a tool for reducing greenhouse gas concentrations in the atmosphere, this question has many practical implications. 66 REFERENCES Aranibar, J., Anderson, I., Ringrose, S., and Macko, S., 2003. Importance of nitrogen fixation in soil crusts of southern African arid ecosystems: acetylene reduction and stable isotope studies. Journal o f Arid Environments, 54(2), pp.345-358. Belnap, J., 1996. Soil surface disturbances in cold deserts: effects on nitrogenase activity in cyanobacterial-lichen soil crusts. Biology and Fertility o f Soils, 23(4), pp.362-367. Belnap, J., 2001. Factors influencing nitrogen fixation and nitrogen release in biological soil crusts. In J. Belnap & O. L. Lange, eds. Biological Soil Crusts: Structure, Function, and Management. Berlin: Springer-Verlag, pp. 241-261. Belnap, J., 2002. Nitrogen fixation in biological soil crusts from southeast Utah, USA. Biology and Fertility o f Soils, 35(2), pp. 128—135. Belnap, J., 2003. The world at your feet: desert biological soil crusts. Frontiers in Ecology and the Environment, 1(4), pp. 181-189. Belnap, J., 2006. The potential roles of biological soil crusts in dryland hydrologic cycles. Hydrological Processes, 20(15), pp.3159-3178. Belnap, J., Prasse, R. & Harper, K., 2001. Influence of biological soil crusts on soil environments and vascular plants. In J. Belnap & O. L. Lange, eds. Biological Soil Crusts: Structure, Function, and Management. Berlin: Springer-Verlag, pp. 281-300. Belnap, J., Phillips, S. & Miller, M., 2004. Response of desert biological soil crusts to alterations in precipitation frequency. Oecologia, 141(2), pp.306-316. Belnap, J., Phillips, S. & Troxler, T., 2006. Soil lichen and moss cover and species richness can be highly dynamic: The effects of invasion by the annual exotic grass Bromus tectorum, precipitation, and temperature on biological soil crusts in SE Utah. Applied Soil Ecology, 32(1), pp.63—76. Billings, S., Schaeffer, S. & Evans, R., 2003. Nitrogen fixation by biological soil crusts and heterotrophic bacteria in an intact Mojave Desert ecosystem with elevated CO2 and added soil carbon. Soil Biology & Biochemistry, 35(5), pp.643-649. Bowker, M., Maestre, F. & Escolar, C., 2010. Biological crusts as a model system for examining the biodiversity-ecosystem function relationship in soils. Soil Biology and Biochemistry, 42(3), pp.405—417. Coxson, D., 1991. Impedance measurement of thallus moisture content in lichens. The Lichenologist, 23(01), pp.77—84. Coxson, D. & Kershaw, K., 1983a. Nitrogenase activity during chinook snowmelt sequences by Nostoc commune in Stipa-Bouteloa grassland. Canadian Journal o f Microbiology, 29(8), pp.938-944. 67 Coxson, D. & Kershaw, K., 1983b. Rehydration response of nitrogenase activity and carbon fixation in terrestrial Nostoc commune from Stipa-Bouteloa grassland. Canadian Journal o f Botany, 61(10), pp.2658-2668. Crittenden, P. & Kershaw, K., 1979. Studies on lichen-dominated systems. XXII. The environmental control of nitrogenase activity in Stereocaulon paschale in sprucelichen woodland. Canadian Journal o f Botany, 57(3), pp.236-254. Dahlgren, R., Boettinger, J., Huntington, G., and Amundson, R., 1997. Soil development along an elevational transect in the western Sierra Nevada, California. Geoderma, 78(3), pp.207-236. Davidson, D., Bowker, M., George, D., Phillips, S., and Belnap, J., 2002. Treatment effects on performance of N-fixing lichens in disturbed soil crusts of the Colorado Plateau. Ecological Applications, 12(5), pp. 1391—1405. Dawson, R., Werner, A. & Murdock, T., 2008. Preliminary Analysis of Climate Change in the Cariboo-Chilcotin Area o f British Columbia, Victoria B.C.: Pacific Climate Impacts Consortium, University of Victoria. Available at: http://pacificclimate.org/sites/default/files/publicationsAVemer.ClimateChangeCaribo oChilcotin.Sep2008.pdf [Accessed January 17, 2013]. Elbert, W., Weber, B., Burrows, S., Steinkamp, J., Biidel, B., Andreae, M., and Poschl, U., 2012. Contribution of cryptogamic covers to the global cycles of carbon and nitrogen. Nature Geoscience, 5(7), pp.459-462. Eldridge, D., 1998. Trampling of microphytic crusts on calcareous soils, and its impact on erosion under rain-impacted flow. Catena, 33(3-4), pp.221-239. Environment Canada, 2012. Canadian Climate Normals and Averages. Available at: http://www.climate.weatheroffice.gc.ca/climate_normals/index_e.html [Accessed January 21,2013]. Evans, R. & Belnap, J., 1999. Long-term consequences of disturbance on nitrogen dynamics in an arid ecosystem. Ecology, 80(1), pp. 150-160. Evans, R. & Ehleringer, J., 1993. A break in the nitrogen cycle in aridlands? Evidence from delta 15N of soils. Oecologia, 94(3), pp.314-317. Evans, R. & Johansen, J., 1999. Microbiotic crusts and ecosystem processes. Critical Reviews in Plant Sciences, 18(2), pp. 183—225. Evans, R. & Lange, O., 2001. Biological soil crusts and ecosystem nitrogen and carbon dynamics. In J. Belnap & O. Lange, eds. Biological Soil Crusts: Structure, Function, and Management. Berlin: Springer-Verlag, pp. 263-279. Fenger, M., 1982. Characteristics and distribution of chemozemic soils in the Ashcroft map area. In A. Nicholson, A. McLean, & T. Baker, eds. Grassland Ecology and 68 Classification: Symposium Proceedings June 1982. Information Services Branch, B.C. Ministry of Forests, pp. 133-146. Available at: http://www.for.gov.bc.ca/hfd/pubs/docs/mr/Mr037.htm [Accessed February 12, 2013]. Frank, D. & Evans, R., 1997. Effects of native grazers on grassland N cycling in Yellowstone National Park. Ecology, 78(7), pp.2238-2248. GCC, 2004. B.C. grasslands mapping project: a conservation risk assessment (Final Report), Kamloops, B.C.: Grasslands Conservation Council of B.C. Available at: http://www.bcgrasslands.org/what-we-do/conservation-a-mapping/bc-grasslandsmapping-project [Accessed February 10, 2013]. Hamann, A. & Wang, T., 2006. Potential effects of climate change on ecosystem and tree species distribution in British Columbia. Ecology, 87(11), pp.2773-2786. Harper, K. & Belnap, J., 2001. The influence of biological soil crusts on mineral uptake by associated vascular plants. Journal o f Arid Environments, 47(3), pp.347—357. Hartley, A. & Schlesinger, W., 2002. Potential environmental controls on nitrogenase activity in biological crusts of the northern Chihuahuan Desert. Journal o f Arid Environments, 52(3), pp.293—304. Hawkes, C., 2003. Nitrogen cycling mediated by biological soil crusts and arbuscular mycorrhizal fungi. Ecology, 84(6), pp. 1553-1562. Hook, P. & Burke, I., 2000. Biogeochemistry in a shortgrass landscape: control by topography, soil texture, and microclimate. Ecology, 81(10), pp.2686-2703. Housman, D., Powers, H., Collins, A., and Belnap, J., 2006. Carbon and nitrogen fixation differ between successional stages of biological soil crusts in the Colorado Plateau and Chihuahuan Desert. Journal o f Arid Environments, 66(4), pp.620-634. Housman, D., Yeager, C., Darby, B., Sanford Jr., R., Kuske, C., Neher, D., and Belnap, J., 2007. Heterogeneity of soil nutrients and subsurface biota in a dryland ecosystem. Soil Biology and Biochemistry, 39(8), pp.2138-2149. IPCC WGI, 2007. Regional Climate Projections. In S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. Averyt, M. Tignor, and H. Miller, eds. Climate Change 2007: The Physical Science Basis. Contribution o f Working Group I to the Fourth Assessment Report o f the Intergovernmental Panel on Climate Change. Cambridge, UK: Cambridge University Press. Available at: http://www.ipcc.ch/publications_and_data/publications_ipcc_fourth_assessment_repo rt_wgl_report_the_physical_science_basis.htm [Accessed January 31,2013]. Johnson, S., Kuske, C., Carney, T., Housman, D., Gallegos-Graves, L., and Belnap, J., 2012. Increased temperature and altered summer precipitation have differential effects on 69 biological soil crusts in a dryland ecosystem. Global Change Biology, 18(8), pp.2583-2593. Johnson, S., Neuer, S. & Garcia-Pichel, F., 2007. Export of nitrogenous compounds due to incomplete cycling within biological soil crusts of arid lands. Environmental Microbiology, 9(3), pp.680-689. Kimmins, J. & Lavender, D., 1992. Ecosystem-level changes that may be expected in a changing global climate: A British Columbia perspective. Environmental Toxicology and Chemistry, 11(8), pp.1061—1068. Lange, O., 2001. Photosynthesis of soil-crust biota as dependent on environmental factors. In J. Belnap & O. Lange, eds. Biological Soil Crusts: Structure, Function, and Management. Berlin: Springer-Verlag, pp. 217—240. Langhans, T., Storm, C. & Schwabe, A., 2009. Biological soil crusts and their microenvironment: Impact on emergence, survival and establishment of seedlings. Flora - Morphology, Distribution, Functional Ecology o f Plants, 204(2), pp. 157—168. Lee, R., 2011. Plant Community Relationships to Soil Properties and Topography in a Southern Interior B.C. Grassland: A Refinement. MSc. Thesis. Vancouver, B.C.: University of British Columbia. Liengen, T., 1999a. Conversion factor between acetylene reduction and nitrogen fixation in free-living cyanobacteria from high arctic habitats. Canadian Journal o f Microbiology, 45(3), pp.223—229. Liengen, T., 1999b. Environmental factors influencing the nitrogen fixation activity of freeliving terrestrial cyanobacteria from a high arctic area, Spitsbergen. Canadian Journal o f Microbiology, 45(7), pp.573—581. Liengen, T. & Olsen, R., 1997. Nitrogen fixation by free-living cyanobacteria from different coastal sites in a high arctic tundra, Spitsbergen. Arctic and Alpine Research, 29(4), p p . 470-477. Link, S., Smith, J., Halvorson, J., and Bolton, H., 2003. A reciprocal transplant experiment within a climatic gradient in a semiarid shrub-steppe ecosystem: effects on bunchgrass growth and reproduction, soil carbon, and soil nitrogen. Global Change Biology, 9(7), pp. 1097-1105. Liu, H., Han, X., Li, L., Huang, J., Liu, H., and Li, X., 2009. Grazing density effects on cover, species composition, and nitrogen fixation of biological soil crust in an inner Mongolia steppe. Rangeland Ecology & Management, 62(4), pp.321—327. Mack, R. & Thompson, J., 1982. Evolution in steppe with few large, hooved mammals. American Naturalist, pp.757—773. 70 Maestre, F., Castillo-Monroy, A., Bowker, M., and Ochoa-Hueso, R., 2012. Species richness effects on ecosystem multifunctionality depend on evenness, composition and spatial pattern. Journal o f Ecology, 100(2), pp.317-330. Malam Issa, O., Defarge, C., Le Bissonnais, Y., Marin, B., Duval, O., Bruand, A., D’Acqui, L., Nordenberg, S., and Annerman, M., 2007. Effects of the inoculation of cyanobacteria on the microstructure and the structural stability of a tropical soil. Plant and Soil, 290(1), pp.209-219. Marsh, J., Nouvet, J., Sanborn, P., and Coxson, D., 2006. Composition and function of biological soil crust communities along topographic gradients in grasslands of central interior British Columbia (Chilcotin) and southwestern Yukon (Kluane). Botany, 84(5), pp.717-736. Meidinger, D. & Pojar, J., 1991. Ecosystems o f British Columbia, Vicotria, B.C.: Ministry of Forests, Reserach Branch. Available at: http://www.for.gov.bc.ca/hfd/pubs/Docs/Srs/Srs06.htm [Accessed February 10, 2013]. Memmott, K., Anderson, V. & Monsen, S., 1998. Seasonal grazing impact on cryptogamic crusts in a cold desert ecosystem. Journal o f Range Management, pp.547-550. Morton, J., 1979. Chilcotin meadowlands: livestock range and culture. Rangelands, 1(5), pp. 183-185. Neff, J., Reynolds, R., Belnap, J., and Lamothe, P., 2005. Multi-decadal impacts of grazing on soil physical and biogeochemical properties in southeast Utah. Ecological Applications, 15(1), pp.87—95. Niederberger, T., Sohm, J., Tirindelli, J., Gunderson, T., Capone, D., Carpenter, E., and Cary, S., 2012. Diverse and highly active diazotrophic assemblages inhabit ephemerally wetted soils of the Antarctic Dry Valleys. FEMS Microbiology Ecology, 82(2), pp.376-390. Nitschke, C. & limes, J., 2008. Integrating climate change into forest management in SouthCentral British Columbia: An assessment of landscape vulnerability and development of a climate-smart framework. Forest Ecology and Management, 256(3), pp.313—327. Pennock, D., Bedard-Haughn, A. & Viaud, V., 2011. Chemozemic soils of Canada: Genesis, distribution, and classification. Canadian Journal o f Soil Science, 91(5), pp.719—747. Ponzetti, J. & McCune, BP, 2001. Biotic soil crusts of Oregon’s shrub steppe: community composition in relation to soil chemistry, climate, and livestock activity. The Bryologist, 104(2), pp.212—225. Robinson, D., 2001. 5I5N as an integrator of the nitrogen cycle. Trends in Ecology & Evolution, 16(3), pp. 153-162. 71 Root, H. & McCune, B, 2012. Regional patterns of biological soil crust lichen species composition related to vegetation, soils, and climate in Oregon, USA. Journal o f Arid Environments, 79, pp.93-100. Rosenzweig, C. & Hillel, D., 2000. Soils and global climate change: Challenges and opportunities. Soil Science, 165(1), pp.47-56. Russow, R., Veste, M. & Littmann, T., 2004. Using the natural l5N abundance to assess the main nitrogen inputs into the sand dune area of the north-western Negev Desert (Israel). Isotopes in Environmental and Health Studies, 40(1), pp.57—67. Schimel, D., Stillwell, M. & Woodmansee, R., 1985. Biogeochemistry of C, N, and P in a soil catena of the shortgrass steppe. Ecology, pp.276-282. Smith, J., Halvorson, J. & Bolton Jr., H., 2002. Soil properties and microbial activity across a 500 m elevation gradient in a semi-arid environment. Soil Biology and Biochemistry, 34(11), pp.1749-1757. Spittlehouse, D., 2008. Climate Change, Impacts and Adaptation Scenarios: Climate change andforest and range management in British Columbia, Victoria, B.C.: B.C. Ministry of Forests. Research Branch. Available at: http://www.for.gov.bc.ca/hfd/pubs/Docs/Tr/Tr045.htm [Accessed February 10, 2013]. Steen, O. & Coupe, R., 1997. A field guide to forest site identification and interpretation fo r the Cariboo Forest Region, Victoria, B.C.: B.C. Ministry of Forests. Available at: http://www.for.gov.bc.ca/hfd/pubs/docs/lmh/lmh39.htm [Accessed February 10, 2013]. Stewart, K., Lamb, E., Coxson, D., and Siciliano, S., 201 lb. Bryophyte-cyanobacterial associations as a key factor in N 2 -fixation across the Canadian Arctic. Plant and Soil, 344(1-2), pp.335—346. Stewart, K., Coxson, D. & Grogan, P., 201 la. Nitrogen inputs by associative cyanobacteria across a low Arctic tundra landscape. Arctic, Antarctic, and Alpine Research, 43(2), pp.267-278. Stewart, W., Fitzgerald, G. & Burris, R., 1967. In situ studies on nitrogen fixation with the acetylene reduction technique. Science, 158(3800), pp.536—536. Su, Y., Zhao, X., Li, A., Li, X., and Huang, G. 2011,2011. Nitrogen fixation in biological soil crusts from the Tengger desert, northern China. European Journal o f Soil Biology, 47(3), pp. 182—187. Tisdale, E., 1947. The grasslands of the southern interior of British Columbia. Ecology, 28(4), pp.346-382. 72 Van Ryswyk, A. & McLean, A., 1989. Geology and soils of grassland ranges, Kamloops, British Columbia. Rangelands, 11(2), pp.65-67. Van Ryswyk, A., McLean, A. & Marchand, L., 1966. The climate, native vegetation, and soils of some grasslands at different elevations in British Columbia. Canadian Journal of Plant Science, 46(1), pp.35—50. Valentine, K., 1978. The soil landscapes o f British Columbia, B.C. Ministry of Envrionment, Resource Analysis branch. Available at: http://www.env.gov.bc.ca/soils/landscape/index.html [Accessed February 10, 2013]. Veluci, R., Neher, D. & Weicht, T., 2006. Nitrogen fixation and leaching of biological soil crust communities in mesic temperate soils. Microbial Ecology, 51(2), pp. 189-196. Verchot, L., Groffinan, P. & Frank, D., 2002. Landscape versus ungulate control of gross mineralization and gross nitrification in semi-arid grasslands of Yellowstone National Park. Soil Biology and Biochemistry, 34(11), pp. 1691—1699. Vincent, L. & Mekis, E., 2006. Changes in daily and extreme temperature and precipitation indices for Canada over the twentieth century. Atmosphere-Ocean, 44(2), pp. 177193. Walker, I. & Sydneysmith, R., 2008. British Columbia. In D. Lemmen, F. Warren, J. Lacroix, and E. Bush., eds. From Impacts to Adaptation: Canada in a Changing Climate 2007. Ottawa, ON: Government of Canada, pp. 329-386. Wang, T., Hamann, A., Spittlehouse, D., and Murdock, T., 2012. ClimateWNA—HighResolution Spatial Climate Data for Western North America. Journal o f Applied Meteorology and Climatology, 51(1), pp. 16-29. Webb, W., Szarek, S., Lauenroth, W., Kinerson, R., and Smith, M., 1978. Primary productivity and water use in native forest, grassland, and desert ecosystems. Ecology, pp. 1239-1247. West, N., 1991. Nutrient cycling in soils of semiarid and arid regions. In J. Skujins, ed. Semiarid Lands and Deserts: Soil Resource and Reclamation. New York: Marcel Dekker, pp. 295-332. West, N. & Skujins, J., 1977. The nitrogen cycle in North American cold-winter semi-desert ecosystems. Oecologia Plantarum, 12(1), pp.45—53. Wikeem, B. & Wikeem, S., 2004. The grasslands o f British Columbia, Kamloops, B.C.: Grasslands Conservation Council of B.C. Available at: http://www.bcgrzisslands.org/images/stories/leammore/mapping/pdf7bcgrasslandsfmal 2004ver3.pdf [Accessed February 10,2013], 73 Williams, A., Buck, B. & Beyene, M., 2012. Biological soil crusts in the Mojave Desert, USA: micromorphology and pedogenesis. Soil Science Society o f America Journal, 76(5), pp. 1685-1695. Wu, N., Zhang, Y. & Downing, A., 2009. Comparative study of nitrogenase activity in different types of biological soil crusts in the Gurbantunggut Desert, Northwestern China. Journal o f Arid Environments, 73(9), pp.828—833. Yeager, C., Komosky, J., Housman, D., Grote, E., Belnap, J., and Kuske, C., 2004. Diazotrophic community structure and function in two successional stages of biological soil crusts from the Colorado Plateau and Chihuahuan Desert. Applied and Environmental Microbiology, 70(2), pp.973—983. Zelikova, T., Housman, D., Grote, E., Neher, D., and Belnap, J., 2012. Warming and increased precipitation frequency on the Colorado Plateau: implications for biological soil crusts and soil processes. Plant and Soil, pp. 1—18. Zhang, X., Vincent, L., Hogg, W., and Niitsoo, A., 2000. Temperature and precipitation trends in Canada dining the 20th century. Atmosphere-Ocean, 38(3), pp.395—429. Zhao, Y., Xu, M. & Belnap, J., 2010. Potential nitrogen fixation activity of different aged biological soil crusts from rehabilitated grasslands of the hilly Loess Plateau, China. Journal o f Arid Environments, 74(10), pp. 1186-1191. 74