Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effects of wildfire and harvest disturbances on forest soil bacterial communities Smith, Nancy Rosalee 2005

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

Item Metadata

Download

Media
831-ubc_2005-0644.pdf [ 10.72MB ]
Metadata
JSON: 831-1.0092163.json
JSON-LD: 831-1.0092163-ld.json
RDF/XML (Pretty): 831-1.0092163-rdf.xml
RDF/JSON: 831-1.0092163-rdf.json
Turtle: 831-1.0092163-turtle.txt
N-Triples: 831-1.0092163-rdf-ntriples.txt
Original Record: 831-1.0092163-source.json
Full Text
831-1.0092163-fulltext.txt
Citation
831-1.0092163.ris

Full Text

EFFECTS OF WILDFIRE AND HARVEST DISTURBANCES ON FOREST SOIL BACTERIAL COMMUNITIES by N A N C Y R O S A L E E SMITH B.Sc , The University of British Columbia, 2001 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF SCIENCE in THE F A C U L T Y OF G R A D U A T E STUDIES (Microbiology and Immunology) THE UNIVERSITY OF BRITISH C O L U M B I A June 2005 © Nancy Rosalee Smith, 2005 11 A B S T R A C T Little attention has been paid to the effects of wildfires on soil bacterial communities. The opposite is true for the effects of harvest treatments on soil bacterial communities. This scarcity of microbiological research post wildfire may be because of the unpredictability of such events or because of a focus on other fire-induced changes such as soil chemistry. In this study, Boreal forest soil bacterial communities were assessed post disturbance in four treatments: control, harvest, burn and burn-salvage. The burn treatments were areas affected by the wildfire near Chisholm, Alberta in May, 2001. Changes in these microbial communities occurred as a consequence of the wildfire or harvest treatment disturbance, with greater effects in the burn treatments. Significant decreases in microbial biomass carbon (C m i C ) were seen as a result of the burn or harvest treatments. Microbial biomass nitrogen (N m j C ) decreased in the harvest treatment, but increased in the burn treatments, probably because of microbial assimilation of the increased amounts of available N H 4 + and NO3" due to burning. The C m j C : N m i C decreased in the harvest, burn and burn-salvage treatments, indicating a probable decrease in fungal biomass. Non-parametric ordination of molecular fingerprint data (ribosomal intergenic spacer analysis and rRNA gene denaturing-gradient gel electrophoresis) of 119 samples indicated clear distinctions between community composition in the burned and unburned treatments. Differences between control versus harvest and between burn versus burn-salvage treatments were less obvious, but multi-response permutation procedures demonstrated statistically significant separations between the two. Sequencing of bands from fingerprints uncovered interesting patterns of bacterial divisions specific to treatment type, y- and a-Proteobacteria were highly characteristic of the unburned treatments, while j3-Proteobacteria and members oi Bacillus were highly characteristic of the burned treatments. Biomass determinations confirmed general trends observed in past literature, while relatively new molecular methods unveiled new and interesting effects to bacterial communities in Boreal forest soils impacted by human and natural disturbances. iii TABLE OF CONTENTS ABSTRACT " TABLE OF CONTENTS Hi LIST OF TABLES iv LIST OF FIGURES v LIST OF ABBREVIATIONS vi ACKNOWLEDGMENTS vii CO-AUTHORSHIP STATEMENT viii CHAPTER 1: LITERATURE REVIEW 1 FOREST FIRES 1 PRESCRIBED FIRES 4 FOREST WEATHER INDEX 5 EFFECTS OF FIRES ON MICROBIAL COMMUNITIES 6 EFFECTS OF WILDFIRES ON MICROBIAL COMMUNITIES 7 EFFECTS OF PRESCRIBED FIRES ON MICROBIAL COMMUNITIES 11 HARVESTING EFFECTS ON SOILS 13 HARVESTING EFFECTS ON MICROBIAL COMMUNITIES 16 FOREST SOIL MICROBIAL COMMUNITIES - RESEARCH TOOLS 21 Extracted DNA as a Biomass Measure 21 Microbial biomass carbon to soil organic carbon ratio 22 Ribosomal Intergenic Spacer Analysis (RISA) 24 Polymerase Chain Reaction-Denaturing Gradient Gel Electrophoresis (PCR-DGGE) 25 Band Sequencing 29 RATIONALE 31 THE RESEARCH PROJECT 33 Chisholm, Alberta Wildfire 33 The Project 34 The Research Questions 35 CHAPTER 2: EFFECTS OF FOREST WILDFIRE AND HARVESTING 36 INTRODUCTION 36 MATERIALS AND METHODS 40 RESULTS 51 DISCUSSION 55 CONCLUSIONS & FUTURE RESEARCH 64 REFERENCES 78 CHAPTER 1 REFERENCES 78 CHAPTER 2 REFERENCES 85 APPENDIX 1 91 APPENDIX 2 95 iv LIST OF TABLES PAGE T A B L E L l . Summary of % change in soil chemical and physical properties from different clear-cut regimes in the Muskeg River Forest Demonstration Area in the Northwest Territories 15 T A B L E 1. Locations of sampling sites in the Chisholm-Slave Lake area 66 T A B L E 2. Comparison of the differences in RISA and D G G E community fingerprints among treatments with nonparametric multi-response permutation procedures (MRPP), based on a ranked Sorensen distance measure (Bray-Curtis method); T = description of the separation between the groups and A = description of the effect size or "chance-corrected within-group agreement" 67 T A B L E 3. Richness, evenness and Shannon diversity indices for the RISA and DGGE fingerprint data. Shannon diversity data followed by a different letter are significantly different using Bonferroni's test at/?<0.05 68 T A B L E 4. Cmjc:Nmjc a n & CmjC:C0rg values in the mineral soils of the control, harvest, burn and burn-salvage treatments (n = 10) 69 T A B L E 5. Descriptive properties of sequences from RISA fingerprint bands with high correlation values. The correlation coefficient, tau, represents the rank relationship between the ordination scores and the individual variables 70 V L I S T O F F I G U R E S P A G E F I G . L I Components of the FWI system. Daily observations of temperature, relative humidity, wind speed and 24-hour rainfall are used to generate a numerical rating scale which indicates relative wildland fire potential 6 F I G . 1 Microbial biomass-C, - N and extracted D N A in the mineral soils of the control, harvest, burn and burn-salvage treatments (n = 10). A l l values are significantly different (p < 0.05) 73 F I G . 2 RISA fingerprint NMS ordination with joint plot of N0 3 " , N H 4 + , Pext and pH superimposed indicating the relative strength and direction of correlation of variables with the ordination. Three axes produced the best ordination. A total cumulative variance of 76.1% was explained by the ordination 74 F I G . 3 DGGE fingerprint NMS ordination with joint plot of N 0 3 , N H 4 , Pext and pH superimposed indicating the relative strength and direction of correlation of variables with the ordination. Three axes produced the best ordination and a total cumulative variance of 63.9% was explained by the ordination 75 F I G . 4 Sample RISA fingerprints with arrows indicating bands that were sequenced 76 F I G . 5 Rooted neighbor-joining phylogenetic tree of sequences from RISA fingerprint bands. Reference sequences used: Uncultured eubacterium clone WD260 - AJ292673.1, Uncultured eubacterium clone WD272 - AJ292684.1, Uncultured bacterium clone 1174-1021 -18 - A B 128891.1, Uncultured alpha proteobacterium clone EBI 127 - 395446.1, Bacterium Ellin 5280 - AY234631.1, Hydrocarboniphaga effusa - AY363244.1, Pseudomonas sp. 273 -AF039488.1, Sinobacter luteus - AY966001.1, Variovoraxparadoxus -AF532868.1, Rhodoplanes elegans - AF487437.1, Bacterium Ellin 5220 - AY234571.1, Flavobacterium ferrugineum - M62798.1, Bacterium Ellin 5025 - AY234442.1, Frankia sp. - AF063641.1, BacUlus naganoensis - AB021193.1, Parachlamydia acanthamoebae -Y07556.1, Nitrospira marina - X82559.1 and rooted with Aquifex aeolicus - AJ309733.1. Numbers at the nodes are bootstrap values 50 made with 1000 bootstrap resamplings. The bar represents 0.1 substitutions per site 77 L I S T O F A B B R E V I A T I O N S Cext extractable C Next extractable N Pext extractable P D N A deoxyribonucleic acid DNAext extracted D N A rDNA ribosomal D N A rRNA ribosomal ribonucleic acid RISA ribosomal intergenic spacer analysis PCR-DGGE polymerase chain reaction-denaturing gradient gel electrophoresis Cmic microbial biomass carbon N m i c microbial biomass nitrogen C o r g organic carbon SON soluble organic nitrogen L F H organic layers primarily composed of leaves, twigs and woody materials L: original structures of organic material F: accumulated organic material, partly decomposed H : original structures of organic material are unrecognizable V l l A C K N O W L E D G M E N T S Many thanks to Dr. B i l l Mohn and all past and present lab techs, research associates and students in the lab for their endless patience, expertise and guidance through this project. It has all been a thoroughly worthwhile and fun time. I especially thank my girls, my family and my friends for their infinite encouragement, assistance and love during these years. You have all made the journey that much more rewarding. In the beginning there was no fire, and the world was cold, until the Thunders (Ani'-Hyun'tikwala'ski), who lived up in GalunTati, sent their lightning and put fire into the bottom of a hollow sycamore tree which grew on an island. — Cherokee tale in Alianor True, Wildfire: A Reader Vll l C O - A U T H O R S H I P S T A T E M E N T Some data for this thesis was received from Dr. Barbara Kishchuk, Research Scientist, for the Forest Soils Northern Forestry Centre of the Canadian Forest Service, Natural Resources Canada. Specifically, her participation has included: planning and designing the sampling strategy and, collection of soil samples. These soil samples were processed by her lab from which the following data were sent to the thesis author: soil water content, inorganic N concentrations (NO3" and N H 4 + ) , extractable phosphorus and, pH. These data were used in the ordination analyses of the RISA and D G G E fingerprint data. A l l other work was conducted or overseen by the thesis author. Thesis Authol-Thesis Supervisor 1 CHAPTER 1: LITERATURE REVIEW FOREST FIRES Fires are a natural phenomenon in forests. Return periods, which range from 15-2000 years, are governed by forest type, climatic conditions, topographic characteristics and season (1, 92). The type of fire that will burn depends on the fuel availability which is related to the type of vegetation present (5, 31). Ground fuels consist of compacted duff, roots and dead or decaying buried logs. Ground fires are slow moving as the fire burns through the 02-limited, compacted ground fuels progressing only a few feet per day. Surface fires burn above ground, igniting standing trees, downed woody materials, herbaceous plants, litter and the F layer of duff. Surface fuels are the most common mode of propagating fires, and surface fire intensities (rate of heat energy release) range from very low to very high. Crown fires spread through the canopies of trees and shrubs and are usually driven by strong winds or aided by the topography of steep slopes. Crown fires tend to be of high intensities. Plant adaptations to fire regimes are well documented and are as diverse as the fire types that affect the plants (59). Some of these survival traits are bark characteristics, with thick bark such as that found on ponderosa pine and western larch enhancing survival in ecosystems with frequent, low-severity surface fires. Thick bark provides a strong insulator and protector of the delicate tissues underneath. Growth characteristics of tree species, such as thick bark, bud scales, deep rooting systems, long needles and self-thinning crowns, lend strength to plant fire resistance in forests (79). Serotinous cones are either tightly closed cones that remain on the tree for many years or cones whose scales are sealed with a thin coat of resin that both open and release seeds when stimulated by the heat of a forest fire. Therefore, germination of some seeds requires fire stimulation and these seeds stored in serotinous cones will contribute to the regeneration of burned forests (31, 79). 2 Fires impact forest soils in several ways. Vegetation removal, litter removal and reduced albedo, the fraction of radiation striking the soil surface that is reflected by the surface, all result in increased surface temperatures (58, 68). Crown removal results in loss of interception, a process where precipitation is interrupted in its path to the soil surface (31). Changes in soil structure/porosity, hydrophobicity and major alterations in soil organic matter, which affect nutrient cycling, are other examples of changes that occur after a fire (5, 30, 32). Microenvironmental changes in forest soils caused by fire will have an affect on microbial communities. Severe heating of soil breaks down the structures of the inorganic parent materials like phyllosilicates, clays and other minerals (106). The result is a reduction of small sized particles and microenvironments where microorganisms live accompanied by an increase in larger, sand-sized, particles, resulting in a less stable soil structure. Decreases in total porosity and pore size distribution result from soil heating (31). The consequences of a loss of macropore structures are decreased water infiltration rates and an increase in overland flow, which contribute to soil erosion. Conversion of humic and fulvic acids to 'pyromorphic humus' is another structural modification found in soils that may contribute to microbial inhibition (2). The soil's ability to hold water is negatively affected by fires. Fires can create hydrophobic layers (2, 30, 32) within the soil structure resulting in decreased water infiltration and increased soil erosion from water runoff. Hydrophobic organic material that accumulates on the surface layers in litter and duff produce a water repellent layer immediately below them. The water repellency of soils caused by low- to moderate-severity fires has faster recovery times than those of high-severity fires (36, 78). Hydrophobic substances in soils can be beneficial in the maintenance of soil structure i f present in the proper concentration (46). Soil hydrophobicity also has a negative effect on plants, which provide niches for microbial colonization, by reducing the depth of root growth and hampering the process of re-vegetation. 3 Nutrient transformations occur when excessive heat is applied to soils. Whether it is a wildfire or prescribed fire (planned burning of forest land), changes to nutrients like nitrogen, which tends to be a limiting factor for many organisms' growth, are affected (23, 72). In a prescribed burn, only 10% of total N was lost in the plant, litter and upper soil layers in a chaparral forest (28); however, another study reported a 67% loss of total N in dry soils after a fire while only a 25% loss was reported in moist soils (29). Ammonium (NFL/") and nitrate (NOV) concentrations are commonly measured after fires, but consistent trends are not found for either nutrient in the literature, because the effects are governed by pre-fire concentrations of the respective nutrients, soil type, moisture and litter levels, vegetation burned, duration and intensity of the fire. Increases in NFL/ concentrations in the soil were seen following fires (23, 64); however, complete volatilization of N H 4 + (as NH 3 ) occurs during fires at temperatures greater than 500°C (35, 66). Immediately after a fire, NO3" concentrations were minimal to none, but after some time, a build-up of NO3" was observed (23, 65). This nitrification process is due to microbial activity in the burned soils (57). Post-fire, increased levels of soil nutrients, phosphorus, potassium, magnesium and calcium, were reported in a trembling aspen woodland (99). In a later study, available phosphorus (P) increased 22 times in a 250°C fire and 13 times in a 600°C fire above the control (65). In the same study, available P concentrations were still 15 times greater than the control soils 12 weeks after the fire. Finally, most studies report an increase in soil pH following fires due to the release of basic cations during combustion and their deposition on the soil surface (2, 31, 65, 87). Charcoal from burned woody species in forests has a strong adsorptive capacity for phenolic compounds. Phenolic compounds inhibit tree growth, mycorrhizal functions and other biological soil processes. These compounds are produced by certain ericaceous shrubs and can reduce forest regeneration and growth. Charcoal reactivation by microbial activity was assessed through microcosm experiments (113). Charcoal placed in sterile humus was not reactivated but 4 charcoal placed in intact humus was fully reactivated suggesting that microbial activity was crucial for the degradation of the adsorbed phenolic compounds. Similar processes have been observed for clay particles in mineral soils (108). Suppression of natural wildfires can lead to an increased domination in late-successional Boreal forests of plant species which produce phenolic secondary metabolites. Greater nitrogen availability resulting from reduced phenolic concentrations is one of the beneficial effects of adsorption of phenolics by charcoal (109). Processes such as these are key factors aiding in forest rejuvenation and are examples where it has been shown that microbes are major players in that role. P R E S C R I B E D FIRES Prescribed fires are used by forest managers for a variety of objectives. Some uses of prescribed burning are site preparation, stand composition regulation, competition reduction, insect and disease reduction, fire hazard reduction, wildlife habitat management, range improvement and water management (68, 77). Also the types of prescribed burns vary. Head fires, which spread with the wind or upslope, burn at high intensities, while, backing fires, which spread by backing into the wind or downslope, burn at low intensities. However, fast spreading head fires may disrupt the surface organic matter less than a backing fire which has a higher residence time and causes greater heat damage to surface organic matter. Flanking fires spread across the wind and/or slope. These are moderate in intensity, but, depending on the fuel availability, sideways movement may increase or decrease. Removal of coarse woody debris from harvested sites is a tactic that reduces the fire intensity on the forest floor; however, whether the fire intensity is high or low, similar effects, varied as the fire that produced them, occur. Fires are historically part of the natural ecology of forests and fire-adapted vegetation persists because of the particular historical fire regime. Loss of biodiversity can be expected when alterations are made to natural fire regimes such as prescription treatment. The capacity for these 5 fire-adapted traits to remain diminishes and natural selection processes may eliminate species which have adapted to certain types of fire regimes. It is not difficult to imagine that this biodiversity loss would also apply to other macro- and microorganisms as well. Increases in nutrient availability and pH, and decreases in moisture content post-fire treatment impact microbial compositions in forest soils. However, species loss may or may not result in significant ecological impact. If the functional loss from one species has overlap from other species in the environment, then little to no impact may be observed. However, i f the functional loss from the species is an unique quality possessed only by that one species, then the ecological impact may significantly alter the environment. There is a lack of knowledge about microbial diversity and effects to microbial communities impacted by harvest treatments and wildfire events (103). Although destructive, natural fire regimes in forest ecosystems govern the environmental successions that follow (6). Understanding the microbiological adaptations to natural forest fire regimes will aid in the development of strategies and policies that deal with biodiversity conservation and emulation of these regimes may be a management strategy. FOREST WEATHER INDEX The Canadian Forest Fire Danger Rating System (CFFDRS) was initiated by the federal government in 1925. It is used as a tool to evaluate and integrate factors, individual and combined, influencing fire danger. The CFFDRS is composed mainly of two components, the Fire Weather Index (FWI) and the Fire Behaviour Prediction (FBP) system. The FWI is built from components of weather observations that, when assessed together, generate a numerical rating that accounts for ignition potential and probable fire potential (Fig. LI) . In Alberta, a FWI > 30 is classified as extreme. Several other countries have modeled the Canadian fire danger rating system when implementing their own systems and the CFFDRS remains one of the few nationally implemented fire danger rating systems in the world. 6 Fire Temperature, Wind weather relative humidity, observations wind, rain Fuel moisture codes Fire behavior indices Temperature, relative humidity, Temperature, rain I- ire weather Index (FWI) Fig. LI. Components of the FWI system. Daily observations of temperature, relative humidity, wind speed and 24-hour rainfall are used to generate a numerical rating scale which indicates relative wildland fire potential. (Natural Resources Canada: Canadian Forest Service http://fire.cfs.mcan.gc.ca/research/environment/cffdrs/fwi_e.htm). EFFECTS OF FIRES ON MICROBIAL COMMUNITIES One of the earliest studies which assessed microbial populations over two years from soils where timber and brushwood were piled in heaps and burned was by Corbet (1934) (22). At this point, it was already known that differences occurred in microbial numbers at the diurnal level, corresponding to temperature differences, although no correlation could be found for changes in microbial numbers and soil moisture content. One aspect of Corbet's study, in the equatorial regions where monthly temperature variations are negligible and rainfall is high, hypothesized that under these conditions, microbe numbers would find some constant value. A second aspect of his research sought to verify previous research on the anticipated changes in soil microbes following harvesting and clearing of tropical forests. Large increases in microbial populations were found in short periods following the burning operations, while at other times populations were uniformly low. It was suggested that the presence of the wood-ash on the soil influenced 7 the rapid growth of the soil microorganisms following the burning period. A l l colonies were counted on the plates, which comprised fungi and bacteria, and it was noted in the earlier stages of the experiment that more fungi were present compared to bacterial colonies. No mention was made whether this trend changed after burning. EFFECTS OF WILDFIRES ON MICROBIAL COMMUNITIES Wildfires produce extreme and long-lasting effects on soil bacterial communities. While prescribed burns have low to moderate fire intensities, natural wildfires are much different (103). Heat intensities in wildfires, nutrient vaporization and mineralization of soil organic matter mostly appear to surpass the damage level that occurs during a prescribed fire. These characteristics may account for the sustained effects that existed in a subalpine forest that was burned 6 years previously and still demonstrated statistically significant changes in its bacterial and fungal biomasses (13). The immediate effects of a moderate intensity bushfire in Australia were mostly confined to the upper 0-2 cm layer of soil (105). Initially, significantly lower microbial counts were observed within the first month after the burn. Microbial numbers 2 and 7 months post burn were markedly above those found in the unburned treatment. After observance of these increased abundances of microbial numbers, counts in the burned treatment decreased until approximately 20 months post burn, when the microbial numbers were similar to the unburned treatment. The fluorescent pseudomonads, part of the y-Proteobacteria subdivision, displayed the most sensitive reactions to the burn treatment which may indicate that this subdivision of bacteria would be the best indicators of soil disturbance post burn treatments. Soil microbe changes were evaluated one month and one year post wildfire in a Maritime pine forest soil (107). Bacterial populations one month post-fire ranged from 2 to 16 times greater in the burned soils compared to the unburned soils, while photoautotrophs, algae and fungi decreased 19 to 126 times in the burned soils. One year post-fire, the aerobic heterotrophic and 8 acidophilic bacterial populations decreased significantly and, in contrast, the photoautotrophs and algae increased in abundance 51 and 10 times, respectively. The populations oi Bacillus spp. increased significantly in the burned treatment one year post-fire. Most fires result in increased pH and nutrient levels in burned soils; hence, initially bacterial growth was enhanced over fungal growth, due partly to the increased nutrient status. The massive fall in the fungal population one month post-fire also shows that the bacteria were better able to survive soil heating compared to the fungi. A post-fire chronosequence was conducted in Aleppo pine forest soils to describe aspects of microbial biomass changes in soils that experienced wildfires (34). Initial decreases were seen in C m ic and a slight increase in N m j c one month post-fire. The C m j C to N m j C ratio is used as a crude measure to describe structures of the microbial communities with large ratios indicating fiingal-dominated communities and smaller ratios indicating bacterial-dominated communities (4). Cmic:NmJc decreased from 6.5 to 3.6 suggesting a definite fire stress on the fungal community. Large decreases in fungal relative abundance were observed in the one month post-fire treatment, as is commonly found in fire treated soils (73,105,107, Aida E. Jimenez-Esquilin, personal communication). One year post-fire and beyond, C m i C : N m j C values were returning to control levels, and both C m j c and N m j C content steadily decreased in the burned treatments over the 11-year chronosequence. Evidently, the effects on soil bacterial communities that are impacted by high intensity fires have long-lasting consequences. A wildfire in a Maritime pine forest in Spain had detrimental effects on the microbial population in the soil surface layer, but only reduced the microbial population by 50% in the subsurface layer (91). Differences in microbial biomass were reduced in subsequent years and any negative effects of burning were barely detectable 4 years after the fire. For the 4 year time period after the fire, burning explained the variation in C m j c and N m i c , but from year 5 to year 13 soil depth accounted for a significant proportion of the variation in microbial biomass. 9 Following a wildfire and salvage-logging treatment, a parallel analysis of bacterial diversity associated with ectomycorrhizae was conducted via an rDNA amplicon analysis from isolates from standard culturing techniques and rDNA amplicons obtained by direct D N A cloning (61). Four to five years post-fire, the ectomycorrhizae-associated bacterial community showed no significant difference from the control community. Like most microbial community studies post-fire treatment, gram-positive bacteria still had high numbers several years later contributing to the bacterial composition in the burned treatments. Clone analysis determined that restriction patterns of many clones did not match the 16S rDNA of any of the cultured isolates; hence, a high percentage of clones were classified as unidentified bacteria. Such parallel research techniques demonstrate the duality of information obtained and suggest the importance of the use of various methods to study natural, microbiologically unknown ecosystems. Wildfires had adverse effects on soil microbial communities in coniferous forests in Hiroshima prefecture (73). Less than half of the control site's C m j C was reported in both burned treatments; one burned two years prior to sampling; and, the other sampled immediately post-fire. The gram-negative bacteria appeared to be the most drastically affected, yielding less than half of the colony-forming units than the control site. The gram-positive bacteria also yielded less than half the colony-forming units of the control site for approximately the first year, then a sharp increase in colony-forming units was observed on solid media from the 2-year and immediate post-fire samples. Significant differences in the fungal counts were noted for all three sample sites, although a sharp increase in the fungal counts during the third summer post-fire was observed, closing the gap between the control site and the site burned two years prior. A l l microbial counts, gram-negative, gram-positive and fungal, were still lower than the control site values at the end of the two year study. Changes in the nitrogen-fixing and ammonia-oxidizing soil bacterial communities were observed in fire-impacted fir and pine forests in New Mexico (110). While 16S rDNA genes 10 were always successfully amplified from all samples, PCR amplification of nifH and amoA was less successful from the burned sites, suggesting that the proportion of the nitrogen-fixing (based on nifH sequences) species and ammonia-oxidizing (based on amoA sequences) species decreased due to negative impacts from the fire. The dominant nifH sequence types associated with the burned soils were affiliated with Clostridium spp. and Paenibacillus spp., both spore-forming taxa, and a number of other rare nifH fragment types were observed in the burn treatments. The ammonia-oxidizing community demonstrated a large shift in composition from dominance of one genotype to dominance of a different genotype after the wildfire. One molecular research project currently underway is analyzing the microbial structural characterization and diversity of a fire-ravaged forest soil in the Hayman fire site in southern Colorado, U S A (Aida E. Jimenez-Esquilin, personal communication). Ester-linked fatty acid methyl ester (EL-FAME) analysis, together with denaturing gradient gel electrophoresis (DGGE), are being used to trace the changes in the viable bacteria and fungi and analyze changes in the overall community fingerprint patterns. Preliminary E L - F A M E results demonstrate that the microbial fatty acids, representing different viable divisions of bacteria and fungi, were substantially changed by the fire. Also, D G G E fingerprints have demonstrated differences among the various treatments, with unique bands possibly representing organisms distinctive to the disturbed treatments. Fire impacts to microbial communities in forest soils across studies are varied. However, similar trends exist, such as significant reductions in microbial biomass, especially in the upper layers of soil horizons. Direct killing of microorganisms is the result of severe heating of soils. Soil moisture content, i f high, can increase the heat transfer down into lower layers of soil, resulting in greater microbial biomass loss, while i f the soil is dryer, then microbial biomass loss will not be as great, because dry soils are effective heat transfer insulators (17, 31). Heat from fires appears to be more detrimental to the fungal component in soils than the bacterial one (73, 11 105, 107, Aida E. Jimenez-Esquilin, personal communication). Ash deposition on forest soils increases pH levels, which also appears to be unfavourable to fungi and favorable to bacteria. Increases in Bacillus spp. are common in fire-treated soils, most probably due to the structural integrity of the gram-positive cell wall (73, 105, 107) or their ability to form heat-resistant endospores which ensure their survival. The most sensitive division of bacteria in response to fire treatment in forest soils appears to be the y-Proteobacteria (105). Novel restriction patterns of direct 16S clones that did not match restriction patterns of cultured isolates were a result following a wildfire and salvage-logging treatment, suggesting that the burned soil had increased bacterial diversity (61). The changes seen in microbial communities from fire-treated forest soils can be long-lasting; however, some groups found recoveries of microbial compositions 4 years post-fire (91), while others still saw significant differences 11 years post-fire (34). Finally, while most studies have used a traditional culturing approach to study the microbial composition in burned forest soils, only recently have culture-independent techniques, for example E L - F A M E and DGGE fingerprinting methods, been employed to analyze changes in microbial community compositions in forest soils (Aida E. Jimenez-Esquilin, personal communication). EFFECTS OF PRESCRIBED FIRES ON MICROBIAL COMMUNITIES Changes to the microbial biomass in forest soils subjected to prescribed burns are variable. Fire-induced changes to the soil C m j C and N m j C in a predominantly Scots pine (Pinus sylvestris) forest resulted in decreased values, below those of the control treatment (87). Mean CmiC:NmiC values in the prescribed burned area 3-35 days post-fire were higher than the control treatment suggesting changes in the microbial community occurred. Drastic reductions in microbial biomass measured by phospholipid fatty acid (PLFA) content were seen in fungal communities (63% and 79%) compared to bacterial communities (27% and 40%) in two sites that were subjected to a prescribed burn even though the specific respiration rate was highest in both burned treatments (44, 86). In comparison, some studies found no effects on the microbial 12 fraction or only decreased microbial biomass that was not statistically significant in soils that experienced a prescribed fire (38, 45), while others only found fire effects in the upper soil layer (12, 53). Prescribed fires result in less nutrient losses compared to wildfires, due to reduced intensities; however, one prescribed fire resulted in more adverse affects compared to a simulated forest fire (87). Field et al. (2003) did a site preparation prescribed burn in a loblolly pine forest and found increased concentrations of nutrients in runoff, which peaked after 3 months (40). Total nitrogen levels decreased by volatilization due to the fire and changes were seen in the soil's physical properties and nutrient cycling. Even though the fire was of low intensity (above ground temperature) and light severity (heat penetration into the soil), nutrients, such as NO3", NFL;"1", PO43", K + , M g 2 + and C a 2 + were lost in soil runoff. Slight increases in NO3" in the control site runoff were found after the burn treatment. The authors' attribute this to wind-blown sediment contamination in the control site on account of the dry period that immediately followed the fire. No mention was made of any microbial process that would have contributed to these NO3" increases, and they suggested that these increases may be accounted for by sediment nutrient levels which were never measured. Prescribed burns in two different studies, a mixed oak-pine forest and a jack pine stand, produced similar results (63, 71). Both studies followed soil changes for 5 or 10 years, respectively. In the Knoepp study, increases in Ca 2 + , K + and M g 2 + were observed post treatment in the A horizon, while increases in C a 2 + and K + were observed in the B horizon. The increases in cation responses in the B horizon were ascribed to leaching. No changes in soil C and N were attributed to the prescribed burn treatment; although, total C and N amounts varied in both the A and B horizons. Lynham studied soil changes and early vegetation succession post experimental burn and found that burning increased P e x t , K + , C a 2 + and M g 2 + in mineral soils and that the P e x t and K + were still greater than the pre-burn levels 10 years post-fire. Total N also increased in the 13 mineral soils and was still increasing 10 years later. These total N increases were greater in the 5-10 cm depth compared to the 0-5 cm depth. Both studies reported increases in pH which is common in all forest soils post-fire treatments. Generally, reductions in microbial biomass C and N are observed in soils after prescribed burns, and the reductions seen in the fungal component of microbial biomass are greater than the reductions in the bacterial component (44, 86, 87). However, some studies found no effects on the microbial community composition, while others found effects only in the upper layers of the soil (12, 38, 45, 53). H A R V E S T I N G E F F E C T S O N S O I L S Changes in soil chemistry are induced by harvest treatments in forests. Wider fluctuations in soil temperature and moisture content due to unimpeded exposure to the elements occur in the forest floor following harvest treatments with generally higher soil temperatures and lower moisture levels (43); however, another study reported increased soil moisture post harvest treatment (67). Inconsistent results have been reported for concentrations of soluble organic nitrogen (SON) in soils following clear-cut treatments. While some research has reported increased concentrations of SON in a clear-cut site compared to its control (101), others have reported decreased concentrations of SON (51). The decreased concentrations were attributed to removal of the F layer at one site, which is rich in SON, or to a reduction in the forest floor at another site. Increased N mineralization was observed in clear-cut sites of a Norway spruce stand (101). The chemical nature of the dissolved organic matter was determined after incubation experiments and increases of hydrophilic N-containing acids were found, which are microbiologically derived, in the clear-cut sites but not in the undisturbed forest sites. There was also a concomitant decrease in hydrophobic substances. It was presumed that the changes in the chemical nature of the dissolved organic matter composition were due to an increase in microbial 14 activity in the clear-cut sites. Other studies have reported increased availability of magnesium in clear-cut treatments (43) as well as short term increases in N H / (11). Quantification of the effects of different clear-cut regimes, a patch clear-cut, a strip clear-cut and a clear-cut, was conducted in the Muskeg River Forest Demonstration Area in the Northwest Territories (14). Nine soil properties were assessed, eight of them in the L F H and the mineral soil horizons. Test results are summarized in Table L l , but generally, increases in all the measured parameters were observed in the patch clear-cut and the clear-cut. The decreases were attributed to the addition of twigs and needles of high C/N content at the time of the harvest, the disruption of the litter input and canopy wash, increased decomposition and initial rapid mineralization and N leaching. The clear-cut treatments were conducted in the winter months, therefore, minimal physical disturbance of the soil surface occurred. Aside from evidence of accelerated organic matter decomposition and nutrient leaching into the mineral horizons, the changes observed did not significantly impact the majority of the soil properties measured 3-4 years post-harvest treatment. 15 Patch clear-cut Strip clear-cut Clear-cut Soil property % change % change % change LFH thickness 10 12 3 Bulk density LFH 12 0 12 Mineral soil 1 3 7 pH LFH 0.2 units 0.2 units 0.0 units Mineral soil 0.4 units -0.1 units 0.5 units Total organic carbon LFH 3 4 4 Mineral soil 6 9 30 Total N LFH -6 -9 -6 Mineral soil 17 33 33 Cation Exchange Capacity LFH 16 -10 -2 Mineral soil 20 -3 24 Exchangeable Ca 2 + LFH 1 1 -2 Mineral soil 26 -10 43 Exchangeable Mg 2 + LFH 33 -2 8 Mineral soil 43 -20 44 Exchangeable K + LFH -24 -8 -32 Mineral soil -6 0 -50 Table LI. Summary of % change in soil chemical and physical properties from different clear-cut regimes in the Muskeg River Forest Demonstration Area in the Northwest Territories. Changes due to harvesting appear to be relatively minor, but would have an effect on the microbial community composition of the native microbes. Most increases in soil parameters are insignificant; however, the fluctuations appear to depend on the initial conditions of the affected site. Consistent trends are seen in decreases in total N in LFH horizons and exchangeable K + in LFH and mineral horizons, while increases in bulk densities and pH in LFH and mineral horizons and total N in mineral horizons are observed. Indeed, these small changes in soil properties precede changes in the microbial component which, although slight and short-lasting, do occur. 16 HARVESTING EFFECTS ON MICROBIAL COMMUNITIES Although to a lesser extent than fires, harvesting treatments effect changes in soil microbial communities. In an earlier study assessing the microbiological effects after a clear-cut and unsuccessful reforestation, changes were observed in the physiological traits of isolates from a spruce stand soil (83). Factor analysis highlighted changes that appeared to occur at the four year mark post clear-cut, and specifically, one factor detected microbial population changes that occurred in the mineral and humus soil layers. When analyzing the similarities between samples in the mineral and humus populations separately, the 4- and 7-year populations were distinctly different from the control populations but the 13-year populations indicated a return to the control population state. Even now that we know that ca. only 1% of the microorganisms in environmental samples are currently cultivable, the changes affected by harvest treatments were observable even with the small community that was able to be isolated. However, even though reversion to control community physiological state was observed, it is not known whether this community, at a genus or species level, is the same as the control community, only that the community which developed was able to produce similar results in the biochemical tests conducted. Clear-cutting affected active fungal hyphal length and biomass in a Scots pine forest (9). Three treatments were tested: control, clear-cut with slash removed and clear-cut with slash left. There was a general trend of decreased fungal length and decreased biomass in the clear-cut treatments with larger reductions in the clear-cut site with slash removed. Three soil horizons were analyzed, the A01/A02, A2 and B horizons, and the greatest reductions were observed in the mineral horizons (A 2 and B). Removal of trees results in eventual cessation of root exudation and any fungi that rely on mycorrhizal associations with tree root systems would be compromised. Mycorrhizal fungi may account for the majority of fungi in the mineral horizon and, hence, were lost due to tree root death. 17 The effects of harvesting on bacterial biomass in the same forest project as the above-mentioned study were assessed by microscopic counts using acridine orange and fluorescein diacetate (70). Bacterial biomass in the humus layer of the clear-cut treatment with slash left was greater than the clear-cut treatment with slash removed but both clear-cut treatments were greater than the control site. This trend continued for 3 years post harvest, after which bacterial biomass measures in the control site exceeded those in both clear-cut sites for the next two years. In the A 2 and B layers of the mineral soil, both clear-cut treatments supported bacterial biomass greater than the control site for all 5 years of the study. pH measurements in the humus layer ranged from 3.9 to 4.4, while in the A2 layer pH ranged from 4.3 to 5.2. The pH in the control treatments for both of the above layers fell approximately in the center of both its respective ranges. Hence, pH changes did not appear to significantly impact the fungal or bacterial components in these studies. On the contrary, the addition of nutrients via harvesting residue appeared to significantly impact the bacterial biomass in this study, which was in contrast to the fungal component study of this forest project. Increases in bacterial biomass may have occurred due to the reduction of competition with fungi for nutrient sources. Seasonal variations in microbial biomass occurred with peaks in spring and fall months in a Montana subalpine fir habitat (37). Depending on the type of harvest treatment (clear-cut -residue left, clear-cut - residue removed, clear-cut - residue broadcast burned) differences in amounts of bacterial biomass were observed. However, all treatments followed the same trends with peaks in spring and fall months, which coincided with periods of high soil moisture. The clear-cut - residue left site sustained the highest amount of microbial biomass even over the control site and had a higher bacteria to fungi ratio. All harvest treatments increased the pH of the soil except the clear-cut - residue removed site. Sites with increased pH appeared to support increased bacterial biomass over fungal biomass. 18 More recently, studies have focused on methods that characterize the microbial communities in forest soils based on sole carbon source utilization (SCSU) or specific culture media, fatty acid methyl ester (FAME) analysis and DNA sequence analysis of cloned 16S rRNA gene fragments (7, 8, 104). These methods enable assessment of microbial community structure resulting in a better understanding of microbial functional capabilities and taxonomic relationships. While organic soil horizons exhibited higher microbial diversity than mineral soil horizons, within one mineral horizon, clear-cut treatments resulted in higher diversity than burned, scarified or control treatment sites (104). Results from SCSU experiments suggested that only a subset of the carbon sources produced significant and biologically meaningful data. Cultivation studies determined changes in bacterial community composition in harvested sites with no soil compaction and those that had heavy soil compaction plus removal of all surface organic matter (7). Although Arthrobacter was frequently isolated from the heavily compacted harvested soil treatment, it was not found in a companion molecular characterization study from the same treatment (8). This suggests that Arthrobacter is not a predominant genus in this treatment, but rather is easily isolated via a cultivation approach. A broader representation of bacterial divisions was determined from the molecular study than that from the corresponding cultivation study, which examined 1.7 times more organisms than the molecular study, but found only one-third the number of bacterial divisions (8). Coupling cultivation-independent approaches to a cultivation-dependent approach uncovered results that would not have been found otherwise. Certainly, the bacterial divisions, Verrucomicrobia and Acidobacteria, which have proven elusive to cultivation, have become recognized as being predominant only since the onset of cultivation-independent studies. The effects of clear-cutting on the ammonia-oxidizing bacterial community were examined in limed and non-limed plots in a Norway spruce site in the central part of southern Sweden (11). A D G G E analysis was conducted using $-Proteobacteria primers specific for members of the 19 genera, Nitrosomonas and Nitrosospira. Two to three bands were isolated and were subsequently labeled NScl2, NScl4 and SpX. Both bands labeled NS- were affiliated with Nitrosospira and the SpX band was affiliated with Herbaspirillum chlorophenolicum, a p-Proteobacterium. Before the clear-cut, only NScl2 was found in the non-limed plot, but after the clear-cut, both NScl2 and NScl4 were found in all plots. In the limed plots, NScl2 and NScl4 were found both before and after the clear-cut and the non-ammonia-oxidizer-like SpX organism was found in the non-limed plot only. The relative abundance of each ammonia-oxidizing population was estimated, assuming that the ratio between the amplified products was in proportion to those ammonia-oxidizers in the template D N A pools. NScl2 increased slightly shortly after the clear-cut and in the second year when NScl4 appeared, the relative abundance of NScl2 decreased by ca. 50%, leaving the two groups representing the ammonia-oxidizing populations in equal proportions. Later in the same year, the total size of the ammonia-oxidizing community was at a maximum so that the total size of this community was 50% larger after the clear-cut than before. Significant correlations were also found between the total ammonia-oxidizing community and potential nitrification rates. Therefore, clear-cutting affected the ammonia-oxidizing community in both limed and non-limed plots and nitrification rates were related to the shift in that community. Changes in forest soils due to harvesting treatments effect changes in microbial communities. Earlier work realized changes in microbial communities with return to control physiological state 13 years post harvest; however, organisms responsible for the physiological state may have been different species or genera with overlapping capabilities as those in the control community (83). Fungal hyphal-length reductions were observed in soils post clear-cut with the greatest reduction observed in mineral soil horizons (9) and bacterial biomass in the humus layers in the clear-cut site exceeded that determined in the control site for the first 3 years of the study and for the latter portion of the study bacterial biomass in the clear-cut site was less than the control. In the 20 mineral horizons, bacterial biomass in the clear-cut site exceeded that in the control site for the duration of the study (70). Microbial biomass peaks in spring and fall were observed after a clear-cut treatment. A l l clear-cut treatments increased pH and bacterial biomass over the control site except in the clear-cut - residue removed treatment (37). Culture-dependent and culture-independent studies can complement each another, but sometimes contrasting information is determined (7, 8, 104). Changes in bacterial numbers, and possibly bacterial associations, are closely tied to environmental conditions, as was found in a DGGE analysis of the effects of clear-cutting on the ammonia-oxidizing community in a forest soil (11). Microbial changes in forest soils following natural and human disturbances are inevitable results of the extreme changes in nutrient availability, pH, soil moisture levels and temperature. Past studies have examined these microbial changes through traditional culturing techniques and by measuring C m j c and N m j c . Newer molecular methods provide greater insight into the effects of disturbances on bacterial communities in forest soils and also point to which environmental factors are most responsible for these changes. Sequencing studies have revealed clades of organisms unknown to our existing knowledge of phylogeny of bacteria, while molecular fingerprinting methods are sensitive tools that allow for the unveiling of these new organisms. However, the complicated nutritional and environmental requirements of soil bacteria remain lacunae in our understanding of this underground life. More work is needed relating environmental parameters to bacterial community composition, function and evolution, so as to appreciate the intricacies of the microbial world, to predict human impacts on bacterial community composition and to direct forest management practices towards a 'close to nature ecosystem' approach. 21 FOREST SOIL MICROBIAL COMMUNITIES - RESEARCH TOOLS Extracted DNA as a Biomass Measure Using D N A as a bio-material to measure microbial biomass is problematic. Studies have shown that when D N A is bound to montmorillonite clay and sand it is protected from degradation by nucleases (3, 69, 95). Alvarez also demonstrated that when D N A is bound to clay it can still be amplified by PCR, which can be good for studying palaeontological D N A or for providing genetic material to an appropriate host, however, this not only contributes to problems in estimating microbial biomass, especially i f the extracellular D N A is desorbed from the mineral matrix, but also can contribute to irrelevant bands in gels in fingerprinting assessments of microbial communities. Another study examined humic acid binding of D N A and its effect on transformation of Bacillus subtilis and DNase resistance (24) and found reduced levels of transformation occurred in the presence of DNase and humic acid-DNA complexes. Indeed, there appear to be competitive strategies that involve persistence of free D N A in soil environments; however, these strategies can cause inaccuracy in microbial biomass measurement results. In a study evaluating three commercial D N A extraction kits, FastDNA® SPIN Kit had the most reproducible recovery efficiencies and yielded the highest D N A recoveries. However, plasmid D N A added as whole cells or as purified D N A was recovered with similar efficiency with all kits suggesting that extracellular D N A in soils and sediments will contribute to biomass determinations and complicate PCR results (80). In contrast, some studies have found good agreement between D N A and R N A or double-stranded D N A and microbial biomass C (26, 75), but in both of these studies, more than one microbial biomass measurement technique was used. In another study, D N A concentrations and cell number estimates were strongly correlated in glacier forefield soil samples, suggesting that DNAe X t from the glacier forefield samples was a reasonable predictor of microbial biomass (97). So, while several authors continue to use D N A -based techniques to report microbial biomass estimates, methods such as fumigation-extraction 22 and substrate-induced respiration provide more reliable figures of biomass measure and are adopted by national authorities for routine soil microbial biomass surveys (74, 76). Finally, it is a sound practice to employ more than one method to estimate microbial biomass i f only to confirm results of microbial biomass determinations (55). Microbial biomass carbon to soil organic carbon ratio Using the microbial biomass carbon (C m ic) to soil organic carbon (C o r g ) ratio to monitor soil fertility or degradation was suggested as an extremely useful index for land resource managers, research scientists and other interested parties in the long term trends of soil organic matter (102). Changes in soil organic carbon due to perturbations are gradual and difficult to monitor due to high background carbon levels and natural heterogeneity of soils. Conversely, C m i C has a faster recovery time following soil perturbations than does C o r g , due to the rapid turnover rate of microorganisms (55). A n investigation of C o r g , Cmic and C m i ^ C o r g o n s e v e r a l soil types under varying climate-soil, vegetation and cropping parameters revealed consistent trends for the ratio while either C o r g or C m i C alone was more variable. To be an effective monitoring index, and, since the ratio appears to be affected by soil type and vegetation, an empirically determined reference value, or baseline, is required for each type of soil to establish an equilibrium range. Cmj C : C 0 rg provides a more sensitive index to monitor soil dynamics than do either C m i c or C o r g alone, and C m i C : C 0 r g is often referred to in the literature as an indicator of microbial activity and accumulation/degradation of soil organic matter (52, 91, 102). Low Cmic to C o r g ratios were reported in forest soils (102, 114). This may indicate low substrate quality or low substrate availability for soil microbes due to physical barriers, such as hydrophobic layers, in the soil matrix. A chronosequence on continuous cropping practices with maize using C m i c to C o r g ratios demonstrated a smooth, gradual decrease compared to the erratic overall decline in the individual C m j C and C o r g data (102). C m i c to C o r g ratios of soil returned to 23 pasture following maize cropping rapidly recovered to the level of continuous pasture soils. The inconsistent patterns of the Cmic and C o r g data gave inconsistent trends when considered alone. To examine the long-term effects of clear-cutting and site preparation in the boreal forest on microbial processes related to soil organic matter, the C m i C : C o r g ratio was used. Seven to nine years post-harvest, adjacent forest soils, harvested and unharvested, were compared to assess the impact of the harvest treatment on microbiologically mediated processes (54). In the mineral layer, while C m j c was not significantly different between the two treatments, C m i C : C o r g was lower in the harvest site indicating larger quantities of organic carbon even though slash residues had been mechanically pushed aside and removed from the soil sample sites. Apparently, organic soils had been slightly mixed with mineral soils at the harvest site caused by the machines used during the site preparation. No significant differences were evident 7-9 years post harvest between the clear-cut and control sites for C m i c : C o r g or the other measured microbial process indicators except respiration which was slightly lower in the organic soil layer in the clear-cut sites. The effects of wildfires on soil and the microbial component in those soils were investigated 9 months after a wildfire event in a Mediterranean pine forest (52). A gamut of chemical, biological and enzyme analyses were conducted on unburned and burned soils collected from the top 5 cm of each sample site. Values of C m i c : C o r g were significantly lower in the burned sites compared to the unburned sites. This decrease was mostly due to the extreme effects the fire had on fulvic acids, carbohydrates, lipids and other simple organic compounds, the most labile carbon fractions. Of all the measured factors, increases were seen in electrical conductivity, pH, available P, available K + , N0 3 " and N H 4 + , while decreases were seen in total P, total organic C , C m j C and all enzyme activities. Prieto-Fernandez et al. (1998) investigated the changes over 13 years in microbial and extractable carbon ( C e x t ) and nitrogen (Ne xt) brought about by a wildfire in a pine forest (91). 24 Past studies evaluated the effects of prescribed fires on microbial biomass (42, 86, 88), and because wildfire intensities greatly surpass those of prescribed fires, changes in microbial biomass, C e x t and N e xt were interesting parameters to follow. Initially, C m i C in the burned sites was 4% of the unburned site at the 0-5 cm depth and 54% of the unburned site at the 5-10 cm depth. This is consistent with other studies that observe greater microbial biomass loss in the upper soil horizons post-fire compared to losses in the deeper soil horizons (12, 53). After one month, losses were 27% at the 0-5 cm depth and 62% at the 5-10 cm depth of the unburned site. This trend continued and at 13 years post-fire C m i C and NmjC in the burned sites averaged roughly half of those values determined in the unburned sites. Reduced C m i C : C o r g values during the first 5 years were generally associated with decreased amounts of labile carbon post-fire, but positive correlations between C m i C : C 0 r g and C m j C suggested that other factors adversely affected C m i C values after the fire. After the 5-year mark, no significant effects of the fire appeared to affect C m i C : C 0 r g , rather soil depth appeared to be the only factor that most affected changes in this ratio. The relevance of C m j C : C o r g as a monitoring index of soil health can only be seen if previous and subsequent determinations are made from the exact same sample sites, but because this ratio considers the rapidly changing microbial biomass against the background of gradually changing soil organic matter, changes in a soil environment can be detected before they are even detectable in the soil component alone. Ribosomal Intergenic Spacer Analysis (RISA) One DNA molecular tool used to analyze bacterial, fungal and archaeal community structures is RISA. This method differentiates the bacterial D N A e X t in samples utilizing the length heterogeneity of the ribosomal intergenic spacer (RIS) region between the small (16S) subunit and large (23 S) subunit rRNA genes, which is assessed as a fingerprint band pattern in acrylamide gels. RISA was first used to assess the variability in the bacterial community structures in a mature forest soil and an adjacent pasture soil (15). Analysis of 100 16S rRNA 25 clones derived from each of the mature forest and pasture soils was insufficient to describe the changes in the microbial composition between the two sites; hence, RISA was developed to rapidly compare the different microbial fingerprint patterns between the two soil types. Different RISA fingerprint patterns were determined from each site indicating unique phylotypes existed in each soil type. It was presumed that these differences arose due to changes in soil chemistry that was associated with the conversion of the mature forest to pasture land. Bacterial communities were characterized in two Swiss glacial forefronts using RISA and DGGE (97). Microbial diversities, calculated from numbers and intensities of bands in RISA and DGGE gels (discussed later), decreased as the distance from the glacier terminus increased. Geological history of the two sites was similar, while the exposure, slope, valley width and vegetation differed. Similar trends in band numbers and diversity measures between RISA and DGGE analyses were evident, although, RISA consistently produced lower values in both measures than DGGE. RISA has also been used to assess the variability of microbial community structures in wastewater treatment systems (100, 111). In both of these studies, cluster analysis, based on percent similarities of RIS-length polymorphism, quantitatively assessed the variability of the microbial community structures of the biomass in two aerated systems of pulp mill effluent. Coupled with an analysis of RIS-restriction fragment length polymorphism and rDNA sequencing, assessments of the bacterial community compositions in these two systems were achieved. Polymerase Chain Reaction-Denaturing Gradient Gel Electrophoresis ( P C R - D G G E ) PCR-DGGE was initially developed to analyze DNA fragments from single organisms (82). Now, PCR-DGGE is a popular molecular fingerprint tool used to profile complex microbial populations (81, 98). PCR-DGGE produces distinct banding patterns based on the amplified DNA fragment's mobility in denaturing gradient gels, which is governed by both G/C content and distribution within the amplified DNA region. DGGE has been used in assess microbial 26 variability in a range of environmental systems from hydrocarbon-contaminated seawater and freshwater cultures (16), changes in community structures in a bioremediation project (85), distributions of bacterial populations within a hot spring microbial mat (39) to an analysis of stool microbiota of hospital patients undergoing antibiotic therapy (33). The utility of DGGE has broadened from single organisms to whole bacterial communities despite flaws inherent in the method. Kisand et al. (2003) investigated the cause of wide, fuzzy bands on their PCR-DGGE gels from environmental samples (62). Flavobacterium isolates, 7 out of 10 times, showed "fuzzy band" morphology on DGGE gels. Direct sequence analysis of the 580 bp 16S rDNA amplicons showed multiple melting domains (MMD), regions interspersed along the amplified DNA fragment with 'step-wise' melting characteristics, between 69-72°C. These regions corresponded to a denaturing gradient between 28% and 35% urea and formamide at 61°C. MMD are common features in their subset of Flavobacteria and these MMD may exist in other taxa as well. Fuzzy or wide band patterns lead to erroneous analysis of fingerprint patterns because smeary bands may be overlooked as background or an organism's total contribution to a fingerprint pattern may be underestimated. Kisand et al., suggested shortening the length of the PCR amplicon, thereby reducing the probability of including MMD, to alleviate the fuzzy band problem. Resolution improved with shorter PCR amplicons but, because of less sequence data, phylogenetic information was lost. Another problem was a convergence of melting temperatures among isolates, therefore, bands were products of larger numbers of phylotypes or, in other words, there was less resolution. Kisand et al. (2003) also tested the accuracy of DGGE for 7 highly similar §-Proteobacteria by comparing phylogenetic analysis of sequence data to melting temperature clustering patterns (62). Different relationships for melting temperature clustering patterns and phylogenetic analysis existed. This was interpreted to mean that the DGGE banding patterns did not 27 accurately reflect the phylogenetic relationships of the taxa used. This group concluded that D G G E profiles were not sufficiently accurate for analyzing species composition when using universal primer pairs in communities of unknown diversity. However, D G G E analysis can provide information about changes in the relative abundances of the dominant phylotypes; hence, monitoring the variability of bacterial community composition. Three years following a clear-cut, the ammonia-oxidizer population was investigated in a Norway spruce stand (11). Since it was shown that N mineralization increases after clear-cutting (90), it was hypothesized that changes would occur in the ammonia-oxidizer community of these spruce forest soils. The plots were limed and non-limed, with the clear-cut treatment following 8 years later. Generally, results showed a strong correlation between potential nitrification and a shift in the ammonia-oxidizer community. Of interest, P-subgroup ammonia-oxidizer-specific primers were used to amplify this group from the soil samples, and the amplicons were electrophoretically separated by DGGE. Fingerprint patterns resulted in 2-3 bands in a narrow denaturant range and another fuzzy band found between the clear, distinct bands. Sequencing determined that the fuzzy band had the exact same sequence as the lower clear band; hence, the former was not a unique sequence. It is possible that this fuzzy band was a heteroduplex, a product of two strands of amplified D N A from different organisms (56), with the template strand being that of the lower clear band. This raises the point of another mode of bias that can result using molecular fingerprint methods, so it may be prudent to empirically determine which span of the 16S rRNA gene is most effective for the analysis that needs to be done. The variable (V) regions within the 16S rRNA gene were systematically assessed to justify the choice of which V region to include in a PCR-DGGE analysis of microbial community structure in rumen digesta (112). Their experiments showed that the choice of V region had serious implications for the substantive information of the data retrieved. Depending on the V region chosen, differences in band intensities and a range of richness indices from 14.3 to 31.3 28 were seen. This resulted in contrasting evenness indices where the evenness indices for the single V regions amplified (VI, V3 and V8) were lower than those for the multiple V region profiles (V1-V3, V3-V5, V4-V5 and V6-V8). It was recommended that the V3 region be routinely used in PCR-DGGE analysis, or if longer amplicons were desired, then the V3 to V5 or the V6 to V8 regions should be used. Similarities in soil bacterial divisions and those of rumen and gastrointestinal tract microbial flora such as Firmicutes, y-Proteobacteria and Actinobacteria suggest that their recommendations can be followed in a soil system with a likelihood of success. Both methods, RISA and DGGE, are plagued with procedural problems that influence the results of the data in various ways. Efficient DNA extraction (discussed previously) and effective PCR amplification (89) are both well documented areas of bias. Another serious problem using molecular fingerprint methods is the fact that most bacteria have multiple copies of the rRNA (rrri) genes (27, 50) and within one bacterium, sequence heterogeneity often exists (25). For example, Bacillus subtilis has 10 rrn operons and using length heterogeneity PCR (LH-PCR) (94), 4 fragments were seen for this one organism; using automated ribosomal intergenic spacer analysis (ARISA) (41), 4 fragments were seen; using DGGE, 2 fragments were seen; and using terminal restriction fragment length polymorphism (T-RFLP) with Rsal, 5 ribotypes were seen; and using Hhal, 3 ribotypes were seen (25). Indeed the variability is great and can lead to over-estimation of diversity and/or erroneous representation of an organism in an environment; hence, complicating a community characterization. Nevertheless, either method, if normalized well, can produce useful comparisons for measuring bacterial community changes or comparing experimental treatments. The versatility of molecular fingerprint analyses to be used in a variety of systems has enabled environmental research to go beyond qualitative assessments of ecosystems to more sophisticated analyses using multivariate ordination including other biotic 29 or abiotic data, correlation analyses and determining key factors or species that distinguish one environmental sample from another. Band Sequencing Sequencing rDNA from bands in RISA and DGGE gels provides the opportunity to determine phylogenetic affiliations of relatively abundant populations, represented by the bands that have been impacted by a particular treatment. Ranjard et al., 2000 excised and made clone libraries from two emerging bands found in bacterial populations from soil spiked with inorganic mercury (93). Neither band represented a single organism, and restriction fragment length polymorphism analysis was used to isolate predominant phylotypes from each library that were subsequently sequenced. One band contained 30 phylotypes, of which one dominant phylotype comprised 41 clones and three other major groups comprised 4, 2 and 2 clones while the remaining phylotypes were singletons. The second band was represented by 54 phylotypes with two predominant groups containing ca. 15 and 13 clones and, again, the remainder of the phylotypes were singletons. This study confirmed that RISA was a beneficial tool for monitoring changes in bacterial communities and it afforded the opportunity to spot populations of bacteria that responded to a mercury spike in soil. A shotgun cloning technique was employed in assessing the bacterial composition of pulp mill wastewater treatment systems (100, 111). In these studies, RISA amplification included ca. 500 base pairs (bp) of the 3' end of the 16S rRNA gene. This provided enough sequence information to conduct a phylogenetic analysis on clones from different spatial and temporal wastewater samples. A comparison was made between bacterial biofilms that grow on plant biomass in herbivore gastrointestinal tracts and those communities in the planktonic fraction of herbivore digesta using RISA and DGGE fingerprints (66). Both RISA and DGGE banding patterns clustered with respect to the animal used, but only RISA patterns clustered with respect to diet. DGGE had less 30 consistent clustering with respect to diet or fraction sampled. Subsequent phylogenetic analyses used only RISA amplicons which provided ca. 500 bp of sequence information while the DGGE bands would have provided ca. only 162 bp of sequence data. An advantage of the RISA method is that it discriminates at a lower taxonomic level than the D G G E method because amplicons include the ribosomal intergenic spacer region which is more variable than the 16S rRNA gene, as evidenced by the ability of the intergenic spacer to discriminate within and between species (49). Further, sequencing RISA bands containing four variable regions within the 16S rRNA gene compared to one or two variable region(s) in DGGE bands provides more phylogenetic information. This may explain the greater discrimination provided by RISA data over D G G E data, even though both techniques are plagued with problems arising because of multiple copies of the rRNA operons that exist in most bacteria (described above). This situation can lead to over-estimates of bacterial diversity. Nevertheless, the additional genetic information in RISA amplicons compared to D G G E amplicons and the importance of 16S rRNA gene variable regions in bacterial identification (47) suggests that RISA is more efficient at distinguishing microbial communities compared to DGGE. A non-parametric ordination analysis is a statistical analysis best suited for ecological data. Establishing band classes (operational taxonomic units - OTUs) from band mobility and intensity data from the molecular fingerprints can be used as the input source to generate a similarity matrix. One way of relating variables is to use correlation analysis. Kendall's x (tau) is a non-parametric correlation measure that is used when the standard correlational analysis assumptions are not met. Kendall's x is based on an order, or ranking, of the observations, and the distribution of Kendall's x does not rely on the distribution of the individual variables in the actual data (60). Assuming that the distribution function of bacterial communities is not a normal distribution, then rank scores expressing the relationship between the ordination score and the individual variables used to build the ordination are preferred, because the theory of 31 probability, in many cases, does not apply to non-normal distributions (21). Band classes designated in the ordination analyses can be ranked and chosen based on high positive or negative correlations with respect to each ordination axis. Sequencing bands from highly correlated band classes that distinguish sample fingerprints elucidates which divisions of bacteria are substantially associated with any one treatment. Finally, including abiotic environmental factors as an overlay on the ordination analyses uncovers the types of environmental niches that are preferred by the highly correlated bacterial divisions sequenced. The relatively new molecular fingerprinting methods for analyzing bacterial communities, in conjunction with non-parametric statistical analyses, provides opportunities to determine which populations of bacteria are characteristic and of potential functional importance in harvested or burned forest soils. While CmjC and Nmic measurements can only confirm results from previous studies, molecular fingerprinting provides valuable new information about the changes in bacterial communities following natural and human disturbances. Not only are these changes important from a scientific perspective, but also information gained will beneficially serve forest management practices oriented towards a close-to-nature management approach R A T I O N A L E Significant changes to forest soil microbial communities that mediate decomposition and nutrient cycling impact critical processes such as forest productivity and regeneration. Many studies of forest soils have looked at microbial diversity (48), microbial functional capabilities (84) and microbe abundance (73), and some of these studies were conducted in the rhizosphere, non-rhizosphere, mineral soils and litter of forest soils (7, 8, 104, 115). Of particular relevance to this study, some groups have examined microbial activity and other effects of microbial composition following fire and harvest experimental treatments in forest ecosystems. It is known that microbial biomass carbon decreases in harvested forest sites and that even greater decreases are measured after fire treatments (10, 17, 86). Phospholipid fatty acid (PLFA) 32 analyses have quantified fungal and bacterial biomasses in harvested and burned sites and have also strived to characterize the community based on P L F A signatures (10, 88). Spread plate methods using selective media for heterotrophic bacteria, aerobic sporeforming bacteria, filamentous fungi and cellulose decomposers are still being using to study soil microflora after fires and harvest treatments (42, 54). The above fire studies have involved controlled burns, in laboratory experiments or in the field. However, controlled burns are not as hot or as disruptive as natural wildfires. Variability of the effects of fire is high due to seasonal influences, climatic factors, such as rain and temperature (73), the soil sampling depth, fire intensity (31) and length of time post-fire for sampling. Wildfires are natural events in most forests and have return times ranging from 15-2000 years depending on the forest type, climate regime and topographic characteristics (1). A successful forest management strategy might closely mimic the natural processes from which the forest has evolved; hence, preserving the ecological regimes for the area and maintaining the natural in situ biodiversity. A long term goal of this project is to employ molecular methods to develop a means to use the forest soil microbial communities as biological indicators of forest soil health. In the short term, assessments of microbial community composition in forest soils will generate baseline knowledge of microbial communities and their responses to perturbations that can be used in forest management strategies. In this research, the effects of wildfire and clear-cut treatments on the forest soil microbial community were examined using cultivation independent methods. Microbial biomass measurements and bacterial community fingerprint changes were assessed. Microbial changes may serve as more rapid indicators of the ecological status of forests compared to soil, geomorphology or botanical assessments. 33 T H E R E S E A R C H P R O J E C T Chisholm, Alberta Wildfire Fire season in Alberta, Canada was at its longest and most severe intensity in 2000 and 2001 (19). In fact, extreme fire behaviour was already beingexperienced by municipalities in agricultural zones prior to the ignition of the Chisholm fire. On May 23, 2001, a man-caused fire was ignited, which turned out to be a record fire event. The Chisholm fire was a composite of five fires that merged into one. The main types of fire behaviour that were observed were intermittent to continuous surface and crown fires. The CFFDRS FWI rating of the Chisholm fire was 116 at its highest, and averaged at ca. 98. These values significantly exceeded the 43 year average of 59.6. In 1968, the Vega fire site, which burned parallel to the Chisholm fire site, had a regenerated forest of young aspen trees. Aspen can survive high intensity fires or slow the rate of fire spread (59, 79); consequently, the Chisholm fire progress, with the aid of the wind pattern, was slowed due to the serai aspen growth in the Vega fire site. The re-growth of aspen also likely prevented the Chisholm fire from reaching the subdivision of Poplar Estates on May 28, 2001 (19). Compared to the Vega fire and a 1980 wildfire in Cold Lake, Alberta, the peak burning conditions reached similar or higher values in the Chisholm fire, wind speeds were greater and the fire weather indices reached higher maxima. The Chisholm fire, in comparison, was 8 times more intense than the 1968 Vega fire and greater than 8 times more intense than the 1980 Cold Lake fire. The Chisholm fire surpassed the Vega fire in intensity, fuel consumption and human impact (18). As with the Vega and Cold Lake fires, the Chisholm fire site was experiencing 20-30% below normal precipitation levels at the time of onset. General weather patterns diverted moisture potential north and south of the affected Chisholm area. Indeed, the extremely dry conditions prevailed up to April, 2001 when the Chisholm forestry tower recorded a 93% below normal precipitation amount for that area. Extreme wind speeds were occurring and fuels were 34 critically dry; these conditions continued throughout most of the fire event. At the fire front, deep consumption of the forest floor and vigorous surface fires ensured high disturbance impacts at deeper levels of soil. Community-level impacts were high (20). A total of 75 structures were destroyed and forest resources, such as growing stocks and annual allowable cuts, were lost. Significant disturbances to wildlife, mortality due to fire and loss of habitat, were assessed, but long-term effects are still unknown. Fisheries and stream habitats were significantly affected, for example by the loss of large, woody debris and riparian zones. Studies done on fires in Yellowstone Park suggest that re-establishment of sensitive areas in Chisholm may take up to 10 years (96). Other impacts of the Chisholm fire include destruction of areas of significance to First Nation's people, protected areas, rail lines, electrical power systems and public health facilities. The Project After the 2001 Chisholm fire, a long-term productivity study was set-up in the Chisholm-Slave Lake area to analyze the effects of wildfire, harvest treatment and post-fire salvage-logged stands on boreal forest ecosystems. It is important to understand the natural evolution of microbial communities post-fire and, possibly, attempt to emulate that succession in a harvest management scheme. Ecological effects of altering natural wildfire regimes result in a loss of biodiversity of vegetation that has adapted to a specific fire regime. M y particular project analyzes the changes in the bacterial communities in a spruce-dominated boreal forest soil one year post wildfire and eight months post harvest and salvage-logged treatments. Soil samples were collected from individual spruce-dominated stands in a completely randomized design so that any effects on the bacterial communities may be attributed to the treatment rather than by chance alone. Microbial biomass was measured using C m j C , N m j c and DNA^xt. Bacterial community changes in the burned and unburned mineral layers of the various treatments were analyzed and compared using RISA and DGGE. These molecular fingerprints enabled: 1) 35 nonparametric multivariate ordination analyses to be conducted on RISA and DGGE fingerprint data to determine the relationships of the bacterial community sample fingerprints with some measured abiotic environmental factors; NO3", NH44", P e x t and pH; and, 2) a comparison of the two fingerprinting methods. Finally, I sequenced bands that predominated in RISA fingerprints or strongly distinguished a sample's bacterial community and conducted a phylogenetic analysis of those bacterial sequences. The Research Questions The following questions were addressed: 1. How do wildfire and harvest disturbances in a boreal forest affect the native soil bacterial communities? 2. Are bacterial communities significantly different among the treatments? 3. Are bacterial communities more or less diverse in disturbed treatments? 4. Are there key organisms/band classes that distinguish any one community from the others; i f so, what are these key organisms? 5. Is any one abiotic environmental factor significant with respect to the observed changes in bacterial community compositions? 36 C H A P T E R 2: E F F E C T S O F F O R E S T W I L D F I R E A N D H A R V E S T I N G I N T R O D U C T I O N Significant changes to the microbial communities that mediate decomposition and nutrient cycling are thought to impact processes such as forest regeneration. Many studies of forest soils have looked at microbial diversity (35), microbial functional capabilities (68) and microbe abundance (59). Some of these studies were conducted in the rhizosphere, mineral soils and litter of forest soils (8, 9, 83, 95). Of particular relevance to this study, some groups have examined microbial activity and other effects of microbial community composition following fire and harvest experimental treatments in forest ecosystems. It is known that microbial biomass carbon usually decreases in harvested forest sites and decreases to an even greater degree after fire treatments (11, 20, 70). Phospholipid fatty acid (PLFA) analyses have quantified fungal and bacterial biomass at harvested and burned sites and have also attempted to characterize the community based on P L F A signatures (11, 72). Spread plate methods using selective media for heterotrophic bacteria, aerobic sporeforming bacteria, filamentous fungi and cellulose decomposers are still used to study soil microflora after fires and harvest treatments (32, 41). The above fire studies have involved controlled/prescribed burns, in laboratory experiments or in the field. However, prescribed burns are not as hot or as disruptive as natural wildfires. Wildfires are natural events in most forests and have return times ranging from 15-2000 years, depending on the forest type, climate regime and topographic characteristics (1). One strategy for forest management is to closely mimic the natural processes from which the forest has evolved. Fires impact the physicochemical environment of forest soils in several ways. Vegetation and litter removal and reduced albedo result in increased surface temperatures (48, 56). Changes in soil structure, hydrophobicity and alterations in nutrient cycling are other examples of changes that occur after a fire (7, 26, 27). Severe heating of soil breaks down the structures of the 37 inorganic parent materials (90) and the result is a less stable soil structure and degradation of the microorganisms' habitat. A soil's ability to hold water is negatively affected by fires. Fires can create hydrophobic layers within the soil structure (4, 26, 27), resulting in decreased water infiltration and increased soil erosion from water runoff. Soil hydrophobicity also has a negative effect on plants, which provide niches for microbial colonization, by reducing the depth of root growth and hampering the process of re-vegetation. Nutrient transformations occur when excessive heat is applied to soils. In either a prescribed burn or wildfire, changes to nutrients like nitrogen, which tends to be a limiting factor for many organisms, are affected (23). Ammonium (NH44") and nitrate (NO3") concentrations are most commonly measured after fires, but consistent trends in either nutrient are not found in the literature, because the effects are governed by multiple factors including pre-fire concentrations of the respective nutrients, soil type, moisture and litter levels, vegetation burned and duration and intensity of the fire (23, 29, 46, 50, 55). Post fire, increased levels of soil nutrients, phosphorus, potassium, magnesium and calcium, were reported in a trembling aspen woodland (79). Finally, most studies report an increase in soil pH at fire sites (4, 55, 71). Wildfires produce drastic and long-lasting effects on soil bacterial communities. Decreases in microbial biomass were seen in fire ravaged soils (28, 59, 73, 93); however, some literature reports initial surges of microbial growth immediately after fires preceding subsequent drastic decreases (93). Generally following fires, an overlay of ash resulting in an increase in soil pH also causes increases in the availability of phosphorus, calcium, magnesium and potassium. These changes appear to favour bacterial growth over fungal growth (14, 59, 93). Increased populations of gram-positive organisms such as Bacillus spp. are common in fire-treated soils (49, 59, 93). Finally, similar microbial biomass levels were reported in unburned and burned treatments 6 years post fire (14), while other studies report sustained decreased biomass measurements in fire-treated soils compared to control soils 11 years or beyond (28, 73). 38 Harvesting also effects changes to microbial communities in soils, although to a lesser extent than fire. Wider fluctuations in soil temperature and moisture content due to unimpeded exposure to the elements occur in the forest floor following harvest treatments with generally higher soil temperatures and lower moisture levels (33, 43). Increased availability of magnesium has been reported in clear-cut treatments (33) as well as short term increases in N H / (12). After an initial increase, reductions in soil microbial biomass were seen after clear-cut treatments (31, 58, 67). Some literature reports, in contrast, immediate reductions in soil microbial biomass (70, 78). Greater reductions were seen in fungal biomass compared to bacterial biomass (70). The observed fungal reductions are hypothesized to be due to cessation of annual litter fall, root growth and root exudation (10, 31). Earlier studies on the bacterial component in a clear-cut spruce forest reported considerable changes in bacterial biomass as assessed by random isolation of heterotrophic bacteria passed through several biochemical tests (67). Initially, significant increases in groups of organisms occurred which lasted at least 7 years. A return to the control forest state was evident after 13 years. A follow-up study determined that the observed changes were due to an increase in decomposable matter, consisting of dead roots and felling residue remaining in or on the soil, because when this supply was exhausted biomass measures returned to control levels (84). A denaturing gradient gel electrophoresis analysis of limed and non-limed spruce forest soils ammonia-oxidizing community revealed that clear-cutting changed the composition of the community and that these changes were correlated with potential nitrification rates (12). Microbial alterations have faster response times to natural or forest management disturbances than to changes observed in the soil organic matter (6, 42). Indeed, bacteria and fungi are the bottleneck that all degradable carbon, nitrogen and other nutrients flow through in soil. Sparling (1992) observed a consistent trend for the microbial biomass C (Cmj C) and total organic carbon (Corg) ratios in soil to be higher in long-term, undisturbed soils than under afforestation or 39 cropping and suggested that this introduced parameter would be extremely useful as a monitoring index (81). Hence, C m i C : C o r g values would quickly provide an indication of the state and health of forest soils. Also, present day molecular tools allow for detection of the natural range of microbial community variability providing the ability to generate a frame of reference for evaluating the extent of change in disturbed ecosystems (85). Despite the above advances, knowledge of forest soil microbial communities impacted by harvest and, particularly, by wildfire disturbances is limited. This limitation hinders our ability to manage forestlands. Because the number of factors that govern forest soil microbial communities are many and overlapping (16), further microbial studies are needed to analyze complex community reactions to human and natural disturbances. It is important to understand the natural evolution of microbial communities post-fire and, possibly, to attempt to emulate that succession in a harvest management scheme. This is the first extensive molecular study, assessing how a wildfire and harvest treatment impacted Boreal soil bacterial community compositions. We evaluated i f any abiotic environmental factors were significant with respect to the changes in community compositions and whether these bacterial communities were significantly different among the treatments analyzed. A n analysis of the microbial biomass quantified the extent the treatments changed the microbial communities in soils impacted by wildfire and harvesting, and finally, sequence data from distinguishing band classes, possibly key species, in the sample fingerprint patterns provided a survey of the important bacterial divisions that comprise the particular treatments. In the present study, the bacterial communities in a spruce-dominated Boreal forest soil were examined one year post wildfire and eight months post harvest. This examination had several facets: 1) at a molecular level, two different D N A fingerprinting methods were used and compared; 2) at a biomass level, extracted D N A ( D N A e x t ) , microbial biomass carbon ( C m j c ) and microbial biomass nitrogen ( N m j C ) were measured and compared; and 3) at a phylogenetic level, 40 populations that distinguished community fingerprints were identified by sequencing bands from the D N A fingerprints. The two molecular fingerprinting methods used were ribosomal intergenic spacer analysis (RISA, (15)) and denaturing gradient gel electrophoresis (DGGE, (66)). The former produces banding patterns based on length heterogeneity of polymerase chain reaction (PCR) amplicons, while the latter produces its distinct banding patterns based on the PCR amplicon's mobility in denaturant gradient gels, characterized by the G/C content of the amplified region. A n advantage of the RISA method is that it discriminates at a lower taxonomic level than the D G G E method. RISA amplicons include the ribosomal intergenic spacer region which is more variable than the 16S rRNA gene, as evidenced by the ability of the intergenic spacer to discriminate within and between species (36). Lastly, sequencing bands from RISA fingerprints that predominate or discriminate sample fingerprints from specific treatments provides insight into the types of bacteria that are present and clues to the functional capabilities that are possible or required within an individual treatment (12, 75). M A T E R I A L S A N D M E T H O D S Sample Sites. Soil samples were taken from the Chisholm-Slave Lake area, approximately 150 kilometers north of Edmonton, Alberta (Table 1). A 116,000 hectare wildfire occurred in this area between May 23,2001 and June 4, 2001. Twelve sites, each approximately 2.5-5 hectares in size, half burned and half unburned were selected in a completely randomized design. Four treatments in triplicate were control (G), harvested (H), burned (B) and burned-salvaged (S). The harvest and burn-salvage treatments were clear-cut over the 2001/2002 winter. An example of a sample plot name is GSpl P10, which indicates a soil sample taken from the control (G) treatment, in a spruce (Sp) predominant forest, replicate site 1 at plot (P) 10. A l l sites were comprised of productive mixed wood, trembling aspen (Populus tremuloides Michx.), balsam poplar (Populus balsamifera L.) and, predominantly, white spruce (Picea glauca (Moench) Voss). The sites were classified as productive based on the size, density and 41 composition of trees located in the stand as classified by the Alberta Vegetation Inventory (AVI). Selected sites comprised >80% of white spruce and the mixed wood composition consisted of at least a 60%/40% split between coniferous and deciduous species. Optimum site characteristics consisted of a stand density of 51-70% crown closure with a tree height of 25 meters. Mostly the sites were level to nearly level (0-0.5% slope) while one harvest site (HSpl) had some plots, P11,P13,P14 and PI6, on a west-facing aspect. The slope was roughly classified as very gentle to gentle (2-9% slope). The soils in the area were generally brunisols and dark gray luvisols and the clay content averaged at 19.6% and sand was 37.7%. Each site had a 16-point grid with 20 meters between plots, from which 10 sample plots were randomly chosen, and was surrounded by a 50-meter buffer zone. In the control and harvest treatments, litter was removed from the sample plots and a soil core was taken from the top 0-10 centimeters of the mineral soil with a sterile scoop. Litter layers were completely burned in the two burned treatments. Soil samples were homogenized in sterile, plastic bags and placed in sterile, sample bottles (I-CHEM certified) for transport. A l l soil samples were collected in July, 2002 and the soil samples were kept cold prior to processing which was done within one week of collection. D N A extraction. Total microbial community D N A (DNAeXt) was extracted from 0.5 g fresh weight soil samples using the FastDNA® SPIN Kit for Soil (Q-BIOgene, CA). Soil samples were added to 978 pi of Sodium Phosphate Buffer, 122 pi of M T Buffer and lysing matrix, a mixture of ceramic and silica particles. Samples were secured in the FastPrep® Instrument and processed at 5.5 m/s for 40 seconds. Samples were centrifuged at 14,000 x g for 1 minute and the supernatant of each sample was transferred to a sterile Eppendorf tube to which 250 pi of PPS reagent (protein precipitating solution) was added. The tubes were mixed by hand shaking the tube 12 times and the precipitated proteins were pelleted by centrifuging at 14,000 x g for 5 minutes. The supernatant was transferred to a sterile 15-ml centrifuge tube with 1.0 ml of D N A 42 Binding Matrix suspension and placed on a rotating wheel for 2 minutes. After mixing, the silica matrix was allowed to settle for 3 minutes and then 670 pi of supernatant was removed, avoiding the settled Binding Matrix, to decrease the total volume of the mixture. The Binding Matrix was gently suspended and transferred, in two rounds, to a SPIN™ Filter, centrifuging 1 minute at 14,000 x g and emptying the catch tube each time. Samples were washed with 500 pi of SEWS-M (salt-ethanol wash) solution, which was added to the SPIN™ Filter and removed by centrifuging at 14,000 x g for 1 minute. The eluate was discarded, and samples were dried by re-centrifuging at 14,000 x g for 2 minutes. SPIN™ Filters were transferred to sterile Catch Tubes and left, caps open, in a fume hood for 5 minutes to ensure evaporation of the ethanol from the Binding Matrix. To the dry Binding Matrix, 100 pi of DNase/Pyrogen Free Water was added, and the slurry was gently stirred to allow the D N A e x t to elute from the Binding Matrix. The eluate was collected in a sterile Catch Tube while centrifuging at 14,000 x g for 1 minute. The integrity of the D N A e x t was checked on a 0.8% agarose gel, stained with ethidium bromide and uv-illuminated in an Alphalmager (Alpha Innotech, CA). DNAeXt was quantified using a known amount of 1-kb ladder, and solutions of 3 ng/pl of D N A e x t were prepared for polymerase chain reaction (PCR) fingerprinting methods. Chloroform-fumigation extraction of soluble organics. Chlorofoim-fumigation extraction was done using 15 g (fresh weight) of each soil sample, according to the standard technique (92). Samples were fumigated for five days with 50 ml of ethanol-free chloroform in moist paper towel-lined desiccators, without desiccant, in a fume hood (40). At the beginning of each day, the desiccators were evacuated 3 times, allowing the chloroform to boil for two minutes, and air was allowed back into the desiccator to ensure complete saturation of chloroform throughout the soil samples. Then the desiccators were evacuated one last time and left for 24 hours in the dark at room temperature. At the end of the fifth day, the desiccators, without chloroform, were 43 evacuated 3 times and allowed to stand open for 30 minutes in a fume hood to ensure complete removal of chloroform from the soil samples. The non-fumigated samples were extracted at the time fumigation commenced, using 70 ml of 0.5 M K 2 S O 4 on a reciprocal shaker on ice for 1 hour. Control samples without soil were also processed to determine background carbon and nitrogen values for both the fumigated and non-fumigated samples. Samples were gravity-filtered through presoaked Whatman® 42 ashless filter paper (Whatman International Ltd., England) and then vacuum-filtered through 0.45 pm Millipore filters (Millipore Corporation, MA) . A l l sample solutions were stored at -20°C in acid-washed vessels until further analysis. Microbial biomass carbon. Soluble organic carbon was analyzed in the fumigated and non-fumigated K 2 S O 4 extracts using the high-temperature combustion method with a Shimadzu TOC-500 Carbon Analyzer. Microbial biomass carbon (CmjC) was estimated as the difference between fumigated and non-fumigated samples (both less blank values) divided by a KEC of 0.45 (45, 94). A l l microbial biomass data was expressed as per oven-dry soil basis (105°C, 24 hours). Microbial biomass nitrogen. For the determination of total nitrogen, an alkaline persulfate oxidation, which oxidizes all the nitrogen species to NO3", was done on the K2SO4 extracts (19). Pre-oxidation levels of N C V - N were determined in the non-fumigated samples on the Lachat QuikChem Ae Autoanalyzer at the Soil Science Laboratory at U B C . Available N H 4 + was determined colorimetrically in the non-fumigated samples, using a Technicon Autoanalyser II. Total N03"-N in the oxidized K 2 S 0 4 extracts was measured in the fumigated and non-fumigated samples. Microbial biomass nitrogen (NmiC) was estimated as the difference between fumigated and non-fumigated samples (both less blank values) divided by a KEN of 0.54 (18). R I S A amplification. RISA fragments were amplified using primers S926f (5'-C T Y A A A K G A A T T G A C G G - 3 ' ) and L189r (5 ' -TACTGAGATGYTTMARTTC-3 ' ) which anneal to positions 910 to 926 of the 16S rRNA gene and 189 to 207 of the 23 S rRNA gene 44 (Escherichia coli numbering). The amplicons contain the length-variable ribosomal intergenic spacer (RIS) region and are referred to as rDNA-RIS amplicons. These rDNA-RIS amplicons were electrophoretically separated on an acrylamide gel, which produced a banding pattern specific to the samples' bacterial community composition. A 50-pl PCR contained 39.15 pi dH 2 0 , lx PCR buffer (200 m M Tris-HCl, pH 8.4, 500 m M KC1), 2.0 m M M g C l 2 , 672 pg of bovine serum albumin per ml, 200 p M concentrations of each deoxynucleoside triphosphate, 30 pmol of each primer, 1.25 U of Taq D N A polymerase (Invitrogen™, CA) and 30 ng of template D N A . To minimize contamination, all pipettes, cotton-plugged pipette tips, eppendorf tubes and PCR tubes were uv-irradiated for 15 minutes prior to PCR set-up. Cycling parameters on the Robocycler Gradient 96 (Stratagene®, CA) were 95°C for 5 minutes, 30 cycles of 94°C denaturation for 30 seconds, 47°C annealing for 30 seconds and 72°C extension for 2 minutes and ending with a final extension at 72°C for 5 minutes. rDNA-RIS amplicons were purified with a QIAquick PCR purification kit (Qiagen, CA) and eluted with 30 pi of EB (10 m M Tris-Cl, pH 8.5) prior to electrophoresis. To simplify the handling of the acrylamide gels during staining and imaging, the long glass plates used for the electrophoresis were treated with Sigmacote® (SIGMA C H E M I C A L CO., MO) to prevent the gel from sticking after electrophoresis, and the short glass plates had a Bind Silane (Promega, WI) treatment to make the gel stick to the plate during staining and imaging. The rDNA-RIS amplicons were electrophoretically separated on a 3.5% Duracryl (30% Duracryl 30T, 2.6C; Genomic Solutions, MI) gel at 80V for 16 hours, stained with 3 pi of GelStar (Bio-Whittaker Molecular Applications, Rockland, ME) in 200 ml l x T A E (40 m M Tris Acetate, 2 m M EDTA) for 1.25 hours, then double-stained with 5 pi of S Y B R Green I (Molecular Probes, OR) in 200 ml l x T A E for 1.25 hours. Fingerprint patterns were imaged on a Typhoon Imager (Amersham Biosciences, CA) using a 526 SP/Green (532 nm) laser, 600 photo multiplier tube 45 for detection and quantification of the emitted light, medium sensitivity and a pixel size of 50 pm. Fingerprint banding patterns were analyzed in GelCompar II™ software (Applied Maths, Belgium) and normalized using a 100-bp ladder (Invitrogen™, CA). For the band selection step, minimum profiling was set at 10.0% relative to the maximum value to adjust for fingerprint patterns with uneven intensity. This value appeared to best include real bands while excluding false bands. The band report was exported from GelCompar II™ to a spreadsheet and band classes (operational taxonomic units - OTUs) were established from the band mobility and intensity data. Starting from the smallest value of band mobility, equal ranges (band classes) of band mobility were identified, maximizing the band classes containing a single band. Inevitably, some band classes were empty, while others contained the sum of two, or rarely three, bands. D G G E amplification. DGGE fragments included amplification of the variable 3 (V3) region of the 16S rRNA gene. Primers used for this amplification were 341f-GC (5 ' -CGCCCGCCGC G C G C G G C G G G C G G G G C G G G G G C A C G G G G G G C C T A C G G G A G G C A G C A G - 3 ' ) and 534r (5' - ATT ACCGCGGCTGCTGG-3 ' ) which correspond to positions 341 and 534 E. coli numbering (66). A 50 pl-PCR contained 41.85 pi dH 2 0 , l x PCR buffer (KC1, (NH 4)2S0 4), 672 pg of bovine serum albumin per ml, 200 p M concentrations of each deoxynucleoside triphosphate, 25 pmol of each primer, 1.25 U of Taq D N A polymerase (Qiagen, CA) and 3 ng of template DNA. Cycling parameters on the PTC-200 (MJ Research, M A ) were 95 °C for 1 minute, 30 cycles of 95°C denaturation for 45 seconds, 55°C annealing for 30 seconds and 72°C extension for 45 seconds with a final extension at 72°C for 7 minutes. PCR samples were purified with a QIAquick PCR purification kit (Qiagen, CA) and eluted with 30 pi of EB prior to electrophoresis. Amplified product was run on a 1.5% agarose gel to confirm product size, verify purity and quantify the PCR product. Purified samples were run in 8% polyacrylamide gels (37.5:1) containing a gradient of denaturants (100% denaturant consists of 40% [vol/vol] formamide and 7 M urea) of 40-60% in 0.5x TBE (45 m M Tris, 45 m M boric acid, 1 m M 46 EDTA). Each lane contained 450 ng of purified PCR product, and the gels were run at 64V for 16 hours at 60°C. Gels were stained with 5 pi of S Y B R Green I in 200 ml of 0.5x TBE for 2.5 hours. DGGE fingerprint patterns were imaged on a Typhoon Imager using a 526 SP/Green (532 nm) laser, 600 photo multiplier tube, medium sensitivity and a pixel size of 50 pm. Fingerprint banding patterns were analyzed with GelCompar II™ software and normalized with previously made standards containing selected bands that electrophoresed throughout the 40-60% gradient range. D G G E band selection was done as for the RISA fingerprint data, using a minimum profiling of 5.0% relative to the maximum values to adjust for fingerprint patterns with uneven intensity. Richness, Evenness and Diversity. Richness constitutes the number of species or OTUs in a sample, in this case, the number of band classes in a sample that contained at least one band. Pielou's J ( E = H % r ) is a measure of evenness where high values indicate relatively equal / max abundance of the different species that comprise a community (69). The Shannon-Wiener s diversity index (FT = log pt), based on information theory, is a measure of the «=i information content of a sample. This diversity measure is sensitive to both the number of species (S) and the relative abundances of those species (pt) (51). Statistical analyses. Analyses of variance (ANOVA) were conducted as for a completely randomized design on the M B - C , M B - N and the DNAeXt data which were log-transformed, as the original data violated the assumption of homogeneous variances. A N O V A was also conducted on the Shannon diversity indices. Results were considered significant atp < 0.05. Multiple comparisons were made with Bonferroni contrasts, where adjustments to the alpha level were made to maintain a constant probability of type I error. Nonmetric multidimensional scaling (NMS; (52, 53, 61)) was used for ordination of the sample fingerprint patterns and the abiotic environmental parameters, identifying the 47 relationships among all factors. This nonparametric ordination method best handles ecological community data that are non-normally distributed and maps the sample inter-relationships on a graphical image in a reduced number of dimensions (21, 62). Multivariate data reduction summarizes large datasets with many variables (original species space) into smaller composite variables that express the greatest amount of information from the original multidimensional data (ordination species space). The Bray-Curtis similarity distance measure, which is appropriate for community ecology, represents pairwise sample values as proportions of the maximum distance possible for all comparisons (17) and was used in the ordination analysis. A total of 40 runs were conducted on the real data and 50 runs with randomized data for a Monte Carlo test of significance. A random starting configuration produced from a random number generator was initially used, but for the final ordination, the starting configuration was supplied from a previous ordination run that produced a minimum of stress in the final ordination pattern. Assessment of dimensionality was done by inspecting the scree plot. This plot shows stress as a function of dimensionality, and a three dimensional plot was found to be the best solution for both the RISA and DGGE data. The main matrix was constructed as rows of samples and columns of band classes and was used as the input file for all multivariate statistical analyses conducted using PC-ORD software (62). Abiotic environmental factors were also compiled in a worksheet as rows of samples, that exactly matched the main matrix of fingerprint data, and columns of values for NO3", N H 4 + , P e x t and pH for each respective sample. These data were used as the second matrix input file for the PC-ORD software, and trends were plotted as an overlay on the fingerprint ordination. The relative strength and contribution of a variable's relationship to the fingerprint ordination was indicated by the length and angle of the vector. Multi-response permutation procedures (MRPP) were performed (63, 64) using PC-ORD software to test the null hypothesis of no difference in average within-group ranked distances (5). A Ranked S0renson non-metric distance measure was used, which helped to correct for the loss 48 of sensitivity of distance measures as community heterogeneity increased. Groups were defined according to treatment. T is the test statistic which describes the separation between the groups. The more negative the T value, the stronger the separation. A is the agreement statistic which describes the within-group similarity. When all items within a group are identical, A=l . When the heterogeneity within a group equals expectation by chance, then A=0, and when there is less agreement within groups, A<0. Last, the p-value evaluates the likelihood of getting a 8 as extreme or more extreme than the observed 8. Results were considered significant ifp < 0.05. Mantel tests were used to evaluate the hypothesis of no relationship between two symmetrical similarity matrices. The question asked is, "How often does a randomization of one matrix result in a correlation as strong or stronger than the observed correlation?". Distance matrices were constructed using Sorensen's distance measure and Mantel's randomization approximation (PC-ORD) with 1,000 randomized runs on the distance matrices. Sequencing and phylogenetic analysis. Band sequencing was conducted mostly from the RISA D N A fingerprints. Correlation coefficients indicating the relationship of each band class relevant to the ordination scores were the basis upon which bands were chosen for sequence analysis. Band classes with high positive or negative rankings with respect to the ordination scores and which were represented by intense, well-resolved bands were selected. Attempts were made to pick bands from each treatment in which they were found. A total of 37 bands, including bands of the same band class from different treatments, were chosen for sequencing. The selected bands were extracted using sterile pasteur pipettes, and stored in 10 pi LoTE (3 m M Tris, 0.2 m M EDTA) at 4°C until ready for sequencing. After band excision, gels were re-imaged to ensure desired bands were extracted. Extracted bands were checked for correct size by re-amplification and electrophoresis on a 2% agarose gel. Gels were stained with ethidium bromide and uv-illuminated on an Alphalmager (Alphalnotech, CA). The verified bands were quantified using the 1-D-MULTI program of AlphaEase and the 1 kb ladder (Invitrogen™, CA) to generate a standard curve. Approximately 60-150 ng of purified rDNA-RIS amplicons were used in the sequencing reactions primed with the S926f primer. Each 5-pi reaction contained 2 pi of BigDye (Applied Biosystems BigDye™ v3.1 Terminator Chemistry, CA) and 0.25 pmol S926f primer. Sequencing reactions were subjected to an initial denaturation at 95°C for 5 min. Subsequent steps included 34 cycles of a 30-second denaturation step at 96°C, a 15-second annealing step at 50°C and a 4-minute extension step at 60°C. Completed sequencing reactions were diluted with 15 pi d H 2 0 and cleaned using Sephadex G-50 Fine (Amersham Biosciences, England) in Centrisep columns (Princeton Separations, NJ). Sequencing reactions were conducted on an Applied Biosystems PRISM 377 automated D N A sequencer at the Nucleic Acids Protein Services Unit (NAPS) at the University of British Columbia. Approximately 500 bp of rDNA-RIS sequences were evaluated using the Chimera Check program implemented in the Ribosomal Database Project (RDP) (22) and none were found. The sequences were then compared to those in the National Center for Biotechnology Information (NCBI) database using the Basic Local Alignment Search Tool (BLAST) version 2.0 search program and the most similar rRNA gene sequence was used as the reference affiliation for subsequent phylogenetic analysis. ClustalX v. 1.83 (89) was used to align all rDNA-RIS sequences and reference affiliations. T R E E C O N for Windows v. 1.3b (91) was used to construct the phylogenetic tree using the neighbor-joining distance method (77) with Jukes and Cantor (47) correction. Bootstrap analysis consisted of 1,000 replications. Bands that did not initially yield good sequence data were cloned using the TOPO T A Cloning® system (Invitrogen™, CA). At least three white clones were chosen from each plate and colony PCR was done using the M13f(-20) ( 5 ' - G T A A A A C G A C G G C C A G - 3 ' ) and M13r (5 ' - C A G G A A A C A G C T A T G A C - 3 ' ) primer pair provided with the cloning kit. A 10-pl colony PCR contained 8.57 pi dH 2 0 , lx PCR buffer (KC1, (NH 4 ) 2 S0 4 ) , 672 pg of bovine serum albumin per ml, 200 p M concentrations of each deoxynucleoside triphosphate, 1 pmol of each 50 primer and 1.25 U of Taq D N A polymerase (Qiagen, Canada). Cycling parameters on the PTC-200 (MJ Research, M A ) were 95°C for 5 minutes, 30 cycles of 95°C denaturation for 1 minute, 55°C annealing for 1 minute and 72°C extension for 1 minute and ending with a final extension at 72°C for 7 minutes. The sizes of the colony PCR amplicons were confirmed by electrophoresis and correct sized fragments were subjected to the 5-pi sequencing reaction as outlined above. Appendices. Data not shown in the results for the above experiments can be found at http://www.microbiology.ubc.ca/Mohn/Chisholm.pdf. Nucleotide sequence accession numbers. The partial 16S rDNA sequences determined in this study were deposited in the GenBank under the following accession numbers: GSp2P4-l, AY730464; GSp3P7-2, AY730465; HSplP5-3, AY730466; HSp2P2-4, AY730467; HSplP14-5, AY730468; BSp2Pl-6, AY730469; GSp2P13-7, AY730470; GSp3Pl-8, AY730471; GSp3Pl-9, AY730472; GSp3Pl-10, AY730473; H S p l P 9 - l l , AY730474; SSp3P7-12, AY730475; SSplP12-13, AY730476; BSplP6-14, AY730477; BSplP6-15, AY730478; BSp2Pl-16, AY730479; GSplP l l -17 , AY730480; BSp3P10-18, AY730481; BSplP12-19, AY730482; GSp3P9-20, AY730483; HSp3P9-21, AY730484; BSplpl2-22, AY730485; HSp2P13-23, AY730486; BSp3P8-24, AY730487; BSp2P7-25, AY730488; SSp3P2-26, AY730489; SSp3P7-27, AY730490; GSp2P13-28, AY730491; GSp2P13-29, AY730492; GSp3P12-30, AY730493; GSp3P12-31, AY730494; HSp3P10-32, AY730495; HSp3P10-33, AY730496; HSp3P10-34, AY730497; HSplP9-35, AY730498; HSplP9-36, AY730499; BSp3P13-37, AY730500; BSp2Pll-38, AY730501; BSp2P12-39, AY730502; SSp2P7-40, AY730503; BSp2P12-41, AY730504; SSp2P2-42, AY730505; SSp3P7-43, AY730506; SSp2P2-44, AY730507; SSp2P2-45, AY730508; SSp2P7-46, AY730509; HSp3P10-47, AY730510; GSp2P13-48, AY730511; GSp3P12-49, AY730512; HSp3P10-50, AY730513; HSplP9-51, AY730514; BSp3P13-52, AY730515; BSp2Pll-53, AY730516; BSp2P12-54, AY730517; SSp2P7-55, AY730518; 51 HSp3P10-56, AY730519; HSp3P10-57, AY730520; BSp3P13-58, AY730521; BSp2Pll-59, AY730522; HSplP9-60, AY730523; HSplP9-61, AY730524; SSp3P7-62, AY730525; GSp2P4-1T-DGGE, AY730526; GSp2P4-2B-DGGE, AY730527; HSp2P2-3T-DGGE, AY730528; HSp2P2-4B-DGGE, AY730529; BSplPl -5-DGGE, AY730530; SSp2P2-6-DGGE, AY730531. R E S U L T S Microbial biomass carbon (CmjC) was higher in forest soil from unburned experimental treatments than in soil from burned treatments, while microbial biomass nitrogen (NmjC) was lower in the unburned treatments compared to the burned treatments (Fig. 1). A N O V A and Bonferroni's test determined that all treatments were significantly different in soil content of C m i C and N m i c (p < 0.05). Like Cmic, total microbial extracted D N A (DNA e x t ) from the unburned treatments was higher than from the burned treatments (Fig. 1) and all treatment means were significantly different (p < 0.05) from each other based on Bonferroni's test. C m i C : N m j C was highest in the unburned treatments, and was substantially lower in the burned treatments (Table 4). In the unburned treatments, C m j C :C o r g ranged from 2.3-3.9, while in the burned treatments C m j C :C 0 rg ranged from 0.8-2.0 (Table 4). Burn treatments had a major effect on bacterial community composition, based on the soil D N A fingerprint data, while harvesting treatments appeared to affect community composition to a lesser extent (Table 2). Nonmetric multidimensional scaling (NMS) ordinations of the RISA and D G G E fingerprint patterns provided graphical representations of the patterns of the bacterial community composition in which it was clear that there were differences between fingerprints of the unburned communities (control and harvest) and the burned communities (burn and burn-salvage) (Fig. 2, Fig. 3). Similarities of RISA patterns were higher (maximum 93%) than those of DGGE patterns (maximum 73%). Consequently, the smaller variability in the RISA fingerprint patterns suggests that differences seen from this method can be interpreted with more 52 confidence than those seen from the D G G E method. Community differences among the treatments were tested with an MRPP test, and overall, the RISA ordination species space explained 76.1% of the total cumulative variance represented from the distances in the original, unreduced species space; while, the DGGE ordination explained 63.9% of the total cumulative variance. Increases in NO3", N H 4 + , P e x t and pH measurements were seen in the burned treatments. A joint plot of these abiotic environmental factors was overlaid on the fingerprint community data (Fig. 2, Fig. 3). The direction and length of the vectors radiating from the centroid of the fingerprint ordinations represents the degree of correlation between the abiotic factors and the fingerprint patterns. Thus, the burned sample fingerprints cluster and are positively correlated with increases in NO3", NFL;"1", Pext and pH. The unburned sample fingerprints also cluster but are negatively correlated with increases in NO3", N H 4 + , P e x t and pH. The relative length of the vectors indicates that pH is 4 to 5 times more strongly correlated with the fingerprints than are the other abiotic factors. The overall MRPP with RISA bacterial fingerprint patterns illustrated strong differences (p < 0.001) among the treatments (Table 2). Differences between fingerprints of the two unburned treatments and between fingerprints of the two burned treatments were less obvious (Fig. 2) but were significant on the basis of multiple pairwise comparisons. The T statistic clearly indicates that each treatment's fingerprint patterns reside in distinct areas of the ordination species space, with more negative values indicating stronger separation. Positive values of the A statistic indicate that the null hypothesis can be rejected and that there are real differences between the compared groups (p < 0.01). D G G E fingerprint similarity between pairs of treatments showed similar trends as RISA fingerprint similarity, with obvious differences between the burned and unburned treatments (p < 0.001; Table 2). Again, significant differences were found between the control and harvest 53 fingerprint patterns (p < 0.05), but not between the burn and burn-salvage fingerprint patterns (p > 0.05). The RISA and D G G E analyses agree well with one another, as there is a significant relationship between the RISA and D G G E fingerprint distance matrices. A Mantel test or a test of correlation between both RISA and DGGE fingerprints distance matrices indicates a slight positive correlation that is statistically significant (r = 0.1303,/? < 0.001). The strength of the relationship between the RISA or DGGE fingerprint distance matrix and the abiotic environmental factors distance matrix is positive in both cases but is 2.6 times greater for the RISA fingerprint data (r = 0.1613,/? < 0.001) compared to the D G G E fingerprint data (r = 0.0620, p < 0.027). Both comparisons are statistically significant; however, the RISA/abiotic environmental factor comparison is highly significant while the DGGE/abiotic environmental factor comparison is only slightly so. Both RISA and D G G E fingerprints from the burned treatments had higher diversity than those fingerprints from the unburned treatments. This difference was evident on the basis of three different indices (Table 3), with only the exception of one RISA evenness index. RISA and D G G E Shannon diversity indices were significantly different between all treatments, with only the exception of RISA indices for the control and harvest treatments (both unburned). Populations most responsible for distinguishing between RISA fingerprints from different treatments were identified by rRNA gene sequence analysis (Fig. 4). Using correlation coefficients which indicate the relationship of each band class to the ordination scores, band classes with highly positive or negative correlations to the ordination scores were selected. In all cases where these band classes were represented by intense RISA bands, representative bands from samples of each treatment were excised and sequenced. Out of the 37 highly correlated bands, 26 produced high-quality sequence data from amplicons, and the remaining 11 were cloned for sequencing (Table 5). A total of 23 direct sequences were affiliated with prokaryotes; 54 while, 3 were affiliated with eukaryotes from the Basidiomycota. From each clone library, at least three random clones were chosen for sequencing, and from only one library, SSp3 P7, were all clones most similar to the same reference affiliation. Thus, the inability to directly sequence some bands was probably due to representation of multiple populations in those bands. In most cases, at least 2 of 3 clones from a library represented a common division. The most common reference affiliation that occurred was Hydrocarboniphaga effusa, a y-Proteobacterium (Table 5). Sequences affiliated with this organism were from band classes >.19-.21, >.37-.39, >.39-.41 and >.65-.67, which were found in all treatments, but were most intense in fingerprints from the unburned treatments. Another organism to which several sequences were affiliated was BacUlus aminovorans. These sequences were found in band classes >.53-.55 and >.59-.61, which were only found in fingerprints from the burned treatments. RISA bands that were selected for sequence analysis ranged from 8.3% to 41.8% of the total intensity in their respective fingerprints (Table 5, Fig. 4). Only in fingerprints from GSp2 P4, GSp3 PI and GSp3 P7 did the bands sequenced contribute greater than 25% of the total intensity, suggesting that these are predominant phylotypes in the unburned control treatment. Of the remaining bands sequenced, about half comprised 12.5-25%) total intensity of their fingerprints, and about half comprised <12.5%. Some D G G E bands that migrated identical distances were visually chosen to determine i f both bands corresponded to the same organism. Both GSp2 P4-1T and HSp2 P2-3T were most similar to sequences of y-Proteobacteria, while both GSp2 P4-2B and HSp2 P2-4B were most similar to sequences of a-Proteobacteria. However, two bands from the burn treatments that migrated to the same position, BSpl PI-5 and SSp2 P2-6, were most similar to a y- and a-Proteobacterium, respectively, which may reflect the increased diversity seen in fingerprints from the burned treatments. 55 The sequences determined spanned 8 different divisions (Fig. 5), with certain divisions found only in either the unburned or burned treatments. For instance, the sequences affiliated with the B-Proteobacteria, Bacillus, Parachlamydia and Nitrospirae divisions were exclusively derived from the burned treatments. Similarly, the sequences affiliated with the a-Proteobacteria were exclusively from the unburned treatments. Sequences from both the unburned and the burned treatments were affiliated with the remaining divisions. DISCUSSION Estimating microbial biomass with D N A e x t is problematic. Materials such as clay, sand and humic acid have a high affinity for binding free D N A and rendering the D N A free from nucleolytic enzymes thereby allowing its persistence in the environment (5, 24, 34, 57, 76). Commercial D N A extraction kits can extract and purify free D N A , possibly including soil-bound D N A with D N A from intact microbial cells (65), causing overestimation of biomass. In contrast, some researchers have found that extraction of D N A from mineral soils is less fraught with problems than extraction from humus layers and is therefore a reliable estimate of microbial biomass (60). Because of the uncertainty of estimating microbial biomass using D N A e x t , a microbial biomass estimate generally considered more reliable is extraction and measurement of microbial biomass C and N solubilized by chloroform fumigation. In this study, D N A e x t did not correlate well with either C m j C or N m j C , but D N A e X t did reflect the very general trend of lower C m j C in burned than in unburned treatments (Fig. 1). Clearly, the Chisholm-Slave Lake area wildfire considerably reduced microbial biomass, agreeing with most past research. In this study, reductions in C m j C , compared to the control treatment, were 74% in the burn treatment and 52% in the burn-salvage treatment (Fig. 1). The smaller reduction seen in the burn-salvage treatment compared to the burn treatment may have occurred due to the addition of nutrients to the forest floor from tree residues left behind after the removal of the marketable wood, providing an energy source for surviving microbes to 56 proliferate (41). In general, C m j C measurements post wildfire have showed significant reductions (28, 39, 59), and these reductions were sustained over periods ranging from 11 to 13 years (28, 73). An exception to this general trend was reported (93) in a study conducted one month after a wildfire. However, in this case, different methodologies may explain the differences observed. Harvesting is reported to have variable effects on forest soil C m i C . Summer clear-cuts were reported to reduce humus C m j C , measured by chloroform fumigation-extraction and complemented with reductions in PLFA content (11, 70). Greater reductions were seen in fungal PLFAs (59%) than bacterial PLFAs (24%) compared to a control treatment (11). In contrast, Smolander et al. (2001) found a slight increase in humus C m j C of a Norway spruce stand, from which logging residue from a winter clear-cut was evenly distributed over the harvested site (80). In the Chisholm-Slave Lake harvest sites, logging residue also remained on site; however, significantly lower C m i C determinations (19%) compared to the control treatment were found in the mineral soils of the harvest treatment (Fig. 1). One year after the burn, significant increases in N m j c occurred in both burned treatments in this study (Fig. 1). Burning may be responsible for rendering nitrogen available for microbial growth. In agreement with our results, one year post wildfire in a pine forest, N m j C levels were significantly higher (28). In contrast, lower N m i C was reported in soils ravaged by wildfires (73, ' 87). Significant decreases in N m j c were observed in the harvest treatment in this study (Fig. 1). Significant increases in N m j C were observed in organic horizons while no changes were observed in mineral horizons in a balsam fir forest two growing seasons after pre-commercial thinning treatments (88). Mean N m i C increased in soluble organic and inorganic nitrogen buried bag incubations in high-elevation spruce-fir forest floor and mineral horizons in both a harvest and control treatment, but the net change in N m j C was greater in the clear-cut site compared to the control (38); these studies were conducted 2 and 5 years post harvest treatment respectively, and 57 N m i C one year post harvest was not measured. Variability of C m i C and N m j c trends from harvested and burned forest soils complicate comparative analyses across studies, because initial environmental conditions and fire behaviour are unique for each site. Fires have a sterilizing effect on soils, being more detrimental to fungi than to bacteria (11, 28, 59, 72, 93). Also, increases in pH caused by fires tend to favour bacteria over fungi (14). Large C m i c to N m i C ratios typically indicate that fungal populations are more abundant than bacterial populations (28). As in many other studies, a slightly lower ratio was seen in the harvest treatment while larger reductions were seen in the burn and burn-salvage treatments in this study, suggesting a fungal predominance (Table 4) (28, 31). C m j C to C o r g ratios provide not only an indication of carbon availability and quality (44, 81), but also an indication of microbial activity (39, 73). In this study, the substantial decrease in C m i C : C o r g was mostly due to the large decrease in microbial biomass in the burned treatments, although there was a small increase in C o r g in both burned treatments. Changes in bacterial community composition among the various treatments were evident from the RISA and DGGE community fingerprint patterns (Table 2, Figs. 2 and 3). Past studies have analyzed microbial community composition through the use of Biolog microplates (49), the spread-plate method using selective media (3, 8, 30, 59, 93), as well as cloning of 16S rRNA gene sequences from isolates and direct PCR amplicons (9). The clear differences observed in this study between the unburned and burned treatments agree with previous research which analyzed PLFAs of microbial communities in forest humus after various heat shock treatments (72). Compared to the heated treatments, a significantly different subset of PLFAs contributed to the principal components determined for the control treatment. Baath (1995) tested if altered P L F A patterns, also from humus layers, were due to pH or soil organic matter changes in clear-cut and burn treatments (11). The clear-cut and clear-cut/burn communities clustered more closely than the control treatment which was distinct. These studies were done along side a 58 treatment in which ash from a fire was added to humus soils, with the effect of increasing the pH, as in the burn treatment. Similar P L F A patterns were observed in this ash treatment; therefore, this supported previous studies that suggested increasing soil pH affects microbial community composition in the same way in all forest humus. A caveat to the results of the Baath study was that the fire-treated and lowest ash addition treatments had similar pH but did not cluster in the principal component analysis (PCA), so had different P L F A patterns, and the fire-treated samples in one forest separated to the same extent as the fire-treated samples from another forest but both had different pH. Expectedly, it appears that other factors are at work and that pH differences are not the only determinant of microbial community changes. Recently, the Hayman 2002 wildfire in Colorado, USA provided the opportunity to study the microbial communities in the burned soils across a gradient of wildfire intensities (Aida E. Jimenez-Esquilin, personal communication). These communities are being analyzed using ester-linked fatty acid methyl ester (EL-FAME) analysis and with DGGE. Preliminary E L - F A M E results demonstrate that the microbial fatty acids, representing different viable divisions of bacteria and fungi, were substantially different among treatments. Also, initial D G G E patterns are demonstrating differences among the various treatments. D N A fingerprint analyses indicated that bacterial communities differed much less between control versus harvest and between burn versus burn-salvage treatments than they did between unburned and burned treatments (Table 2, Figs. 2 and 3). MRPPs demonstrated that significant separations existed between the control and harvest treatments and the burn and burn-salvage treatments. A P C A done in the Dumontet (1996) study showed the burn-salvage treatment, representing an 11 year old burned and logged site, was more similar to the control treatment compared to the four other recently burned treatments (28). However, even after 11 years, the burn-salvage treatment was still clustering with two of the older burned sites. Clear-cut humus and mineral sites of increasing age were microbiologically investigated to elucidate reasons why 59 reforestation was unsuccessful (67). Greater changes were seen in the mineral soil populations, but regression to the control community profile commenced 7 years post harvest. In the present study, the differences seen between the control and harvest treatments and the burn and burn-salvage treatments, i f they follow the trends of previous studies, have the potential to persist for many years. At a forest management level, since wildfires appear to cause much bigger changes in bacterial communities compared to harvest treatments, then harvest treatments alone may never be able to successfully mimic wildfire disturbances at the microbial level. Analysis of D N A fingerprints suggests that microbial diversity increased as a result of the burned treatments (burned and burn-salvaged) in this study. Diversity values calculated from the RISA data are greater in the burned treatments, and although the evenness is relatively the same in all the disturbed treatments, the increase in diversity can only be attributed to an increase in microbial richness. A proliferation of organisms that are taking advantage of the newly released nutrients caused by the burn or the addition of nutrients from the logging residue would be responsible for these increases in richness, evenness and diversity. Percent contribution of total band intensity in the fingerprints had larger ranges in the control and harvest treatments. This is indicative of some species having larger contributions to the total fingerprint pattern while others were less intense. The burn and burn-salvage treatments had higher richness but a smaller range of percent contribution of total band intensity which suggests a more even distribution of its members from their respective groups. A loss of biodiversity is a concern for forest managers because of the increase of forest management practices. Biodiversity losses can lead to less stable ecosystems and treatments, such as harvesting and burning, disturb the forest microbial community stability (82). The increases in richness and/or evenness in the harvest and burn treatments are good indicators that the soil ecosystem has not lost its biological ability to adapt. Sequence analysis identified organisms most responsible for N M S axes and so most important in distinguishing among treatments. The rRNA gene sequences divided into 8 different divisions 60 with the most dominant clearly being the y-subgroup of the Proteobacteria. In this group, the distribution of sequences was weighted to the unburned treatments, although there were 4 of 18 sequences from the burn and burn-salvage treatments. Most characteristic of the burn treatments were members of the Firmicutes {Actinobacteria and Bacillus groups), which were highly correlated in the burn and burn-salvage treatments with only one sequence from the control treatment. Bacillus spp. are common organisms in burned soils (87, 93). It appears that members of the a-Proteobacteria were strongly selected for within the unburned treatments, while members of the /3-Proteobacteria were strongly selected for within the burned treatments. Unique to the burned treatments were sequences affiliated with the divisions Parachlamydia and Nitrospirae, as well as one uncultured organism, further suggesting increased diversity in those treatments. Bacterial diversity from a long-term soil productivity project in British Columbia, Canada forests determined organisms from soil in a harvested treatment with and without heavy soil compaction classified as a-, y-, /3-, and S-Proteobacteria, Actinobacteria, Acidobacteria, Verrucomicrobia, Bacillus/Clostridium group, Cytophaga-Flexibacter-Bacteroides, green non-sulfur bacteria, the Planctomycetales, candidate divisions, TM6 and OP 10 and unclassified divisions (9). In the Chisholm-Slave Lake study, no sequences or clones from the harvest or control treatment were members of the /3-Proteobacteria or Bacillus groups; however, our sequence determinations came from selected bands that highly discriminated the RISA sample fingerprints rather than the whole soil D N A extracts and, therefore, these divisions were represented in only the burn treatments. This suggests that although P-Proteobacteria or Bacillus groups may be present in harvest treatment soil bacterial DNA, these are neither predominant members found in these treatments, nor ones which discriminate these communities from those in the other treatments. Some different discriminating band class sequences were most similar to the same organism. Ribotypes clustering with Hydrocarboniphaga effusa, a y-Proteobacterium, were examples of 61 this finding. They spanned four band classes, and it is impossible to say whether these ribotypes represented different strains of the same genus or products of multiple copies of the rRNA operon found within fewer than four strains (25, 37). The percent contribution of the above ribotypes to total band intensity of samples ranged from 9.0-41.8%. Indeed, H. effusa's presence throughout all treatments, but more abundant in the unburned treatments, was consistent with the diversity measures where lower evenness values in the control site indicated more uneven distribution of species compared to burned treatments that had higher diversity and more even distributions. Percent identities with reference affiliations were quite low for most submitted sequences, suggesting that they belong to unidentified species. One exception are sequences from the burn treatments which were most similar to Bacillus aminovorans with identities ranging from 98-99%. Based on RISA and DGGE molecular fingerprint methods, it is clear that bacterial communities were affected by the burning and harvesting disturbances, whether by immediate killing from the heat of the fire or by altered selective pressures due to disturbance-induced changes in the soil environment. The loss of tree and shrub biomass which affords habitats to microorganisms involved in nutrient exchange and the concomitant changes in available substrates caused by the fire are some factors that likely led to changes in microbial populations. Fire intensities can be great, indeed surface temperatures have been reported to reach 1,000°C (2); however, the temperature gradient down to the mineral soil is steep, especially when the moisture content in the humus layer is high. In a study where flames reached temperatures of 800°C, the moist humus layer was reported to only reach a temperature of 60°C (86). This suggests that the greatest changes to the microbial community are through its lost interactions with plant roots and available substrates. Wheat and clover pot experiments were set up in burned and unburned soil from a pine forest that experienced a devastating wildfire burn (54). Shoot and root biomasses were higher in the plants grown in the burned soils compared to those 62 plants grown in the unburned soils. Also the uptake of nitrogen, phosphorus and several cations were greater in the plants grown in the burned soils. In conclusion, there was a short-term flush of mineral elements mobilized from the burned woody plants and their litter which were returned to the environment through the vigorous growth of herbaceous plants. With a large number of samples analyzed in this study, MRPPs showed significant spatial variability in community composition within the treatments. Past studies have concluded this point by showing, for example, that isolates grown on soil-water extracts from one location did not grow as well on soil-water extracts from a site only one meter away (13). More relevant, Rainey and Travisano (1998) grew populations of Pseudomonas fluorescens from a single ancestral cell type (based on colony morphology) in a heterogeneous (without shaking) environment and in a homogeneous (with shaking) environment (74). Given the ecological opportunity to diverge, the population in the heterogeneous environment diversified to different cell types, supporting the ideas of ecological specialization and maintenance of diversity by spatial heterogeneity. However, despite the variability, we found that communities within each treatment, including replicates that were, at most, 43 kilometers apart, shared similarity and were characterized by particular predominant ribotypes. Thus, key aspects of community composition are common over a large spatial scale in this environment. Further, the disturbances in this study had severe effects on community composition, leading to changes greater than normal variability and different predominant populations that were distinct. Knowledge of the potential ecological roles played by the specific phylotypes revealed in this study is unknown, but it is clear that significant changes occurred when these communities were impacted by a wildfire or harvest treatment. Forest management schemes that can closely simulate the effects of wildfires, especially where wildfires are a natural part of the forest's regime, are beneficial to maintaining a close-to-nature ecosystem and would avoid problems such as impaired ecosystem function and biodiversity losses. Further studies are required to 63 understand the importance of each group's contributions to maintaining processes such as decomposition and nutrient cycling and how any one group works towards recovery or restoration of its environmental niche. A C K N O W L E D G E M E N T S This work was supported in part by Discovery Grants from the Canadian Natural Sciences and Engineering Research Council to W. Mohn and B. Kishchuk and sponsorship of N . Smith from the Wikwemikong Board of Education. Assistance from Paul A . Lythgo in the sequencing component of this research is greatly appreciated. C O N C L U S I O N S & F U T U R E R E S E A R C H This study was consistent with previous reports in which wildfire and harvest treatments reduced C m j C in forest soils. Fires not only have a sterilizing effect on microbial communities but also reduce plant biomass that supports microbial life. Previously reported effects of fire on N m i C are variable. One reason for the inconsistencies in biomass results is that pre-fire or pre-harvest environmental conditions are variable, and unpredictable fire behaviour can further complicate experimental outcomes. Few studies have examined effects of wildfires, as opposed to prescribed burns. Thus, this study provides valuable insight into effects of wildfire. There is a lack of research with respect to microbial biomass post wildfire; however, more importantly, an understanding of the range of natural variation in forests (structure, composition and ecological functions) post disturbance is the goal that needs to be attained. Both human and natural disturbances significantly affected Boreal soil bacterial communities' compositions. This fact is found in the literature and while return to 'close-to-original' composition occurs, the time factor has been of short and long durations. A temporal experimental design would address how long it would take the Chisholm-Slave Lake area to return to original community compositions. In this research, differences in community 64 fingerprint patterns were significant with burned treatments obviously differing from unburned treatments. Bacterial communities in harvest- and fire-impacted forest soils appear to be more diverse than the control treatment. This is because increases in the number of bands in the D N A fingerprints from harvested and wildfire disturbed sites in this study suggest that bacterial richness, or the total number of species, increased. The uniformity of band intensity in the fingerprints appears to be more so in the burned treatments compared to the unburned treatments. This would suggest that in the unburned treatments, predominant bacterial communities reside, while in the burned treatments a more equal distribution of communities exist; however, calculated evenness indices do not strongly support this conclusion. Additional studies, similar to this research, in other forest sites would complement this study and provide better comparisons to see trends and further understanding of the tendencies of bacterial communities impacted by harvest or wildfire disturbances. Key divisions of bacteria were observed in the burned and harvest treatments. Previous research has also uncovered some of these same divisions. D N A fingerprint bands that are highly correlated to treatments suggest that the populations represented are predominant and functionally important to the corresponding community. Traditional microbiology methods, such as spread plating, targeting these specific bacterial divisions are needed to better understand their functional significance. In this study, pH appeared to be the physicochemical factor that most distinguished burned treatment bacterial communities from the unburned. Subtle differences of the abiotic environmental factors (N03~, N H 4 + and P e x t ) in conjunction with the pH increases likely contributed to differences in community composition; however, the extent of their impact may only be discovered through highly controlled experiments. Such experiments may include controlled burns of forest soils or in-lab burning experiments using soil slabs collected from forests. TABLE 1. Locations of sampling sites in the Chisholm-Slave Lake area. Site Name Latitude/Longitude (degrees-minutes) Area GSpl 54°54'N, 114°3'W Chisholm, Alberta GSp2 55°17'N, 115°6'W Canyon Creek GSp3 55°13'N, 114°42'W Vanderwell Haul Road HSpl 55°17'N, 115°5'W Canyon Creek HSp2 55°17'N, 115°6'W Canyon Creek HSp3 55°17'N, 115°6'W Canyon Creek BSpl 54°54'N, 114°6'W Chisholm, Alberta BSp2 54°56'N, 114°6'W Chisholm, Alberta BSp3 55°8'N, 114°12'W Saulteaux SSpl 54°55'N, 114°6'W Chisholm, Alberta SSp2 55°1'N, 114°22'W Five Corners Road SSp3 55°1'N, 114°23'W . Five Corners Road 67 TABLE 2. Comparison of the differences in RISA and DGGE community fingerprints among treatments with nonparametric multi-response permutation procedures (MRPP), based on a ranked Sarensen distance measure (Bray-Curtis method); T = description of the separation between the groups and A = description of the effect size or "chance-corrected within-group agreement". RISA DGGE T A T A Ranked Sorensen -26.087** 0.261 -14.377** 0.153 Multiple comparisons: Control & Harvest -8.210** 0.093 -2.313+ 0.027 Control & Burn -23.629** 0.266 -15.619** 0.192 Control & Salvage -16.214** 0.187 -9.168** 0.109 Harvest & Burn -20.845** 0.243 -11.825** 0.150 Harvest & Salvage -17.664** 0.211 -9.042** 0.109 Burn & Salvage -3.918* 0.040 -1.785++ 0.021 *_p<0.01, **p< 0.001 +p> 0.01, ++/>> 0.05 68 TABLE 3. Richness, evenness and Shannon diversity indices for the RISA and DGGE fingerprint data. Shannon diversity data followed by a different letter are significantly different using Bonferroni's test aXp < 0.05. RISA DGGE Richness Evenness Shannon Richness Evenness Shannon Control 16 0.843 2.31a 26 0.857 2.77a Harvest 16 0.893 2.49a 25 0.861 2.76b Burn 18 0.879 2.55b 29 0.889 2.99c Salvage 18 0.895 2.58c 28 0.885 2.94d TABLE 4. Average C m j C :N m j c and C m j c :C 0 rg values in the mineral soils of the control, harvest, burn and burn-salvage treatments (n = 10). C . *N • c •r M n i c ' ^ m i c ^ - m i c ^ o r g Control 8.0 3.9 Harvest 7.0 2.3 Burn 1.2 0.8 Salvage 3.5 2.0 70 T A B L E 5 . Descriptive properties of sequences from RISA fingerprint bands with high correlation values. The correlation coefficient, tau, represents the rank relationship between the ordination scores and the individual variables. % of total Band tau value Closest GenBank reference affiliation & GenBank band intensity Coordinate Sample Class Axis 1 Axis 3 Division/Subdivision % Identity in fingerprint GSpl PI 1-17 >.39-.41 0.237 0.508 Hydrocarboniphaga effiisaly- Proteobacteria 92 17.8 GSp2 P4-1 >.39-.41 0.237 0.508 Hydrocarboniphaga effiisaly- Proteobacteria 91 25.7 GSp2P13-28 >!l9-.21c -0.352 0.171 Rhodoplanes elegans/a- Proteobacteria 96 12.6 GSp2P13-29 >.19-.21c -0.352 0.171 Acidisphaera sp. NO-15/a- Proteobacteria 94 GSp2P13-48 >.19-.21c -0.352 0.171 Methylosmus sp. LW2/a- Proteobacteria 96 GSp2P13-7 >.37-.39 -0.686 0.074 Hydrocarboniphaga effiisaly- Proteobacteria 92 22.0 GSp3 PI2-30 >.53-.55c 0.522 -0.193 Achromatium oxaliferumly- Proteobacteria 91 11.1 GSp3P12-31 >.53-.55c 0.522 -0.193 Achromatium oxaliferumly- Proteobacteria 91 GSp3 PI2-49 >.53-.55c 0.522 -0.193 Frankia sp. MRn2-2/Actinobacteria 89 GSp3 P9-20 >.31-.33 -0.198 0.373 Desulfatibacillum aliphaticivoranslo-Prnteobaclena 87 17.1 GSp3 PI-8 >.19-.21 -0.352 0.171 Asterophora /wras/ft'cfl/Basidiornycota 97 10.4 GSp3 Pl-9 >.37-.39 -0.686 0.074 Hydrocarboniphaga effiisaly- Proteobacteria 97 41.8 GSp3Pl-10 >.37-.39 -0.686 0.074 Hydrocarboniphaga effiisaly- Proteobacteria 93 GSp3 P7-2 >.39-.41 0.237 0.508 Hydrocarboniphaga effiisaly- Proteobacteria 92 33.2 HSpl P5-3 >.39-.41 0.237 0.508 Hydrocarboniphaga eflusa/y- Proteobacteria 91 11.3 HSpl P9-35 >.19-.21c -0.352 0.171 Hyphomicrobium vulgarela- Proteobacteria 91 10.3 HSpl P9-36 >.19-.21c -0.352 0.171 Desulfatibacillum aliphaticivorans/5-Pmteobacteria 90 HSpl P9-51 >.19-.21c -0.352 0.171 Hydrocarboniphaga effiisaly- Proteobacteria 86 HSpl P9-60 >.19-.21c -0.352 0.171 Desulfatibacillum aliphaticivorans/o-Proteobacteria 90 HSpl P9-61 >.19-.21c -0.352 0.171 Desulfatibacillum aliphaticivoranslS-Proteobactem 90 HSpl P9-11 >.37-.39 -0.686 0.074 Hydrocarboniphaga effiisaly- Proteobacteria 92 9.3 HSpl P14-5 >.39-.41 0.237 0.508 Asterophora lycoperdoideslBasidiomycota 96 15.6 HSp2 P2-4 >.37-.39 -0.686 0.074 Beggiatoa sp. A ASA/y- Proteobacteria 91 23.4 HSp2P13-23 >.39-.41 0.237 0.508 Hydrocarboniphaga e#Ksa/7-Proteobacteria 92 17.4 71 % of total Band tau value Closest GenBank reference affiliation & GenBank band intensity Coordinate Sample Class Axis 1 Axis 3 Division/Subdivision % Identity in fingerprint HSp3 PI0-56 >.19-.21c -0.352 0.171 Bradyrhizobium sp. ORS 3 2 60/a-Proteobacteria 94 9.2 HSp3P10-57 >.19-.21c -0.352 0.171 Bradyrhizobium sp. bfslb/a-Proteobacteria 90 HSp3P10-50 >.19-.21c -0.352 0.171 Methylosinus sp. LW2/a-Proteobacteria 96 HSp3P10-32 >.39-.41c 0.237 0.508 Hydrocarboniphaga e$/sa/y-Proteobacteria 92 15.4 HSp3 P10-33 >.39-.41c 0.237 0.508 Hydrocarboniphaga e$«sa/y-Proteobacteria 92 HSp3 P10-34 >.39-.41c 0.237 0.508 Hydrocarboniphaga ej?usa/y-Proteobacteria 92 HSp3 PI0-47 >.39-.41c 0.237 0.508 Methylocystis parvus/a-Proteobacteria 95 HSp3 P9-21 >.31-.33 -0.198 0.373 Flavobacterium ferrugineumlCFW 90 18.3 BSpl P6-14 >.53-.55 0.522 -0.193 Thermodesulforhabdus norveg/cws/8-Proteobacteria 89 10.8 BSpl P6-15 >.59-.61 0.404 -0.275 Bacillus aminovorans/Fkmicutes 99 15.6 BSpl PI2-22 >.31-.33 -0.198 0.373 Acidovorax de/7uw7^3-Proteobacteria 97 12.7 BSpl P12-19 >.39-.41 0.237 0.508 Flavobacterium omnivorum/CFB 93 15.9 BSp2Pl-16 >.53-.55 0.522 -0.193 Bacillus aminovorans/Fkmicutes 98 15.2 BSp2Pl-6 >.59-.61 0.404 -0.275 Bacillus aminovorans/Fkmicutes 99 13.7 BSp2Pll-38 >.37-.39c -0.686 0.074 Parachlamydia acanthamoebae/Ch\amydiae 92 9.0 BSp2PU-59 >.37-.39c -0.686 0.074 Hydrocarboniphaga e^uM/y-Proteobacteria 92 BSp2Pll-53 >.37-.39c -0.686 0.074 Hydrocarboniphaga e/Tusfl/y-Proteobacteria 91 BSp2P12-39 >.37-.39c -0.686 0.074 Acidovorax rfe/7«w7/p-Proteobacteria 94 9.6 BSp2P12-41 >.37-.39c -0.686 0.074 Burkholderia (ropi'co/is/|3-Proteobacteria 95 BSp2P12-54 >.37-.39c -0.686 0.074 Flexibacter aggregans/CFB 88 BSp2 P7-25 >.59-.61 0.404 -0.275 Bacillus aminovorans/Fvmicutes 98 12.1 BSp3 PI3-52 >.39-.41c 0.237 0.508 Lysobacter 6/nne.scefl^y-Proteobacteria 87 11.7 BSp3 PI3-58 >.39-.41c 0.237 0.508 Paucimonas /emoi'gnei/p-Proteobacteria 97 BSp3 PI3-37 >.39-.41c 0.237 0.508 Rhizoctonia jo/ari'tfBasidiomycota 95 72 % of to tal Band tau value Cbsest GenBank reference affiliation & GenBank band intensity Coordinate Sample Class Axis 1 Axis 3 Division/Subdivision % Identity in fingerprint BSp3P10-18 >.65-.67 0.330 -0.299 Hydrocarboniphaga e#usa/y-Proteobacteria 92 10.4 BSp3 P8-24 >.59-.61 0.404 -0.275 Baciilus naganoensis/Firmicutes 96 9.5 SSpl P12-13 >.37-.39 -0.686 0.074 Hydrocarboniphaga e^uttj/y-Proteobacteria 92 20.7 SSp2 P7-40 >.39-.41c 0.237 0.508 Nitrospira man'na/Nitrospirae 92 18.8 SSp2 P7-46 >.39-.41c 0.237 0.508 Rathayibacter tritici/Gram positive 97 SSp2 P7-55 >.39-.41c 0.237 0.508 Flavobacterium heparinum/CFB 92 SSp2 P2-42 >.53-.55c 0.522 -0.193 Frankia ^ ^./Actinobacteria 89 17.0 SSp2 P2-45 >.53-.55c 0.522 -0.193 Achromatium oxa/^rw/n/y-Proteobacteria 90 SSp2 P2-44 >.53-.55c 0.522 -0.193 Frankia sp./Actinobacteria 89 SSp3 P7-12 >.39-.41 0.237 0.508 Asterophora /wras/ft'ca/Basidiomycota 99 12.0 SSp3 P7-27 >.59-.61c 0.404 -0.275 Bacillus a/m'flovora/is/Firmicutes 98 8.3 SSp3 P7-43 >.59-.61c 0.404 -0.275 Bacillus ammovorans/Firmicutes 98 SSp3 P7-62 >.59-.61c 0.404 -0.275 Bacillus ammovorans/Firmicutes 92 SSp3 P2-26 >.31-.33 -0.198 0.373 Variovorax paradoxus/fi-Vmteobacteria 97 10.8 GSp2 P4-1T DGGE band Uncultured y-Proteobacteria 92 8.6 GSp2 P4-2B DGGE band Chelatococcus asaccfiarowrans/a-Proteobacteria 90 13.0 HSp2 P2-3T DGGE band Uncultured y-Proteobacteria 92 3.5 HSp2 P2-4B DGGE band Rhodopseudomonas pa/usfr/s/a-Proteobacteria 96 7.1 BSpl Pl-5 DGGE band Thermomonas hydrothermalis/y-Proteobacteria 94 8.3 SSp2 P2-6 DGGE band Rhodopseudomonas pa/wsfm/a-Proteobacteria 89 5.9 a Cytophaga-Flexibacter-Bacteroides c clones from a band that didn't directly sequence 73 FIG 1. Microbial biomass-C, - N and extracted D N A in the mineral soils of the control, harvest, burn and burn-salvage treatments (n = 10). A l l values are significantly different (p < 0.05). 74 FIG 2. RISA fingerprint N M S ordination with joint plot of NO3-, NH4+ P e x t and pH superimposed indicating the relative strength and direction of correlation of variables with the ordination. Three axes produced the best ordination. A total cumulative variance of 76.1% was explained by the ordination. < O O O O Treatment O Control O Harvest I • Burn ^ Salvagel O o o o Axis 1 F I G 3. D G G E fingerprint NMS ordination with joint plot of N O 3 - , N H 4 + P e x t and pH superimposed indicating the relative strength and direction of correlation of variables with the ordination. Three axes produced the best ordination and a total cumulative variance of 63.9% was explained by the ordination. o o o o o • PN J3 o o Ok Treatment O Control O Harvest! • Burn A Salvagel Axis 1 FIG. 4. Sample RISA fingerprints with arrows indicating bands that were sequenced. 77 FIG. 5. Rooted neighbor-joining phylogenetic tree of sequences from RISA fingerprint bands. Reference sequences used: Uncultured eubacterium clone WD260 - AJ292673.1, Uncultured eubacterium clone WD272 - AJ292684.1, Uncultured bacterium clone 1174-1021-18 -AB128891.1, Uncultured alpha proteobacterium clone EBI 127 - 395446.1, Bacterium Ellin 5280 - AY234631.1, Hydrocarboniphaga effusa - AY363244.1, Pseudomonas sp. 273 -AF039488.1, Sinobacter luteus - AY966001.1, Variovoraxparadoxus - AF532868.1, Rhodoplanes elegans - AF487437.1, Bacterium Ellin 5220 - AY234571.1, Flavobacterium ferrugineum - M62798.1, Bacterium Ellin 5025 - AY234442.1, Frankia sp. - AF063641.1, Bacillus naganoensis - AB021193.1, Parachlamydia acanthamoebae - Y07556.1, Nitrospira marina - X82559.1 and rooted with Aquifex aeolicus - AJ309733.1. Numbers at the nodes are bootstrap values > 50 made with 1000 bootstrap resamplings. The bar represents 0.1 substitutions per site. 0.1 substitutions/site I 1 Si l l 52i-"it! GSp3 Pl-9, SSpl P12-13 f-HSpl P5-3 GSp3 Pl-10 GSp2 PI 3-7 GSp2 P4-1 HSp2P2-4 GSp3 P7-2, HSpl P9-11, GSpl PI 1-17, HSp3 PI 0-33 BSp3 P10-18 HSp3 P10-34 HSp3 P10-32 BSp2 PI 1-53 l_r Uncultured eubacterium clone WD260 96l-BSp2 PI 1-59 HSp2 P13-23 97r Bacterium Ellin 5280 ~ HSpl P9-51 Hydrocarboniphaga effusa Pseudomonas sp. 273 \W>r-Sinohacter luteus fl LBSp3 P13-58 I 1 BSp2 PI 2-41 BSp2 P12-39 BSpl P12-22 Variovorax paradoxus SSp3 P2-26 GSp2 P13-29 HSpl P9-35 GSp2 P13-48, HSp3 P10-50 HSp3 P10-47 Rhodoplanes elegans GSp2 P13-28 Uncultured clone EBI 127 HSp3 P10-57 «5iHSp3 P10-56 Parachlamydia acanthamoebae BSp2 PI 1-38 a-•2! a o GSp3 PI 2-49 SSp2 P2-42, SSp2 P2-44 Bacterium Ellin 5025 •Frankia sp. SSp2 P7-46 •Bacillus naganoensis BSp3 P8-24 ,nnrBSp2 P7-25 W BSpl P6-15, BSp2 Pl-6, SSp3 P7-27, SSp3 P7-43 SSp3 P7-62 SSp2 P7-55 I BSpl P12-19 Flavobacterium ferrugineum HSp3 P9-21 Uncultured bacterium 1174-1021-18 GSp3 P9-20 HSpl P9-60, HSpl P9-61 HSpl P9-36 Uncultured eubacterium WD272 J •Parachlamydia "Actinobacteria 'Bacillus - Gemmatimonadetes • Bacteroidetes . Environmental ' sample loor -Aquifex aeolicus -Nitrospira marina -SSp2 P7-40 -BSp3 PI 3-52 • Nitrospirae 78 R E F E R E N C E S C H A P T E R 1 R E F E R E N C E S 1. Agee, J . K . 1990. The historical role of fire in Pacific Northwest forests, p. 25-38. In J. D. Walstad, S. R. Radosevich, and D. V . Sandberg (eds), Natural and prescribed fire in Pacific Northwest forests. Oregon State University Press, Corvallis, Oregon. 2. Almendros, G . , F . J . Gonzalez-Vila, and F . Mart in . 1990. Fire-induced transformation of soil organic matter from an oak forest: An experimental approach to the effects of fire on humic substances. Soil Sci. 149:158-168. 3. Alvarez, A . J . , M . Khanna, G . A . Toranzos, and G . Stotzky. 1998. Amplification of D N A bound on clay minerals. Mol . Ecol. 7:775-778. 4. Anderson, T . - H . , and K . H . Domsch. 1989. Ratios of microbial biomass carbon to total organic carbon in arable soils. Soil Biol. Biochem. 21:471-479. 5. A r n o , S. F . , and S. Allison-Bunnell. 2002b. Environmental Impacts: Fire's Influence on Soils, Water, and Air, p. 90-91, 94-95, In Flames in Our Forest: disaster or renewal? Island Press, Washington, D.C. 6. A r n o , S. F . , and S. Allison-Bunnell. 2002a. Fire History: Discovering Effects of Past Fires in a Forest, p. 103-117, Flames in our Forest: disaster or renewal? Island Press, Washington, D.C. 7. Axelrood, P. E . , M . L . Chow, C . S. Arno ld , K . L u , J . M . McDermott , and J . Davies. 2002a. Cultivation-dependent characterization of bacterial diversity from British Columbia forest soils subjected to disturbance. Can. J. Microbiol. 48:643-654. 8. Axelrood, P. E . , M . L . Chow, C . C . Radomski, J . M . McDermott , and J . Davies. 2002b. Molecular characterization of bacterial diversity from British Columbia forest soils subjected to disturbance. Can. J. Microbiol. 48:655-674. 9. Baath, E . 1980. Soil fungal biomass after clear-cutting of a pine forest in central Sweden. Soil Biol. Biochem. 12:495-500. 10. Baath, E . , A. Frostegard, T . Pennanen, and H . Fritze. 1995. Microbial community structure and pH response in relation to soil organic matter quality in wood-ash fertilized, clear-cut or burned coniferous forest soils. Soil Biol. Biochem. 27:229-240. 11. Backman, J . S. K . , A. K . Klemedtsson, L . Klemedtsson, and P . - E . Lindgren. 2004. Clear-cutting affects the ammonia-oxidising community differently in limed and non-limed coniferous forest soils. Biol. Fertil. Soils 40:260-267. 12. Bauhus, J . , P. K . Khanna , and R. J . Raison. 1993. The effect of fire on carbon and nitrogen mineralization and nitrification in an Australian forest soil. Aust. J. Res. 31:621-639. 13. Bissett, J . , and D . Parkinson. 1980. Long-term effects of fire on the composition and activity of the soil microflora of a subalpine, coniferous forest. Can. J. Bot. 58:1704-1721. 14. Bock, M . D. , and K . C . J . V a n Rees. 2002. Forest harvesting impacts on soil properties and vegetation communities in the Northwest Territories. Can. J. For. Res. 32:713-724. 15. Borneman, J . , and E . W . Triplett. 1997. Molecular microbial diversity in soils from Eastern Amazonia: Evidence for unusual microorganisms and microbial population shifts associated with deforestation. Appl. Environ. Microbiol. 63:2647-2653. 16. Chang, Y . - J . , J . R. Stephen, A . P. Richter, A . D . Venosa, J . Briiggemann, S. J . Macnaughton, G . A . Kowalchuk, J . R. Haines, E . Kline, and D . C . White. 2000. Phylogenetic analysis of aerobic freshwater and marine enrichment cultures efficient in hydrocarbon degradation: effect of profiling method. J. Microbiol. Methods 40:19-31. 79 17. Choromanska, U . , and T . H . DeLuca. 2002. Microbial activity and nitrogen mineralization in forest mineral soils following heating: evaluation of post-fire effects. Soil Biol. Biochem. 34:263-271. 18. Chisholm Fire Review Committee. 2001b. Final documentation report Chisholm fire (LWF-063) - Fire behaviour documentation. Submitted to the Minister of Alberta Sustainable Resource Development - Forest protection division. 19. Chisholm Fire Review Committee. 2001a. Final documentation report Chisholm fire (LWF-063) - Overview. Submitted to the Minister of Alberta Sustainable Resource Development - Forest protection division. 20. Chisholm Fire Review Committee. 2001c. Final documentation report Chisholm fire (LWF-063) - Resource impacts and values. Submitted to the Minister of Alberta Sustainable Resource Development - Forest protection division. 21. Conover, W . J . 1980. Some methods based on ranks, p. 213-216. In R. A . Bradley, J. S. Hunter, D. G. Kendall, and G. S. Watson (eds), Practical nonparametric statistics. John Wiley & Sons, Inc., New York. 22. Corbet, A . S. 1934. Studies on tropical soil microbiology: II. The bacterial numbers in the soil of the Malay peninsula. Soil Sci. 38:407-416. 23. Covington, W . W . , and S. S. Sackett. 1992. Soil mineral nitrogen changes following prescribed burning in ponderosa pine. For. Ecol. Manag. 54:175-191. 24. Crecchio, C , and G . Stotzky. 1998. Binding of D N A on humic acids: effect on transformation on Bacillus subtilis and resistance to DNase. Soil Biol. Biochem. 30:1061-1067. 25. Crosby, L . D . , and C . S. Criddle. 2003. Understanding bias in microbial community analysis techniques due to rrn operon copy number heterogeneity. BioTechniques 34:790-802. 26. C u r c i , M . , M . D . R. Pizzigallo, C . Crecchio, R. Mininni , and P. Ruggiero. 1997. Effects of conventional tillage on biochemical properties of soils. Biol. Fertil. Soils 25:1-6. 27. Daffonchio, D . , A . Cherif, and S. Borin. 2000. Homoduplex and heteroduplex polymorphisms of the amplified ribosomal 16S-23S internal transcribed spacers describe genetic relationships in the "Bacillus cereus Group". Appl. Environ. Microbiol. 66:5460-5468. 28. DeBano, L . F . , and C . E . Conrad. 1978. The effect of fire on nutrients in a chaparral ecosystem. Ecology 59:489-497. 29. DeBano, L . F . , G . E . Eberlein, and P. H . Dunn. 1979. Effects of burning on chaparral soils. I. Soil nitrogen. Soil Sci. Soc. Am. J. 43:504-509. 30. DeBano, L . F . , L . D . M a n n , and D . A . Hamilton. 1970. Translocation of hydrophobic substances into soil by burning organic litter. Soil Sci. Soc. Am. Proc. 34:130-133. 31. DeBano, L . F . , D . G . Neary, and P. F . Ffolliott. 1998. Fire's effects on ecosystems. J. Wiley, New York. 32. DeBano, L . F . , S. M . Savage, and D . A . Hamilton. 1976. The transfer of heat and hydrophobic substances during burning. Soil Sci. Soc. Am. J. 40:779-782. 33. Donskey, C . J . , A . M . Hujer, S. M . Das, N . J . Pultz, R. A . Bonomo, and L . B . Rice. 2003. Use of denaturing gradient gel electrophoresis for analysis of the stool microbiota of hospitalized patients. J. Microbiol. Methods 54:249-256. 34. Dumontet, S., H . Dinel, A . Scopa, A . Mazzatura, and A . Saracino. 1996. Post-fire soil microbial biomass and nutrient content of a pine forest soil from a dunal Mediterranean environment. Soil Biol. Biochem. 10/11:1467-1475. 35. Dunn, P. H . , and L . F . DeBano. 1977. Fire's effect on physical and chemical properties of chaparral soils, p. 75-84. Proceedings of the Symposium on the Environmental 80 Consequences of Fire and Fuel Management in Mediterranean Ecosystems. U S D A Forest Service General Technical Report. WO-3. 36. Dyrness, C . T . 1976. Effect of wildfire on soil wettability in the High Cascades of Oregon. U S D A Forest Service, Research Paper PNW-202. 37. Entry, J . A . , N . M . Stark, and H . Loewenstein. 1986. Effect of timber harvesting on microbial biomass fluxes in a northern Rocky Mountain forest soil. Can. J. For. Res. 16:1076-1081. 38. Fenn, M . E . , M . A . Poth, P. H . Dunn, and S. C . Barro. 1993. Microbial N and biomass, respiration and N mineralization in soils beneath two chaparral species along a fire-induced age gradient. Soil Biol. Biochem. 25:457-466. 39. Ferris, M . J . , and D . M . W a r d . 1997. Seasonal distributions of dominant 16S rRNA-defined populations in a hot spring microbial mat examined by denaturing gradient gel electrophoresis. Appl. Environ. Microbiol. 63:1375-1381. 40. Field, J . P., K . W . Farrish, and E . A . Carter. 2003. Soil and nutrient losses following site preparation burning in a harvested loblolly pine site. Trans. A S A E (Am. Soc. Agric. Eng.) 46:1697-1703. 41. Fisher, M . M . , and E . W . Triplett. 1999. Automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Appl. Environ. Microbiol. 65:4630-4636. 42. Fonturbel, M . T . , J . A . Vega, S. Bara , and I. Bernardez. 1995. Influence of prescribed burning of pine stands in N W Spain on soil microorganisms. Eur. J. Soil Biol. 31:13-20. 43. Frey, B . R. , V . J . Lieffers, A . D . Munson, and P. V . Blenis. 2003. The influence of partial harvesting and forest floor disturbance on nutrient availability and understory vegetation in boreal mixedwoods. Can. J. For. Res. 33:1180-1188. 44. Fritze, H . , A . Smolander, T . Levula, V . Kitunen, and E . Malkonen. 1994. Wood-ash fertilization and fire treatments in a Scots pine forest stand: effects on the organic layer, microbial biomass and microbial activity. Biol. Fertil. Soils 17:57-63. 45. Garcia , F . O . , and C . Rice. 1994. Microbial biomass dynamics in tallgrass prairie. Soil Sci. Soc. Am. J. 58:816-823. 46. Giovannini, G . , S. Lucchesi, and S. Cervelli. 1983. Water-repellent substances and aggregate stability in hydrophobic soil. Soil Sci. 135:110-113. 47. Goto, K . , T . O m u r a , Y . H a r a , and Y . Sadaie. 2000. Application of the partial 16S rDNA sequence as an index for rapid identification of species in the genus Bacillus. J. Gen. Appl. Microbiol. 46:1-8. 48. Grayston, S. J . , and C . D . Campbell . 1996. Functional biodiversity of microbial communities in the rhizospheres of hybrid larch (Larix eurolepis) and Sitka spruce (Picea sitchensis). Tree Physiol. 16:1031-1038. 49. Giirtler, V . 1993. Typing of Clostridium difficile strains by PCR-amplification of variable length 16S-23S rDNA spacer regions. J. Gen. Microbiol. 139:3089-3097. 50. Giirtler, V . , and V . A . Stanisich. 1996. New approaches to typing and identification of bacteria using the 16S-23S rDNA spacer region. Microbiology 14:3-16. 51. Hannam, K . D . , and C . E . Prescott. 2003. Soluble organic nitrogen in forests and adjacent clearcuts in British Columbia, Canada. Can. J. For. Res. 33:1709-1718. 52. Hernandez, T . , C . Garc ia , and I. Reinhardt. 1997. Short-term effect of wildfire on the chemical, biochemical and microbiological properties of Mediterranean pine forest soils. Biol. Fertil. Soils 25:109-116. 53. Hossain, A . K . M . A . , R. J . Raison, and P. K . Khanna. 1995. Effects of fertilizer application and fire regime on soil microbial biomass carbon and nitrogen, and nitrogen mineralization in an Australian subalpine eucalypt forest. Biol . Fertil. Soils 19:246-252. 81 54. Houston, A . P. C , S. Visser, and R. A . Lautenschlager. 1998. Microbial processes and fungal community structure in soils from clear-cut and unharvested areas of two mixedwood forests. Can. J. Bot. 76:630-640. 55. Jenkinson, D . S., and J . N . L a d d . 1981. Microbial biomass in soil: Measurement and turnover, p. 415-471. In E. A. Paul, and J. N . Ladd (eds), Soil biochemistry, vol. 5. Marcel Dekker, New York. 56. Jensen, M . A . , and N . Straus. 1993. Effect of PCR conditions on the formation of heteroduplex and single-stranded D N A products in the amplification of bacterial ribosomal D N A spacer regions. PCR Methods and Appl. 3:186-194. 57. Jorgensen, J . R. , and C . G . Wells. 1971. Apparent nitrogen fixation in soil influenced by prescribed burning. Soil Sci. Soc. Am. Proc. 35:806-810. 58. Kasischke, E . S. 2000. Boreal Ecosystems in the Global Carbon Cycle, p. 24-25. In G. H . M . M . Caldwell, O. L. Lange, H . A . Mooney, E.-D. Schulze, and U . Sommer (ed.), Fire, Climate Change, and Carbon Cycling in the Boreal Forest, Ecological Studies, vol. 138. Springer-Verlag, New York, N Y . 59. Kauffman, J . B . 1990. Ecological relationships of vegetation and fire in Pacific Northwest forests, p. 39-52. In J. D. Walstad, S. R. Radosevich, and D. V . Sandberg (eds), Natural and prescribed fire in Pacific Northwest forests. Oregon State University Press, Corvallis, Oregon. 60. Kendall , M . G . 1938. A new measure of rank correlation. Biometrika 30:81-93. 61. Khetmalas, M . B. , K . N . Egger, H . B . Massicotte, L . E . Tackaberry, and M . J . Clapperton. 2002. Bacterial diversity associated with subalpine fir (Abies lasiocarpa) ectomycorrhizae following wildfire and salvage-logging in central British Columbia. Can. J. Microbiol. 48:611-625. 62. Kisand, V . , and J . Wikner. 2003. Limited resolution of 16S rDNA D G G E caused by melting properties and closely related D N A sequences. J. Microbiol. Methods 54:183-191. 63. Knoepp, J . D. , J . M . Vose, and W . T . Swank. 2004. Long-term soil responses to site preparation burning in the southern appalachians. For. Sci. 50:540-550. 64. Kovacic, D . A . , D . M . Swift, J . E . Ellis, and T . E . Hakonson. 1986. Immediate effects of prescribed burning on mineral soil nitrogen in ponderosa pine of New Mexico. Soil Sci. 141:71-76. 65. Kutiel , P., and A . Shaviv. 1989. Effect of simulated forest fire on the availability of N and P in mediterranean soils. Plant Soil 120:57-63. 66. Larue , R. , Z . Y u , V . A . Parisi, A . R. Egan, and M . Morrison. 2005. Novel microbial diversity adherent to plant biomass in the herbivore gastrointestinal tract, as revealed by ribosomal intergenic spacer analysis and rrs gene sequencing. Environ. Microbiol. 7:530-543. 67. Lieffers, V . J . , S. E . Macdonald, and E . H . Hogg. 1993. Ecology of and control strategies for Calamagrostis canadensis in boreal forest sites. Can. J. For. Res. 23:2070-2077. 68. Lindeburgh, S. B . 1990. Effects of prescribed fire on site productivity: a literature review 66. Ministry of Forests. 69. Lorenz, M . G . , and W . Wackernagel. 1987. Adsorption of D N A to sand and variable degradation rates of adsorbed DNA. Appl. Environ. Microbiol. 53:2498-2952. 70. Lundgren, B . 1982. Bacteria in a pine forest soil as affected by clear-cutting. Soil Biol. Biochem. 14:537-542. 71. Lynham, T . J . , G . M . Wickware, and J . A . Mason. 1998. Soil chemical changes and plant succession following experimental burning in immature jack pine. Can. J. Soil Sci. 78: 93-104. 82 72. Maars , R. H . , R. D . Roberts, R. A . Skeffinton, and A . D . Bradshaw (eds). 1983 Nitrogen in the development of ecosystems. Blackwell Science Publishing, Oxford, England. 73. Mabuhay, J . A . , N . Nakagoshi, and T . Horikoshi. 2003. Microbial biomass and abundance after forest fire in pine forests in Japan. Ecol. Res. 18:431-441. 74. Machul la , G . 2003. Soil microbial indicators and their environmental significance. J. Soils Sediments 3:229. 75. Marstorp, H . , X . G u a n , and P. Gong. 2000. Relationship between dsDNA, chloroform labile C and ergosterol in soils of different organic matter contents and pH. Soil Biol. Biochem. 32:879-882. 76. Martikainen, P. J . , and A . Palojarvi. 1990. Evaluation of the fumigation-extraction method for the determination of microbial C and N in a range of forest soils. Soil Biol . Biochem 22:797-802. 77. Mart in , R. E . 1990. Goals, methods, and elements of prescribed burning, p. 55-66. In J. D. Walstad, S. R. Radosevich, and D. V . Sandberg (eds), Natural and prescribed fire in Pacific Northwest forests. Oregon State University Press, Corvallis, Oregon. 78. M c N a b b , D . H . , F . Gaweda, and H . A . Froehlich. 1989. Infiltration, water repellency, and soil moisture content after broadcast burning a forest site in southwest Oregon. J. Soil Water Conserv. 44:87-90. 79. Mil ler , M . 2000. Fire autecology, p. 9-34. In J. K . Brown, and J. K . Smith (eds), Wildland fire in ecosystems: effects of fire on flora, vol. General Technical Report RM-42-Vol. 2. U S D A Forest Service, Rocky Mountain Research Station, Fort Collins, Colorado. 80. M u m y , K . L . , and R. H . Findlay. 2004. Convenient determination of D N A extraction efficiency using an external D N A recovery standard and quantitative-competitive PCR. J. Microbiol. Methods 57:259-268. 81. Muyzer , G . , E . C . De Waal , and A . G . Uitterlinden. 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59:695-700. 82. Myers, R. M . , S. G . Fischer, L . S. Lerman, and T . Maniatis. 1985. Nearly all single base substitutions in D N A fragments joined to a GC-clamp can be detected by denaturing gradient gel electrophoresis. Nucleic Acids Res. 13:3131-3145. 83. Niemela, S., and V . Sundman. 1977. Effects of clear-cutting on the composition of bacterial populations of northern spruce forest soil. Can. J. Microbiol. 23:131-138. 84. Nsabimana, D . , R. J . Haynes, and F . M . Wallis. 2004. Size, activity and catabolic diversity of the soil microbial biomass as affected by land use. Appl. Soil Ecology 26:81-92. 85. Ogino, A . , H . Koshikawa, T . Nakahara, and H . Uchiyama. 2001. Succession of microbial communities during a biostimulation process as evaluated by D G G E and clone library analyses. J. Appl. Microbiol. 91:625-635. 86. Pietikainen, J . , and H . Fritze. 1995. Clear-cutting and prescribed burning in coniferous forest: Comparison of effects on soil fungal and total microbial biomass, respiration activity and nitrification. Soil Biol. Biochem. 27:101-109. 87. Pietikainen, J . , and H . Fritze. 1993. Microbial biomass and activity in the humus layer following burning: Short-term effects of two different fires. Can. J. For. Res. 23:1275-1285. 88. Pietikainen, J . , R. Hiukka , and H . Fritze. 2000. Does short-term heating of forest humus change its properties as a substrate for microbes? Soil Biol. Biochem. 32:277-288. 83 89. Polz, M . F . , and C . M . Cavanaugh. 1998. Bias in template-to-product ratios in multi-template PCR. Appl. Environ. Microbiol. 64:3724-3730. 90. Prescott, C . E . 1997. Effects of clear cutting and alternative silvicultural systems on rates of decomposition and nitrogen mineralization in a coastal montane coniferous forest. For. Ecol. Manag. 95:253-260. 91. Prieto-Fernandez, A . , M . J . Acea, and T . Carballas. 1998. Soil microbial and extractable C and N after wildfire. Biol. Fertil. Soils 27:132-142. 92. Pyne, S. J . 1984. Introduction to wildland fire management in the United States. John Wiley and Sons, Inc., New York. 93. Ranjard, L . , E . Brothier, and S. Nazaret. 2000. Sequencing bands of ribosomal intergenic spacer analysis fingerprints for characterization and microscale distribution of soil bacterium populations responding to mercury spiking. Appl. Environ. Microbiol. 66:5334-5339. 94. Ritchie, N . J . , M . E . Schutter, R. P. Dick, and D . D . M y r o l d . 2000. Use of length heterogeneity PCR and fatty acid methyl ester profiles to characterize microbial communities in soil. Appl. Environ. Microbiol. 66:1668-1675. 95. Romanowski, G . , M . G . Lorenz, and W . Wackernagel. 1991. Adsorption of plasmid D N A to mineral surfaces and protection against DNase I. Appl. Environ. Microbiol. 57:1057-1061. 96. Romme, W . H . , and M . G . Turner . 2004. Ten years after the 1988 Yellowstone fires: Is restoration needed?, p. 318-361. In L. L. Wallace (ed.), After the fires: the ecology of change in Yellowstone National Park. Yale University Press, New Haven. 97. Sigler, W . V . , and J . Zeyer. 2002. Microbial diversity and activity along the forefields of two receding glaciers. Microb. Ecol. 43:397-407. 98. Smalla, K . , G . Wieland, A . Buchner, A . Zock, J . Parzy, S. Kaiser, N . Roskot, H . Heuer, and G . Berg. 2001. Bulk and rhizosphere soil bacterial communities studied by denaturing gradient gel electrophoresis: Plant-dependent enrichment and seasonal shifts revealed. Appl. Environ. Microbiol. 67:4742-4751. 99. Smith, D . W . , and T . D . James. 1978. Characteristics of prescribed burns and resultant short-term environmental changes in Populus tremuloides woodland in southern Ontario. Can. J. Bot. 56:1782-1791. 100. Smith, N . R. , Z . Y u , and W . W . M o h n . 2003. Stability of the bacterial community in a pulp mill effluent treatment system during normal operation and a system shutdown. Water Res. 37:4873-4884. 101. Smolander, A . , V . Kitunen, and E . Malkonen. 2001. Dissolved soil organic nitrogen and carbon in a Norway spruce stand and an adjacent clear-cut. Biol. Fertil. Soils 33:190-196. 102. Sparling, G . P. 1992. Ratio of microbial biomass carbon to soil organic carbon as a sensitive indicator of changes in soil organic matter. Aust. J. Soil Res. 30:195-207. 103. Staddon, W . J . , L . C . Duchesne, and J . T . Trevors. 1996. Conservation of forest soil microbial diversity: the impact of fire and research needs. Environ. Rev. 4:267-275. 104. Staddon, W . J . , L . C . Duchesne, and J . T . Trevors. 1997. Microbial diversity and community structure of postdisturbance forest soils as determined by sole-carbon-source utilization patterns. Microb. Ecol. 34:125-130. 105. Theodorou, C , and G . D . Bowen. 1982. Effects of a bushfire on the microbiology of a south Australian low open (dry sclerophyll) forest soil. Aust. For. Res. 12:317-327. 106. Ulery, A . L . , R . C . Graham, and L . H . Bowen. 1996. Forest fire effects on soil phyllosilicates in California. Soil Sci. Soc. Am. J. 60:309-315. 107. Vazquez, F . J . , M . J . Acea, and T . Carballas. 1993. Soil microbial populations after wildfire. FEMS Microbiol. Ecol. 13:93-104. 84 108. Wardle , D . A . 1992. A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil. Biol. Rev. 67:321-358. 109. Wardle , D . A . , O . Zackrisson, and M . - C . Nilsson. 1998. The charcoal effect in Boreal forests: mechanisms and ecological consequences. Oecologia 115:419-426. 110. Yeager, C . M . , D . E . Northup, C . C . Grow, S. M . Barns, and C . R. Kuske. 2005. Changes in nitrogen-fixing and ammonia-oxidizing bacterial communities in soil of a mixed conifer forest after wildfire. Appl. Environ. Microbiol. 71:2713-2722. 111. Y u , Z . , and W . W . M o h n . 2001. Bacterial diversity and community structure in an aerated lagoon revealed by ribosomal intergenic spacer analyses and 16S ribosomal D N A sequencing. Appl. Environ. Microbiol. 67:1565-1574. 112. Y u , Z . , and M . Morrison. 2004. Comparisons of different hypervariable regions of rrs genes for use in fingerprinting of microbial communities by PCR-denaturing gradient gel electrophoresis. Appl. Environ. Microbiol. 70:4800-4806. 113. Zackrisson, O . , M . - C . Nilsson, and D . A . Wardle. 1996. Key ecological function of charcoal from wildfire in the Boreal forest. Oikos 77:10-19. 114. Zak , D . R. , D . F . Grigal , and L . F . Ohmann. 1993. Kinetics of microbial respiration and nitrogen mineralization in Great Lakes forests. Soil Sci. Soc. Am. J. 57:1100-1106. 115. Zvyagintsev, D . G . , I. P. Bab'eva, T . G . Dobrovol'skaya, G . M . Zenova, L . V . Lysak, and T . G . Mirchink. 1993. Vertically layered organization of microbial communities in forest biogenocenoses. Mikrobiologiya 62:5-36. 85 C H A P T E R 2 R E F E R E N C E S 1. Agee, J . K . 1990. The historical role of fire in Pacific Northwest forests, p. 25-38. In J. D. Walstad, S. R. Radosevich, and D. V . Sandberg (eds), Natural and prescribed fire in Pacific Northwest forests. Oregon State University Press, Corvallis, Oregon. 2. Ahlgren, I. F . (ed.) 1974 The effect of fire on soil organisms. Academic Press, New York San Francisco London. 3. Ahlgren, I. F . , and C . E . Ahlgren. 1965. Effects of prescribed burning on soil microorganisms in a Minnesota jack pine forest. Ecology 46:304-310. 4. Almendros, G . , F . J . Gonzalez-Vila, and F . Mart in . 1990. Fire-induced transformation of soil organic matter from an oak forest: An experimental approach to the effects of fire onhumic substances. Soil Sci. 149:158-168. 5. Alvarez, A . J . , M . Khanna, G . A . Toranzos, and G . Stotzky. 1998. Amplification of D N A bound on clay minerals. Mol . Ecol. 7:775-778. 6. Anderson, T . - H . , and K . H . Domsch. 1989. Ratios of microbial biomass carbon to total organic carbon in arable soils. Soil Biol. Biochem. 21:471-479. 7. A r n o , S. F . , and S. Allison-Bunnell. 2002b. Environmental Impacts: Fire's Influence on Soils, Water, and Air, p. 90-91, 94-95, In Flames in Our Forest: disaster or renewal? Island Press, Washington, D.C. 8. Axelrood, P. E . , M . L . Chow, C . S. Arnold , K . L u , J . M . McDermott , and J . Davies. 2002a. Cultivation-dependent characterization of bacterial diversity from British Columbia forest soils subjected to disturbance. Can. J. Microbiol. 48:643-654. 9. Axelrood, P. E . , M . L . Chow, C . C . Radomski, J . M . McDermott , and J . Davies. 2002b. Molecular characterization of bacterial diversity from British Columbia forest soils subjected to disturbance. Can. J. Microbiol. 48:655-674. 10. Baath, E . 1980. Soil fungal biomass after clear-cutting of a pine forest in central Sweden. Soil Biol. Biochem. 12:495-500. 11. Baath, E . , A . Frostegard, T . Pennanen, and H . Fritze. 1995. Microbial community structure and pH response in relation to soil organic matter quality in wood-ash fertilized, clear-cut or burned coniferous forest soils. Soil Biol. Biochem. 27:229-240. 12. Backman, J . S. K . , A . K . Klemedtsson, L . Klemedtsson, and P . - E . Lindgren. 2004. Clear-cutting affects the ammonia-oxidising community differently in limed and non-limed coniferous forest soils. Biol . Fertil. Soils 40:260-267. 13. Belotte, D . , J . B . Curien, R. C . Maclean, and G . Bell. 2003. A n experimental test of local adaptation in soil bacteria. Evolution 57:27-36. 14. Bissett, J . , and D . Parkinson. 1980. Long-term effects of fire on the composition and activity of the soil microflora of a subalpine, coniferous forest. Can. J. Bot. 58:1704-1721. 15. Borneman, J . , and E . W . Triplett. 1997. Molecular microbial diversity in soils from Eastern Amazonia: Evidence for unusual microorganisms and microbial population shifts associated with deforestation. Appl. Environ. Microbiol. 63:2647-2653. 16. Bosatta, E . , and G . I. Agren . 1994. Theoretical analysis of microbial biomass dynamics in soils. Soil Biol. Biochem. 26:143-148. 17. Bray, J . R. , and J . T . Curtis. 1957. An Ordination of the Upland Forest Communities of Southern Wisconsin. Ecol. Monogr. 27:325-349. 18. Brookes, P. C , A . Landman, G . Pruden, and D . S. Jenkinson. 1985. Chloroform fumigation and the release of soil nitrogen: A rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biol. Biochem. 17:837-842. 19. Cabrera, M . L . , and M . H . Beare. 1993. Alkaline persulfate oxidation for determining total nitrogen in microbial biomass extracts. Soil Sci. Soc. Am. J. 57:1007-1012. 86 20. Choromanska, U . , and T . H . DeLuca. 2002. Microbial activity and nitrogen mineralization in forest mineral soils following heating: evaluation of post-fire effects. Soil Biol. Biochem. 34:263-271. 21. Clarke, K . R. 1993. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 18:117-143. 22. Cole, J . R. , B . Chai , T . L . M a r s h , R. J . Farris , Q . Wang, S. A . K u l a m , S. Chandra , D . M . M c G a r r e l l , T . M . Schmidt, G . M . Garrity , and J . M . Tiedje. 2003. The Ribosomal Database Project (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy. Nucleic Acids Res. 31:442-3. 23. Covington, W . W . , and S. S. Sackett. 1992. Soil mineral nitrogen changes following prescribed burning in ponderosa pine. For. Ecol. Manag. 54:175-191. 24. Crecchio, C . , and G . Stotzky. 1998. Binding of D N A on humic acids: effect on transformation on Bacillus subtilis and resistance to DNase. Soil Biol. Biochem. 30:1061-1067. 25. Daffonchio, D . , A . Cherif, and S. Borin. 2000. Homoduplex and heteroduplex polymorphisms of the amplified ribosomal 16S-23S internal transcribed spacers describe genetic relationships in the "Bacillus cereus Group". Appl. Environ. Microbiol. 66:5460-5468. 26. DeBano, L . F . , L . D . M a n n , and D . A . Hamilton. 1970. Translocation of hydrophobic substances into soil by burning organic litter. Soil Sci. Soc. Am. Proc. 34:130-133. 27. DeBano, L . F . , S. M . Savage, and D . A . Hamilton. 1976. The transfer of heat and hydrophobic substances during burning. Soil Sci. Soc. Am. J. 40:779-782. 28. Dumontet, S., H . Dinel, A . Scopa, A . Mazzatura, and A . Saracino. 1996. Post-fire soil microbial biomass and nutrient content of a pine forest soil from a dunal Mediterranean environment. Soil Biol. Biochem. 10/11:1467-1475. 29. Dunn, P. H . , and L . F . DeBano 1977. Fire's effect on physical and chemical properties of chaparral soils, p. 75-84. Proceedings of the Symposium on the Environmental Consequences of Fire and Fuel Management in Mediterranean Ecosystems. U S D A Forest Service General Technical Report. WO-3. 30. Dunn, P. H . , L . F . DeBano, and G . E . Eberlein. 1979. Effects of burning on chaparral soils: II. Soil microbes and nitrogen mineralization. Soil Sci. Soc. Am. J. 43:509-514. 31. Entry , J . A . , N . M . Stark, and H . Loewenstein. 1986. Effect of timber harvesting on microbial biomass fluxes in a northern Rocky Mountain forest soil. Can. J. For. Res. 16:1076-1081. 32. Fonturbel, M . T . , J . A . Vega, S. Bara , and I. Bernardez. 1995. Influence of prescribed burning of pine stands in N W Spain on soil microorganisms. Eur. J. Soil Biol. 31:13-20. 33. Frey, B . R., V . J . Lieffers, A . D . Munson, and P. V . Blenis. 2003. The influence of partial harvesting and forest floor disturbance on nutrient availability and understory vegetation in boreal mixedwoods. Can. J. For. Res. 33:1180-1188. 34. Frostegard, A., S. Courtois, V . Ramisse, S. Clerc, D . Bernillon, F . L e G a l l , P. Jeannin, X . Nesme, and P. Simonet. 1999. Quantification of bias related to the extraction of D N A directly from soils. Appl. Environ. Microbiol. 65:5409-5420. 35. Grayston, S. J . , and C . D . Campbell . 1996. Functional biodiversity of microbial communities in the rhizospheres of hybrid larch (Larix eurolepis) and Sitka spruce (Picea sitchensis). Tree Physiol. 16:1031-1038. 36. Giirtler, V . 1993. Typing of Clostridium difficile strains by PCR-amplification of variable length 16S-23S rDNA spacer regions. J. Gen. Microbiol. 139:3089-3097. 37. Giirtler, V . , and V . A . Stanisich. 1996. New approaches to typing and identification of bacteria using the 16S-23S rDNA spacer region. Microbiology 14:3-16. 87 38. Hannam, K . D . , and C . E . Prescott. 2003. Soluble organic nitrogen in forests and adjacent clearcuts in British Columbia, Canada. Can. J. For. Res. 33:1709-1718. 39. Hernandez, T . , C . Garcia , and I. Reinhardt. 1997. Short-term effect of wildfire on the chemical, biochemical and microbiological properties of Mediterranean pine forest soils. Biol. Fertil. Soils 25:109-116. 40. Horwath, W . R. , and E . A . Paul. 1994. Microbial biomass, p. 753-773, Methods of soil analysis. Part 2 - Microbiological and biochemical properties, vol. 5. Soil Science Society of America, Inc. 41. Houston, A . P. C , S. Visser, and R. A . Lautenschlager. 1998. Microbial processes and fungal community structure in soils from clear-cut and unharvested areas of two mixedwood forests. Can. J. Bot. 76:630-640. 42. H u , S., D . C . Coleman, C . R. Carro l l , P. F . Hendrix, and M . H . Beare. 1997. Labile soil carbon pools in subtropical forest and agricultural ecosystems as influenced by management practices and vegetation types. Agric. Ecosyst. Environ. 65:69-78. 43. Huhta, V . 1971. Succession in the spider communities of the forest floor after clear-cutting and prescribed burning. Ann. Zool. Fenn. 8:483-542. 44. Insam, H . , and K . H . Domsch. 1988. Relationship between soil organic carbon and microbial biomass on chronosequences of reclamation sites. Microb. Ecol. 15:177-188. 45. Jenkinson, D . S., and J . N . L a d d . 1981. Microbial biomass in soil: Measurement and turnover, p. 415-471. In E. A . Paul, and J. N . Ladd (eds), Soil biochemistry, vol. 5. Marcel Dekker, New York. 46. Jorgensen, J . R. , and C . G . Wells. 1971. Apparent nitrogen fixation in soil influenced by prescribed burning. Soil Sci. Soc. Am. Proc. 35:806-810. 47. Jukes, T . H . , and C . R. Cantor. 1969. Evolution of protein molecules, p. 21-132. In H . H . Munro (ed.), Mammalian protein metabolism. Academic, New York. 48. Kasischke, E . S. 2000. Boreal Ecosystems in the Global Carbon Cycle, p. 24-25. In G. H . M . M . Caldwell, O. L. Lange, H . A . Mooney, E.-D. Schulze, and U . Sommer (ed.), Fire, Climate Change, and Carbon Cycling in the Boreal Forest, Ecological Studies, vol. 138. Springer-Verlag, New York, N Y . 49. Khetmalas, M . B. , K . N . Egger, H . B . Massicotte, L . E . Tackaberry, and M . J . Clapperton. 2002. Bacterial diversity associated with subalpine fir (Abies lasiocarpa) ectomycorrhizae following wildfire and salvage-logging in central British Columbia. Can. J. Microbiol. 48:611-625. 50. Kovacic, D . A . , D . M . Swift, J . E . Ellis, and T . E . Hakonson. 1986. Immediate effects of prescribed burning on mineral soil nitrogen in ponderosa pine of New Mexico. Soil Sci. 141:71-76. 51. Krebs, C . J . (ed.) 1994 Ecology: The experimental analysis of distribution and abundance, 4th edition ed. HarperCollins College Publishers, New York, N Y . 52. Kruskal , J . B . 1964a. Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika 29:1-27. 53. Kruskal , J . B . 1964b. Nonmetric multidimensional scaling: a numerical method. Psychometrika 29:115-129. 54. Kutiel , P., and Z . Naveh. 1987. The effect of fire on nutrients in a pine forest soil. Plant Soil 104:269-274. 55. Kutiel , P., and A . Shaviv. 1989. Effect of simulated forest fire on the availability of N and P in mediterranean soils. Plant Soil 120:57-63. 56. Lindeburgh, S. B . 1990. Effects of prescribed fire on site productivity: a literature review 66. Ministry of Forests. 88 57. Lorenz, M . G . , and W . Wackernagel. 1987. Adsorption of D N A to sand and variable degradation rates of adsorbed DNA. Appl. Environ. Microbiol. 53:2498-2952. 58. Lundgren, B . 1982. Bacteria in a pine forest soil as affected by clear-cutting. Soil Biol. Biochem. 14:537-542. 59. Mabuhay, J . A . , N . Nakagoshi, and T . Horikoshi. 2003. Microbial biomass and abundance after forest fire in pine forests in Japan. Ecol. Res. 18:431-441. 60. Marstorp, H . , X . Guan , and P. Gong. 2000. Relationship between dsDNA, chloroform labile C and ergosterol in soils of different organic matter contents and pH. Soil Biol. Biochem. 32:879-882. 61. Mather, P. M . 1976. Computational methods of multivariate analysis in physical geography. J. Wiley & Sons, London. 62. M c C u n e , B . , J . B . Grace, and D . L . with a contribution from Urban . 2002. Analysis of Ecological Communities. Bruce McCune, Gleneden Beach, Oregon. 63. Mielke, P. W . 1984. Meteorological applications of permutation techniques based on distance functions, p. 813-830. In P. R. K . P. K . Sen (ed.), Handbook of Statistics, vol. 4. Elsevier Science Publishers. 64. Mielke, P. W . , and K . J . Berry. 2001. Permutation methods: A distance function approach, p. 9-112. Springer-Verlag, New York, N . Y . 65. M u m y , K . L . , and R. H . Findlay. 2004. Convenient determination of D N A extraction efficiency using an external D N A recovery standard and quantitative-competitive PCR. J. Microbiol. Methods 57:259-268. 66. Muyzer, G . , E . C . De Waal , and A . G . Uitterlinden. 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59:695-700. 67. Niemela, S., and V . Sundman. 1977. Effects of clear-cutting on the composition of bacterial populations of northern spruce forest soil. Can. J. Microbiol. 23:131-138. 68. Nsabimana, D . , R. J . Haynes, and F . M . Wallis. 2004. Size, activity and catabolic diversity of the soil microbial biomass as affected by land use. Appl. Soil Ecology 26:81-92. 69. Pielou, E . C . 1966. The measurement of diversity in different types of biological collections. J. Theoret. Biol. 13:131-144. 70. Pietikainen, J . , and H . Fritze. 1995. Clear-cutting and prescribed burning in coniferous forest: Comparison of effects on soil fungal and total microbial biomass, respiration activity and nitrification. Soil Biol. Biochem. 27:101-109. 71. Pietikainen, J . , and H . Fritze. 1993. Microbial biomass and activity in the humus layer following burning: Short-term effects of two different fires. Can. J. For. Res. 23:1275-1285. 72. Pietikainen, J . , R. Hiukka , and H . Fritze. 2000. Does short-term heating of forest humus change its properties as a substrate for microbes? Soil Biol. Biochem. 32:277-288. 73. Prieto-Fernandez, A . , M . J . Acea, and T . Carballas. 1998. Soil microbial and extractable C and N after wildfire. Biol . Fertil. Soils 27:132-142. 74. Rainey, P. B . , and M . Travisano. 1998. Adaptive radiation in a heterogeneous environment. Nature 394:69-72. 75. Ranjard, L . , E . Brothier, and S. Nazaret. 2000. Sequencing bands of ribosomal intergenic spacer analysis fingerprints for characterization and microscale distribution of soil bacterium populations responding to mercury spiking. Appl. Environ. Microbiol. 66:5334-5339. 89 76. Romanowski, G . , M . G . Lorenz, and W . Wackernagel. 1991. Adsorption of plasmid D N A to mineral surfaces and protection against DNase I. Appl. Environ. Microbiol. 57:1057-1061. 77. Saitou, IS., and M . Nei. 1987. The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol . Biol. Evol. 4:406-425. 78. Schilling, E . B . , B . G . Lockaby, and R. Rummer. 1999. Belowground nutrient dynamics following three harvest intensities on the Pearl River floodplain, Mississippi. Soil Sci. Soc. Am. J. 63:1856-1868. 79. Smith, D . W . , and T . D . James. 1978. Characteristics of prescribed burns and resultant short-term environmental changes in Populus tremuloides woodland in southern Ontario. Can. J. Bot. 56:1782-1791. 80. Smolander, A . , V . Kitunen, and E . Malkonen. 2001. Dissolved soil organic nitrogen and carbon in a Norway spruce stand and an adjacent clear-cut. Biol. Fertil. Soils 33:190-196. 81. Sparling, G . P. 1992. Ratio of microbial biomass carbon to soil organic carbon as a sensitive indicator of changes in soil organic matter. Aust. J. Soil Res. 30:195-207. 82. Staddon, W . J . , L . C . Duchesne, and J . T . Trevors. 1996. Conservation of forest soil microbial diversity: the impact of fire and research needs. Environ. Rev. 4:267-275. 83. Staddon, W . J . , L . C . Duchesne, and J . T . Trevors. 1997. Microbial diversity and community structure of postdisturbance forest soils as determined by sole-carbon-source utilization patterns. Microb. Ecol. 34:125-130. 84. Sundman, V . , V . Huhta, and S. Niemela. 1978. Biological changes in northern spruce forest soil after clear-cutting. Soil Biol. Biochem 10:393-397. 85. Swetnam, T . W . , C . D . Allen, and J . L . Betancourt. 1999. Applied historical ecology: Using the past to manage for the future. Ecol. Appl. 9:1189-1206. 86. Swift, L . W . , K . J . Elliott, R. D . Ottmar, and R. E . Vihnanek. 1993. Site preparation burning to improve southern Appalachian pine-hardwood stands: fire characteristics and soil erosion, moisture, and temperature. Can. J. For. Res. 23:2242-2254. 87. Theodorou, C , and G . D . Bowen. 1982. Effects of a bushfire on the microbiology of a south Australian low open (dry sclerophyll) forest soil. Aust. For. Res. 12:317-327. 88. Thibodeau, L . , P. Raymond, C . Camire, and A . D . Munson. 2000. Impact of precommercial thinning in balsam fir stands on soil nitrogen dynamics, microbial biomass, decomposition and foliar nutrition. Can. J. For. Res. 30:229-238. 89. Thompson, J . D . , T . J . Gibson, F . Plewniak, F . Jeanmougin, and D . G . Higgins. 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882. 90. Ulery, A . L . , R . C . Graham, and L . H . Bowen. 1996. Forest fire effects on soil phyllosilicates in California. Soil Sci. Soc. Am. J. 60:309-315. 91. V a n de Peer, Y . , and R. De Wachter. 1994. TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Comput. Applic. Biosci. 10:569-570. 92. Vance, E . D . , P. C . Brookes, and D . S. Jenkinson. 1987. Microbial biomass measurement in forest soils: Determination of Kc values and tests of the hypothesis to explain the failure of the chloroform fumigation-incubation method in acid soils. Soil Biol. Biochem. 19:689-696. 93. Vazquez, F . J . , M . J . Acea, and T . Carballas. 1993. Soil microbial populations after wildfire. FEMS Microbiol. Ecol. 13:93-104. 94. W u , J . , R . G . Joergensen, B . Pommerening, and R. Chaussod. 1990. Measurement of soil microbial biomass C by fumigation-extraction - an automated procedure. Soil Biol. Biochem. 22:1167-1169. 90 95. Zvyagintsev, D. G., I. P. Bab'eva, T. G. Dobrovol'skaya, G. M. Zenova, L. V. Lysak, and T. G. Mirchink. 1993. Vertically layered organization of microbial communities in forest biogenocenoses. Mikrobiologiya 62:5-36. A P P E N D I X 1 RISA Fingerprints Chisholm, Alberta GSpl P16 Chisholm, Alberta GSpl P4 Chisholm, Alberta GSpl P6 Chisholm, Alberta GSpl P15 Chisholm, A Iberta GSpl P l l Chisholm, A Iberta GSpl P14 Chisholm, Alberta GSpl P2 Chisholm, Alberta GSpl P7 Chisholm, A Iberta GSpl PI Chisholm, Alberta GSpl P12 Chisholm, A Iberta GSp2 P10 Chisholm, Alberta GSp2 P2 Chisholm, A Iberta GSp2 P12 Chisholm, Alberta GSp2 P8 Chisholm, A Iberta GSp2 PI Chisholm, Alberta GSp2 P7 Chisholm, A Iberta GSp2 P4 Chisholm, A Iberta GSp2 P3 Chisholm, A Iberta GSp2 P14 Chisholm, Alberta GSp2 P13 Chisholm, Alberta GSp3 P2 Chisholm, Alberta GSp3 P16 Chisholm, A Iberta GSp3 P12 Chisholm, A Iberta GSp3 P6 Chisholm, Alberta GSp3 P8 Chisholm, Alberta GSp3 P10 Chisholm, A Iberta GSp3 P9 Chisholm, Alberta GSp3 PI Chisholm, A Iberta GSp3 P7 Chisholm, Alberta GSp3 P3 Chisholm, A Iberta HSpl P9 Chisholm, A Iberta HSpl P4 Chisholm, Alberta HSpl P13 Chisholm, Alberta HSpl P3 Chisholm, Alberta HSpl P5 Chisholm, A Iberta HSpl P16 Chisholm, Alberta HSpl PI Chisholm, Alberta HSpl P14 Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, A lberta Chisholm Alberta Chisholm Alberta Chisholm Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm A lberta Chisholm, Alberta Chisholm A lberta Chisholm, Alberta HSpl P8 HSpl P l l HSp2 P10 HSp2P15 HSp2 P5 HSp2 P8 HSp2 P9 HSp2 P5 HSp2 P8 HSp2 P9 HSp2P13 HSp2 P2 HSp2 P6 HSp2 P3 HSp2 P7 HSp3P10 HSp3 P6 HSp3P12 HSp3 P3 HSp3Pll HSp3 P5 HSp3 P4 HSp3 P9 BSpl P9 BSpl PI BSpl P10 BSpl P4 BSpl P2 BSpl P5 BSpl P6 BSplPll BSpl P15 BSplP12 BSp2P12 BSp2Pll BSp2Pl BSp2P10 BSp2P13 Chisholm, Alberta Chisholm, A lberta Chisholm Alberta Chisholm, Alberta Chisholm A lberta Chisholm Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, A lberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, A lberta Chisholm, A lberta Chisholm, Alberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm Alberta Chisholm A lberta Chisholm, A lberta Chisholm, Alberta Chisholm, Alberta Chisholm Alberta Chisholm, Alberta Chisholm Alberta Chisholm, Alberta Chisholm Alberta Chisholm, A lberta BSp2 P5 BSp2P15 BSp2P14 BSp2 P7 BSp2 P4 BSp3 P9 BSp3 P10 BSp3 P6 BSp3 P4 BSp3 P3 BSp3 P5 BSp3 PI BSp3 P8 BSp3 P15 BSp3 P13 SSpl P2 SSpl PI SSpl P8 SSpl P14 SSpl P9 SSpl P5 SSpl PI2 SSpl PI4 SSpl P9 SSpl P5 SSpl PI2 SSpl P6 SSpl P7 SSpl PI5 SSp2 P7 SSp2 P4 SSp2 P6 SSp2 P5 SSp2 P2 SSp2 P3 SSp2P12 SSp2P15 SSp2 P9 Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, A Iberta Chisholm, Alberta Chisholm, A Iberta Chisholm, A Iberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta SSp2P16 SSp3 P4 SSp3P12 SSp3 P5 SSp3 P7 SSp3 P2 SSp3 PI 6 SSp3 PI 5 SSp3P14 SSp3 PI SSP3 P9 A P P E N D I X 2 D G G E - V 3 Fingerprints ! ! 1 I 1 ! | 1 1 ... i 1 1 w U Chisholm, Alberta GSpl P16 Chisholm, Alberta GSpl P4 Chisholm, Alberta GSpl P6 Chisholm, Alberta GSpl P15 Chisholm, Alberta GSpl P l l Chisholm, Alberta GSpl P14 Chisholm, Alberta GSpl P2 Chisholm, Alberta GSpl P7 Chisholm, Alberta GSpl PI Chisholm, Alberta GSpl P12 Chisholm, Alberta GSp2 P10 Chisholm, Alberta GSp2 P2 Chisholm, Alberta GSp2 P12 Chisholm, Alberta GSp2 P8 Chisholm, Alberta GSp2 PI Chisholm, Alberta GSp2 P7 Chisholm, Alberta GSp2 P4 Chisholm, Alberta GSp2 P3 Chisholm, Alberta GSp2 P14 Chisholm, Alberta GSp2 P13 Chisholm, Alberta GSp3 P2 Chisholm, Alberta GSp3 P12 Chisholm, Alberta GSp3 P6 Chisholm, Alberta GSp3 P8 Chisholm, Alberta GSp3 P10 Chisholm, Alberta GSp3 P9 Chisholm, Alberta GSp3 PI Chisholm, Alberta GSp3 P7 Chisholm, Alberta GSp3 P3 Chisholm, Alberta HSpl P9 Chisholm, Alberta HSpl P4 Chisholm, Alberta HSpl P13 Chisholm, Alberta HSpl P3 Chisholm, Alberta HSpl P5 Chisholm, Alberta HSpl P16 Chisholm, Alberta HSpl PI Chisholm, Alberta HSpl P14 Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta HSpl P8 HSpl P l l HSp2 P10 HSp2P15 HSp2 P5 HSp2 P8 HSp2 P9 HSp2 P13 HSp2 P2 HSp2 P6 HSp2 P9 HSp2 P13 HSp2 P2 HSp2 P6 HSp2 P3 HSp2 P7 HSp3 P10 HSp3 P6 HSp3P12 HSp3 P3 HSp3Pll HSp3 P5 HSp3 P4 HSp3 P9 BSpl P9 BSpl PI BSpl P10 BSpl P4 BSpl P2 BSpl P5 BSpl P6 BSpl P l l BSpl P15 BSpl P12 BSp2P12 BSp2Pll BSp2Pl BSp2 P10 BSp2P13 BSp2 P5 Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta Chisholm, Alberta BSp2 P15 BSp2P14 BSp2 P7 BSp2 P4 BSp3 P9 BSp3 P10 BSp3 P6 BSp3 P4 BSp3 P3 BSp3 P5 BSp3 PI BSp3 P8 BSp3 P15 BSp3 P13 SSpl P2 SSpl PI SSpl P8 SSplP14 SSpl P9 SSpl P5 SSplP12 SSpl P6 SSpl P7 SSplP12 SSpl P6 SSpl P7 SSplP15 SSp2 P7 SSp2 P4 SSp2 P6 SSp2 P5 SSp2 P2 SSp2 P3 SSp2 P12 SSp2P15 SSp2 P9 SSp2P16 SSp3 P4 SSp3 P12 SSp3 P5 98 Chisholm, Alberta SSp3 P7 Chisholm, Alberta SSp3 P2 Chisholm, Alberta SSp3 PI6 Chisholm, Alberta SSp3 P15 Chisholm, Alberta SSp3 P14 Chisholm, Alberta SSp3 PI | Chisholm, Alberta SSp3 P9 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092163/manifest

Comment

Related Items