Pioneer plant communities five years after the 1988 Yellowstone fires by Robert J Ament A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biological Sciences Montana State University © Copyright by Robert J Ament (1995) Abstract: The Yellowstone fires of 1988 burned many different types of vegetation. This initiated secondary succession in environments from valley bottoms to alpine tundra. Five years after fire, plant communities were measured. Species presence was recorded in 100 m^2 macroplots and cover was sampled in twenty 1000 cm^2 quadrats. Pioneer community composition after severe fire in late-seral vegetation was compared across the elevational gradient in nine environmental types with three replications in each. In two of the subalpine fir environments, communities arising from four different pre-fire serai stages were sampled to test the hypothesis that pioneer community compostion differs when early-seral versus late-seral forests burn in one environmental type. Plant cover tends to decrease with increasing elevation. Along the elevational gradient, the wet grasslands had the strongest recovery from fire (plant cover averaged 97%), while the lowest cover was in the subalpine zone near treeline (39% average cover). Species richness was between 32 and 42 species per 0.01 hectare in the seven lowest environmental types. Diversity in the two highest elevational environmental types was distinctly low (19 and 20 species/0.01 hectare, respectively). Forty-two of the 262 species identified occurred in nearly all environments. Many of the others were concentrated in various portions of the gradient (i.e. grasslands, montane forests, subalpine fir forests). Each species and its distribution was tabulated. To test the hypothesis that pioneer communties were influenced by previous vegetation, ordinations (principal component analysis and principal coordinate analysis) were conducted on postfire communities representing four pre-fire serai stages. Neither method indicated communities arising from any pre-fire serai stages were distinct from any others. Chi-square goodness-of-fit to random distribution and Monte Carlo randomizations of individual species in these environmental types identified only three species that were significantly non-randomly distributed among postfire communities from pre-fire serai stages. All three were more strongly represented in pioneer communities from early prefire serai stages. Eighteen species in each environmental type possibly had non-random distributions (P=0.06 to 0.15) indicating they may deserve further study.  PIONEER PLANT COMMUNITIES FIVE YEARS AFTER THE 1988 YELLOWSTONE FIRES by Robert J. Ament A thesis submitted in partial fulfillment of the requirements, for the degree O1 . Master of Science in Biological Sciences MONTANA STATE UNIVERSITY Bozeman, MT May 1995 A/3'7? f̂iYX 2>4- ii APPROVAL of a thesis submitted by Robert John Ament This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. 1KS t Wr/vo/i_ Date Chairperson, Graduate Committee Approved for the Major Department Head, Major Department 7 7Date Graduate Dean iii STATEMENT.OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a master's degree at Montana State University, I agree that the Library shall make it available to borrowers under rules of the Library. If I have indicated my intention to copyright this thesis by including a copyright notice page, copyright is allowable only for scholarly purposes, consistent with "fair use" as prescribed by U.S. Copyright Law. Requests for permission for extended quotation from or reproduction of this thesis in whole or in parts may be granted only by the copyright holder. Signature iv ACKNOWLEDGEMENTS I would like to express gratitude to those who helped and guided me during this study. The encouragement, advice, and talks of all things ecological with my major professor, Dr. Tad Weaver, was appreciated.. Dr. Don Despain, adjunct professor, and employee of the newly christened National Biological Service, illuminated various aspects of fire ecology and other facets of the Greater Yellowstone Ecosystem for me. A special thank you to Dave Clark, with the support of his family, in helping me relocate most of the study sites in the Yellowstone area. Dr. Jack Rumely helped in verifying identifications of difficult taxa in the MSU herbarium. Dr. Mark Taper gave me needed support for . programming in MATHCAD and understanding the multivariate analyses. Dr. Dan Gustafson wrote a program to present the multivariate analyses in three dimensional figures. Few endeavors are as complete without the help of others and this was no exception. I thank them all. This study was funded by USDA-Forest Service, Intermountain Research Station Grant No. INT-93837-RJVA. I also was supported by a seasonal Yellowstone National Park research position. V TABLE OF CONTENTS Page LIST OF TABLES............ viii LIST OF FIGURES....... xii ABSTRACT................... xiii GENERAL INTRODUCTION...................................... I PART I Pioneer Communities Five Years After Fire in Nine Environmental Types Along the Elevational Gradient.3 INTRODUCTION----------- ----- . . ........... ......... 4 Yellowstone Fires...................... .Objectives................................ Environmental Classification.............. Pioneer Communities of Select Habitat Types....10 i METHODS............ 13 Study Area..................................... 13 Study Site Selection. ........ 16 Environmental Types.. A ........................ 20 Cover Types.................................. 26 Sampling Methods............................... 27 RESULTS.................................. 34 DISCUSSION.......................................... 49 Plants Found........ ................... *..... 49 Pioneer Communities of Major Environmental Types............................ 49 Species Distribution Along the Environmental Gradient......................... 51 Species Richness and Cover Along the Environmental Gradient......................... 56 CONCLUSIONS........... 62 LITERATURE CITED................. 64 sf VD TABLE OF CONTENTS— Continued PART II Effect of Pre-fire Serai Stage on Postfire Pioneer Communities in Subalpine Fir Environments.......69 INTRODUCTION.................. 70 Objective....... 70 LITERATURE REVIEW.............. 71 Community succession........................... 71 Secondary succession models.................... 79 METHODS............................................. 82 Study Area......................... 82 Study Site Selection................. 82 Environmental Types............................ 84 Serai Stages................................... 89 LPO............... 90 LPl................ 90 LP2.............................. 90 LP3......... ^........... ................ 91 Sampling Methods....................... 92 Methods of Analysis............................ 96 Ordination................................ 97 Releve Tables............................ 101 Chi-square Tests......................... 102 Monte Carlo Randomization ............... 103 RESULTS............................................ 106 Pioneer Communities in the Abies lasiocarpa/, Calamagrostis rubescens habitat type.......... 106 Pioneer community species richness....... 106 Ordinations of pioneer communities....... 106Re I eve. tables.............................H O Pioneer Communities in the Abies lasiocarpa/ Vaccinium scoparium Habitat Type.............. 119 Pioneer community species richness........119 Ordinations of pioneer communities....... 120 Releve tables............ 125 DISCUSSION......................... 130 Abies lasiocarpa/Calamagrostis rubescens Habitat Type................................ ..130 Postfire species richness...... 130 Ordination of postfire communities...... .132 vi TABLE OF CONTENTS— Continued Gradient analysis with Chi-square andrandomization tests..... 135 Abies lasiocarpa/Vaccinium scopariumHabitat Type.................................. 141 Postfire species richness...... 141 Ordination of postfire communities....... 142 Gradient analysis with Chi-square and randomization tests................ 144 CONCLUSIONS.... ...... 149 LITERATURE CITED......... 150 Appendix A--Plant Species Collected and Acronyms....158Appendix B— Plant Species Names and Authority...... 165 Appendix C— Summary Data of Species Abundance and Presence for each Habitat Type............ 172 Appendix D— Computer Programs for Principal Component Analysis, Principal Coordinate Analysis, and Monte Carlo Randomization....... 205Appendix E— Results of Monte Carlo Randomization Probabilities...... 212 ■ j vii viii LIST OF TABLES Table Page 1. Previous postfire pioneer community studies..... 11 2. Sample stand codes, legal descriptions.......... 17 3. Study site habitat types and cover types........ 28 4. Cover classes, ranges and midpoints............. 31 5. Pioneer communites of the environmental gradient.36 6. Pioneer species of single occurrence............ 41 7. Species richness along the environmental gradient........................... 45 8. Cover values along the environmental gradient.... 47 9. Tree seedling density in six habitat types....... 48 10. Sample stand location........................... 85 11. Sample site environmental types................. 92 12. Cover classes, ranges and midpoints............. 95 13. Species richness of pioneer communities 1 following fire in four serai stages of the Abies lasiocarpa/Calamagrostis rubescens h.t..;.107 14. Cumulative eigenvalues for principle component analysis and principle coordinate analysis of the Abies lasiocarpa/Calamagrostis rubescens . habitat type.................................. 109 15. Presence of pioneer species following fire in early (LPO) to late (LP3) vegetation in the Abies lasiocarpa/Calamagrostis rubescens h.t....114 16. Species present once in pioneer communities in the Abies lasiocarpa/Calamagrostis rubescens h.t. five years after fire...........117 ix LIST OF TABLES— Continued 17. Chi-square goodness-of-fit tests of < randomness for pioneer species groups afterfire in four serai stages in the Abies lasiocarpa/Calamagrostis rubescens h.t......... 119 18. Species richness of pioneer communities following fire in four serai stages of the Abies lasiocarpa/Vaccinium scoparium h.t....... 120 19. Cumulative eigenvalues for principle component analysis and principle coordinate analysis of the Abies lasiocarpa/Vaccinium scoparium h.t....122 20. Presence of pioneer species following fire in early (LPO) to late (LP3) vegetation in the Abies lasiocarpa/Vaccinium scoparium h.t.......126 21. Species present once in pioneer communities in the Abies lasiocarpa/Vaccinium scoparium habitat type five years after fire.............128 22. Chi-square goodness-of-fit tests of randomness for pioneer species groups five years after fire in four serai stages in the Abies lasiocarpa/Vaccinium scoparium habitat type....129 23. Vascular plant voucher specimens collected by sample stand and species abbreviations (acronyms) used in text and tables.............159 24. Vascular plant species list and authority......166 25. Species presence and cover in three stands in the Deschampsia cespitosa/Carex spp. h.t.......173 26. Species presence and cover in three stands in the Pseudotsuga menziesii/Calamagrostis rubescens habitat type.................... 175 27. Species presence and cover in three stands in the Pseudotsuga menziesii/Symphoricarpos albus habitat type..... 178 28. Species presence and cover in three stands in the Populus tremuloides/Calamagrostis rubescens habitat type................................... 180 X 29. Species presence and cover in three stands in the Art.emisia tridentata\Festuca idahoensis habitat type............................ ......^g2 30. Species presence and cover in three stands in the Abies lasiocarpa/Calamagrostis rubescens habitat type, LPO pre-fire cover type..........184 31. Species presence and cover in three stands in the Abies lasiocarpa/Calamagrostis rubescens habitat type, LPl pre-fire cover type..........186 32. Species presence and cover in three stands in the Abies lasiocarpa/Calamagrostis rubescens habitat type, LP2 pre-fire cover type..........188 33. Species presence and cover in three stands in the Abies lasiocarpa/Calamagrostis rubescens habitat type, LP3 pre-fire cover type..........190 34. Species presence and cover in three stands in the Festuca idahoensis\Agropyron caninum habitat type ....... ........................... . 35. Species presence and cover in three stands in the Abies lasiocarpa/Vaccinium scoparium habitat type, LPO pre-fire cover type..........194 36. Species presence and cover in three stands in the Abies lasiocarpa/Vaccinium scoparium habitat type, LPl pre-fire cover type..........196 37. Species presence and cover in three stands in the Abies lasiocarpa/Vaccinium scoparium habitat type, LP2 pre-fire cover type..........193 38. Species presence and cover in three stands in the Abies lasiocarpa/Vaccinium scoparium habitat type, LP3 pre-fire cover type..........199 39. Species presence and cover in three stands in the Abies Iasiocarpa-Pinus albicaulis/ Vaccinium scoparium habitat type, WB2 pre-fire cover type...................... .201 40. Species presence and cover in three stands in the Abies Iasiocarpa-Pinus albicaulis/ Vaccinium scoparium habitat type,WB3 pre-fire cover type LIST OF TABLES— Continued 203 LIST OF1 TABLES— Continued 41. Results of the Monte Carlo randomization probabilities for the ABLA/CARU habitat type....213 42. Results of the Monte Carlo randomization probabilities for the ABLA/VASC habitat type___215 xi xii LIST OF FIGURES Figure Page 1. The Greater Yellowstone Ecosystem................ 15 2. Sample stand location of the nine majorenvironmental types. ............. 19 3. Schematic of nine major environmental types......21 4. Species richness on the environmental gradient... 46 5. Location of stands sampled in the Greater Yellowstone Area in subalpine fir environments... 86 . 6. Species richness of pioneer communities following fire in four serai stages in the Abies lasiocarpa/Calamagrostis rubescens h.t....108 7. Three major axes of the principle component analysis (PCA) for postfire pioneer communities in the Abies lasiocarpa/Calamagrostis rubescenshabitat type.................................. Ill 8. Three major axes of the principle coordinate analysis (PCoA) for postfire pioneer communities in the Abies lasiocarpa/Calamagrostis rubescens habitat type..... 112 9. Species richness of pioneer communities following severe fire in four serai stages of the Abies lasiocarpa/Vaccinium scoparium habitat type......... ..121 10. Three major axes of the principle component analysis (PCA) for postfire pioneer communities in the Abies lasiocarpa/Vaccinium scoparium habitat type......... .123 11 11. Three major axes of the principle coordinate analysis (PCoA) for postfire pioneer communities in the Abies lasiocarpa/Vaccinium scoparium habitat type..................... 124 xiii ABSTRACT The Yellowstone fires of 1988 burned many different types of vegetation. This initiated secondary succession in environments from valley bottoms to alpine tundra. Five years after fire, plant communities were measured. Species presence was recorded, in 100 m2 macroplots and cover was sampled in twenty 10.00 cm2 quadrats. Pioneer community composition after severe fire in Iate-seraI vegetation was compared across the elevational gradient in nine environmental types with three replications in each. In two of the subalpine fir environments, communities arising from four different pre-fire serai stages were sampled to test the hypothesis that pioneer community compostion differs when early-seral versus late-seral forests burn in one environmental type.Plant.cover tends to decrease with increasing elevation. Along the elevational gradient, the wet grasslands had the strongest recovery from fire (plant cover averaged 97%), while the lowest cover was in.the subalpine zone near treeline (39% average, cover). Species richness was between 32 and 42. species per 0.01 hectare in the seven lowest environmental types. Diversity in the two highest elevational environmental types was distinctly low (19 and 20 species/0.01 hectare, respectively). Forty-two of the 262 species identified occurred in nearly all environments. Many, of the others were concentrated in various portions of the gradient (i.e. grasslands, montane forests, subalpine fir forests). Each species and its distribution was tabulated. To test the hypothesis that pioneer communties were influenced by previous vegetation, ordinations (principal component analysis and principal coordinate analysis) were conducted on postfire communities representing four pre-fire serai stages. Neither method indicated communities arising from any pre-fire serai stages were distinct from any others. Chi-square goodness-of-fit to random distribution and Monte Carlo randomizations of individual species in these environmental types identified only three species that were significantly non-randomly distributed among postfire communities from pre-fire serai stages. All three were more strongly represented in pioneer communities from early pre­ fire serai stages. Eighteen species.in each environmental type'possibly had non-random distributions (P=0.06 to 0.15) indicating they may deserve further study. I GENERAL INTRODUCTION The Yellowstone fires of 1988 burned all types of vegetation. Thus, secondary succession was initiated in numerous environments. Five years after the fires, in the summer of 1993, this study was conducted to describe early pioneer communities appearing after severe fire in climax vegetation of nine environmental types. Seed banks of most of the study sites had been examined by Clark (1991). These sites and nine more were visited’to acguire three replicate pioneer communities in each of nine major environmental types of the northern Rocky Mountains. The pioneer communities of the nine environmental types have never been described and compared. These environments represent relative positions along the elevational gradient so that direct gradient analyses of pioneer communities can be performed. This approach also allows examination of species distribution patterns on the elevational gradient. In comparing pioneer communities following severe fire in late-seral vegetation of different environments, it was also possible to test whether pioneer communities would differ if early-seral vegetation in these environments had burned. Thus, in Part II, in two environmental types, the effects of the pre-fire vegetation (different serai stages) 2 on postfire community composition appearing five years after severe fire was examined. This hypothesis was tested on both integrated communities (via ordination) and on individual species. The distribution of individual species in pioneer communities on a gradient of pre-fire serai stage was assessed with three different statistical tests. I y 3 PART I PIONEER COMMUNITIES FIVE YEARS:AFTER FIRE IN NINE ENVIRONMENTAL TYPES ALONG THE ELEVATIONAL GRADIENT 4 INTRODUCTION Yellowstone Fires. In the summer and fall of 1988 Yellowstone National Park and adjoining lands, both public and private, experienced the most extensive wildfires of the century. In three months, a series of fires burned 276,533 hectares (683,305 acres) of forests and grasslands (Rothermal and others 1994). The fires were started, by both lightning and human ignitions. The northern Rocky Mountains (NRM) had not experienced a combination of fires of this size and extent since 1910 when over 1,200,000 hectares (3,000,000 acres) of the forests in Montana and Idaho burned (Pyne 1982). The burning in Yellowstone was the most significant ecological event that a National Park has experienced (Scullery 1989). The fires also burned large parts of the surrounding National Forests [Bridger-Teton, Gallatin, Shoshone and Targhee], as well as lands under other ownership. The fires left a landscape comprising grasslands, shrublands, forests and alpine tundra effected to varying degrees by fire. The result was a mosaic of burn intensities and patch sizes reaching from valley bottoms to the alpine 5 ridges. The heterogeneity of the burn was superimposed over a pre-fire landscape that contained vegetation of different environmental zones each occupied by vegetation of different ages (Knight and Wallace 1989). The scope of the fires should not have been an unexpected occurrence in the temporal scale of centuries (Whitlock and others 1994). In the late 1600s and early 1700s several equally large, fire events had burned extensive areas in Yellowstone National Park's high volcanic plateaus (Romme 1982). Intervals between stand-replacing fires are often three hundred years in these subalpine fir forests. At lower elevations, where the shrub-steppe is interspersed with coniferous forest, eight to ten large fires burned previous to the twentieth century (Houston 1973). The average fire interval for the lower elevation northern range is twenty to twenty-five years. Thus, studies of the two different environments in YNP, subalpine forests and shrub- steppe, indicates that although they have different fire regimes, the Greater Yellowstone Ecosystem (GYE) has a history of extensive fires that predates settlement, fire management strategies and suppression. Fire size events are described logarithmically with time (Pyne 1982). That is, large events are infrequent, medium-sized ones more common, and smaller fires are much more numerous. Thus, while the 1988 fires' size was a rare ecological occurrence in recent time, burns of a smaller 6 dimension are a common component of the disturbance regime in plant communities of the Greater Yellowstone Ecosystem and the NRM. The many fires of 1988, were of varying intensities, in all types of plant communities on the environmental gradient, and established a landscape-level disturbance pattern typical of the northern Rocky Mountains. Objectives The Yellowstone fires of 1988 created an unusual opportunity to describe and compare postfire pioneer communities that established themselves concurrently in many different environments. Vegetation dominated by sagebrush, Douglas fir forests, aspen groves, mountain meadows, subalpine fir forests, and alpine tundra were burned. The objective was to locate stands (with three replications) burned at maturity (near-climax) in major- environments (nine habitat types) representing the altitudinal gradient. Resultant information will be useful in predicting early postfire succession after future fires on sites in these environments. ^ To make the study finite, sites were selected from those where the pre-fire vegetation was near-climax and all were severely burned. Since this study is to describe pioneer communities of major elevational zones an environmental classification system was used to characterize the study sites. The only 7 ecological classification widely employed in the NRM (Daubenmire 1952, Steele and others 1983, Mueggler and Stewart 1980) was used for this study. The units of this classification are environmental types (habitat types) and they are descriptors of the environment— not of the vegetation. This report will use the term — "environmental type" — synonymously with habitat type (HT). Environmental Classification The environmental type (habitat type) is the basic unit for classifying and identifying land potential (Daubenmire 1966). Thus, it can be used both to distinguish sites with different environmental conditions and to extrapolate observations from such sites to a larger geographic setting. It would be expected that the composition of a pioneer community appearing after stand-replacing fire, in any specified NRM habitat type, will approximate corresponding units of the environmental type observed in Yellowstone after the 1988 fires. This expectation is based on the assumption that similar environmental and historical constraints are operative on all units of a habitat type. In the United States, the habitat type system of environmental classification was developed in eastern Washington and northern Idaho (Daubenmire 1952). It was extended to the forests of Montana (Pfister and others 1977), the forests of eastern Idaho and western Wyoming 8 (Steele and others 1983) and the grasslands/shrublands of Montana (Mueggler and Stewart 1980). Most of the riparian areas of Montana have been added to this system (Hansen and others 1995). The vegetation of Yellowstone National Park has been described with this environmental classification (Despain 1990). While acknowledging some variability between sites this system of classification nonetheless regards all units of an environmental type to be comparable across the landscape.* Thus, units with a similar environment, regardless of the present successional status, are all categorized as the same habitat type. The environmental type oi a land unit is identified by recording the climax vegetation that occupies the site. If the vegetation is not at climax, then the vegetation that would occur at climax is deduced and recorded. On the burned sites of this study, potential was identified by examination of adjacent unburned vegetation, unburned material left within the sample stand, and vegetation maps of Yellowstone National Park (Despain and others, unpubI.) This classification of vegetation appears to conflict with the continuum theory (Curtis and Macintosh 1951, Gleason 1962, Whittaker 1960, Goodall 1963) of plant community change across the landscape. Observation that vegetation sometimes varies continuously over the landscape has prompted many to doubt that plant communities comprise 9 distinct, coevolved plant populations. Classification is useful for studies like this and for explanations of the results whether or not there is a belief in discontinuites in vegetation composition. This study uses the environmental type of classification with the knowledge that the scheme may divide a continuum into arbitrary, somewhat variable units. This choice was to maximize the extrapolation of our observations, in defense of the habitat type classification method, it was stated, "while this debate may be of interest academically, it need not preoccupy the natural resource managers and field biologists who need a logical, ecologically-based classification with which to work" (Pfister and others 1977). It is due to the efficacy of the habitat type that this study utilizes this classification ( system as the basis of describing early postfire pioneer communties in various environments in the Greater Yellowstone Ecosystem. In defense of discontinuities that would sometimes support discrete typal communities, Daubenmire demonstrated the use of his data could also support a continuum theory. He showed that the methodology and interpretation of continuum theory adherents are biased against observation of coevolved units (Daubenmire 1966). Further contrasts of continuum and classification theory are summarized by continuum adherents (Macintosh.1967, Whittaker 1967), a 10 proponent (Dansereau 1968), and an integrator (Allen and lHoekstra 1992). The latter note that viewpoints affect the conclusions drawn; while continuum advocates use the individual spepies as their point of departure, advocates of distinct units consider the ecosystem from the spatial scale of the landscape. Pioneer Communities of NRM Habitat Tvpe= Most vegetation classifications are based on undisturbed, mature plant communities. Early serai communities appearing in major environmental types (habitat types) have rarely been described. The few environments that ,have been investigated in the northern Rocky Mountains are pioneer communities of forested types in western Montana and northern Idaho. Most of these studies have been on lands disturbed by management activities either directly or on adjacent sites. A compilation by habitat type of postfire pioneer communities in coniferous forests that have been described in the NRM is listed in Table I. In another forest type, after a prescribed fire south of Yellowstone NP, on the Bridger-Teton National Forest, a three year study of aspen community response to moderate and high intensity burning was conducted (Bartos and Mueggler 1981). This was described before aspen communities had been classified (Mueggler 1988) for the intermountain western United States. 11 Table I. Northern Rocky Mountain coniferous forest postfire pioneer community studies by habitat type. Primary1Habitat type Investigator Abies lasiocarpa/Xerophyllum tenax H I l I l Il Il Il Il Il Abies lasiocarpa/Menziesia ferruginea Abies lasiocarpa/Linnaea borealis Pseudotsuga menziesii/Vaccinium globulare Pseudotsuga menziesii/Physocarpus malvaceus H it it it Pseudotsuga menziesii/Vaccinium globulare Tsuga heterophylla/Pachistima myrsinites Lyon 1976 Lyon 1984 Arno* 1985 Arno* 1985 Crane* 1983 Arno* 1985 Arno* 1985 Crane* 1983 Crane* 1983 Stickney 1986 1 Other authors contributed to studies marked with an asterisk(*) (Arno and others 1985, Crane and others 1983). There have been several reports of grassland and shrubland pioneer communities after fire; but, these studies did not use habitat types to characterize study site environments. An ungrazed grassland dominated by Rough fescue, Festuca scabrella was found to quickly return to pre-burn composition within three years in western Montana (Antes and others 1980). Three sagebrush studies west and south of the GYE were.in sagebrush, Artemisia tridentata, dominated stands (Harniss and Murray 1973, Humphrey 1984, and Akinsoji 1988). The long-term study in the Snake River plain (Harniss and Murray 1973) recorded vegetative transformation for thirty years. This study describes pioneer communities in nine different habitat types of grassland, shrubland, and forest X • 12 after severe fire. The habitat types considered do not overlap the environments of previous studies. All of the sites were located in a landscape unaffected by multiple-use management both before and after the fires. V 13 METHODS Study Area This study was conducted in Yellowstone National Park and should reasonably represent major environmental types of the Greater Yellowstone Ecosystem (GYE) and the northern Rocky Mountains. The GYE is a mountainous area surrounding Yellowstone. National Park (Figure I),. It straddles the continental divide and is mid-way between the equator and the north pole; the.45th degree northern latitude runs through it from east to west. It includes Yellowstone National Park, six adjacent National Forests, two National Wildlife Refuges as well as other public and private land holdings. Grand Teton National Park is included in the southern portion of the ecosystem. There is a complexity of land ownership and its changing environmental character make definition of exact boundries for the GYE difficult (Click and others 1991). The land base is approximately 17 million acres (7 million hectares) and includes portions of the states of Idaho, Montana,^ and Wyoming. The climate of the area is characterized by long, cold winters and short, cool summers. The temperatures are typical of a cool, dry continental climate and vary with elevation (Weaver 1980, Despain 1990). Most of the area receives between thirty and fifty inches of annual 14 precipitation depending on the elevation (Despain 1990). Lower elevations in the northwest corner of Yellowstone National Park receive only 10 to 12 inches (25 to 31 cm.). Higher elevations receive well over 70 inches (178 cm.) of precipitation annually (Glick and others 1991). The climates along the elevation gradient have been described for the northern Rocky Mountains for each vegetation type (Weaver 1980). Elevations in the region vary from 1300 meters in the river valleys to over 4,000 meters near the summits in several of the ranges. Despain (1990) describes the geology of the Yellowstone area. Five separate blocks of sedimentary and granitic formations were uplifted during the Laramide orogeny: the Beartooth, Targhee, Gallatin, Washakie, and Teton uplifts. During the Eocene, volcanic activity extruded andesite and basalt. Later, in the Pliocene, more faulting and uplift occurred. In the more recent Quartenary, a violent explosion generated a large caldera which contains the present Yellowstone Lake, destroyed portions of Mount Washburn and adjacent mountain ranges. It spread rhyolitic tuff and flows over most of the area. Glacial activity in the Pleistocene left kettle and kame topography, moraines, glacial till, and erratics in the valley bottoms. Soils vary and depend on parent materials, climate, and associated vegetation. Forest soils are commonly alfisols or inceptisols with shallow, rocky profiles (Trettin 1986). 15 Greater Yellowstone Greater Yellowstone Ecosystem ****** NATION AiALLATIN r'V NATIONAL YELLOWSTONE SHOSHONE IATIOI NATIONAL ONAL BRIDGER TETON WIND RIVERF ORE S T : NATIONAL FOREST Legend Figure I. The Greater Yellowstone Ecosystem (courtesy of the Greater Yellowstone Coalition). 16 Grasslands and shrublands occur primarily on mollisols (Munn and others 1978) in the NRM. These are derived mostly from alluvium and are typically well drained. In the higher valleys and mountain meadows mollisols commonly form on fine-grained alluvium and heavy soils derived from shales and andesitic volcanics (Despain 1990). Predominant parent materials of the soils in the Yellowstone area are andesite and rhyolite (Despain 1990). These two underlying bedrocks arose from separate volcanic events in the tertiary and guartenary, respectively. ; Key characteristics of soils; such as carbon/nitrogen ratios, nitrogen concentrations, pH, cation exchange capacity and phosphorous availability change in a predictable manner as one moves from dry, low elevation grasslands up through the forested zone to the alpine in the northern Rockies Mountains (Weaver 1979). Like the vegetational gradient, the soil gradient reflects changes in abiotic and biotic processes with differences in climate. Study Site Selection To characterize pioneer communities of each of the vegetational zones three sample stands were located in each of nine habitat types which were severely burned in 1988. The location of the sites sampled are listed in Table 2 and displayed in Figure 2. Each site's environmental type is listed in Table 3. 17 Table 2. Sample stand codes, names, legal descriptions (Universal Transverse Mercator - Zone 12), elevations and relative fertility. Stand ̂ Elevation Code___Site Name ________ UTM Coordinates fmeters I Soils1Al Crystal Bench 49730 N, 5532 E 1920A2 Lamar 49672 N, 5619 E 2067A3 Canyon West 49521 N, 5619 E 2621BI Bunsen 49765 N, 5243 E ■ 2164B2. Waterplant 49774 N, 5231 E 2042B3 Blackball 49743 N, 5416 E 2195Cl Terrace 49787 N, 5213 E 2269C2 Floating Island 49762 N , 5433 E 2044C3 Wraith 49761 N, 5299 E 2072Dl Bunsen 49746 N, 5244 E 2134D2 . Lamar 49673 N, 5657 E 2127D3 Waterplant West 49776 N, 5230 E 2059El Lamar 49689 N, 5618 E 2012E2 Frog Rock 49780 N, 5342 E 2114E3 Washburn 49660 N, 5450 E 2438Fl Indian Creek 49683 N, 5174 E 2316F2 Heart Junction 48966 N, 5385 E 2245F3 North Indian Creek 49688 N, 5172 E 2316Gl Willow Park North 49681 N, 5208 E 2246G2 North Roaring Mtn. 49598 N, 5208 E 2316G3 Tuff Creek 49441 N, 5132 E 2121Hl Golden Gate 49756 N, 5215 E 2251H2 Madison 49440 N, 5015 E . 2238H3 Grotto Geyser 49246 N, 5126 E 2246Il South Canyon 49495 N, 5401 E, 238012 Southeast Swan 49716 N, 5221 E 219913 Obsidian Creek 49653 N, 5206 E 2253Jl Washburn 49637 N, 5438 E 2626J2 Blackball 49764 N, 5395 E 2275J3 Specimen 49720 N, 5574 E 2074Kl Fan Creek 49781 N, 4967 E 2230K2 Cygnet Lakes 49469 N, 5315 E 2569K3 Canyon - Norris. 49521 N, 5372 E 2469LI Lulu Pass 49865 N, 5866 E 2487L2 Willow Park 49628 N, 5215 E 2254L3 Grizzly Trailhead 49609 N, 5207 E 2322Ml West Thumb 49192 N, 5335 E 2377M2 Lewis Lake 49039 N, 5306 E 2412M3 Lewis Canyon 48974 N, 5275 E 2359NI Little Thumb 49185 N, 5312 E 2503N2 Canyon - Norris 49522 N, 5374 E 2469N3 Virginia Falls 49507 N, 5282 E 241701 Washburn High 49618 N, 5443 E 280402 Observation 49570 N, 5363 E 269203 Washburn Low 49641 N, 5443 E 2534 18 Table 2— Continued Pl Observation East P2 Dunraven P3 West Dunraven 49569 N, 5370 E 49576 N, 5435 E 49600 N, 5427 E 2681 25602659 ' A AA Soils: Substrate of soils derived from: A - Andesite, basalt, sedimentary (fertile soils); R - Rhyolite, gneiss, granite (infertile, well-drained soils) The majority of the sites were originally located by Clark (1991) immediately after the 1988 fires for a seed bank investigation and were relocated with his help. The present study found stands used in that project by locating the metal stakes left,in 1988 with the aid of a metal detector. Sites were located near an unburned portion of the plant community so that verification of vegetative composition and therefore the habitat type and serai stage could be made. Of Clark's 45 transects, 39 were suitable for this study. Six of the original transects were unusable due to logging disturbance, mis-classification, or could not be relocated. Habitat type maps of Yellowstone National Park (Despain and others, unpubl.) were used to find new sites for replacement stands. These replacement stands are noted in Table 2. Three new sites were established to provide additional- replications for cover types originally under-sampled. The new sites were located with Despain and others' (unpubl.) habitat type map and verified by noting adjacent unburned vegetation. 19 Figure 2. Location of stands sampled in or near Yellowstone National Park. Letters indicate habitat type, numerals indicate replications. A-Deschampsia caespitosa/Carex spp. , B- Pseudotsuga menziesia/Symphoricarpos albus, C-Pseudotsuga menziesia/Calamagrostis rubescens, D-Populus tremuloides/ Calamagrostis rubescens, E-Artemisia tridentata/ Festuca idahoensis, F ,G ,H,I-Abies lasiocarpa/Calamagrostis rubescens, J-Festuca idahoensis/ Agropyroncaninum, K,L,M,N- Abies lasiocarpa/Vaccinium scoparium, 0,P-Abies lasiocarpa- Pinus albicaulis/Vaccinium scoparium. 20 All stands sampled for this study were within Yellowstone National Park's boundries except one, this was located on Gallatin National Forest lands north of Yellowstone National Park. Environmental Types The nine environments studied are defined by the habitat type classification. Grasslands and shrublands were described by Mueggler and Stewart (1980), forested sites were classified by Steele and others (1983). Aspen stands (Mueggler 1988) were from a classification but may not be climax communities. The nine environmental types' location in the landscape is shown in Figure 3. There are other habitat types present in the GYE, thus, the listed HTs are not necessarily continuous within a given landscape; they appear instead as a mosaic. Each is described. Artemisia tridentata/Festuca idahoensis (ARTR/FEPDI This moderately mesic shrubland HT predominates on slopes of less than 40 percent. Annual precipitation varies between 16 and 30 inches (Mueggler and Stewart 1980). The moister end of the type supports deeper soils and higher productivity. Bigleaf sagebrush/Idaho fescue HT has high plant and litter cover. Soils are fertile with a dark upper 21 ALPINE TUNDRA WHITEBARK PINE ABLA-Pl AUVASCFORESTS MOUNTAIN MEADOWS FEl DZAGCA LODGEPOLE PINE DOMINATED ABLA/VASC ABLA/CARUSUBALPINE FIR FORESTS POTA/CARU DOUGLAS FIR FORESTS PSME/SYAL DEC A/C AR EX GRASSLANDS / SHRUBLANDS AATR/FE1D Figure 3. Schematic representation of nine major environmental types along an elevational gradient in the Greater Yellowstone Ecosystem. 22 j horizon. Soils are slightly acidic to neutral. Fires are common in this environment. Festuca idahoens is /Agropyron. caninum (FE ID/AGCA1I The Idaho fescue/Bearded wheatgrass HT is characterized by a high forb cover. It is found on moderate to high elevation slopes. Thus, the growing season is shorter than many grasslands and evapotranspiration is lower (Mueggler and Stewart 1980). Soils sampled in this HT in Yellowstone National Park were cryoborolls and had high fertility and a dark surface horizon (Trettin 1986).. Deschamysia cespitosa/Carex spp. 7DECA/CAPEX') The tufted hairgrass/sedge habitat type occurs in grassy depressions or over perched water tables. No clear association, with one particular sedge species was evident in the stands sampled. Moisture is abundant and during portions of the growing season water stands on the soil surface. Gleying of soils indicates saturated soils with some oxidation. On adjacent sites, Idaho fescue grasslands were present. Therefore, due to extremely wet soils this environmental type represents a topoedaphic climax community under the polyclimax concept (Tansley 1935). In this type histosolic soils occur on very poorly drained alluvium (Trettin 1986). Fires only occur here late in the season 23 when the graminoids are cured and standing water has evaporated. Pseudotsucra menziesii/Symphoricarpos albus f'PSME/SYAL'l This habitat type occupies the lower elevations on slopes and benches under Pseudotsuga forests. Douglas- fir/snowberry inhabits moderately warm climates with moist soils. Cryumbrepts and cryoborolls are common to this HT (Trettin 1986). Soils have significant accumulations of organic matter and relatively high moisture retention. Arno (1980) found a 15-30 year fire free interval in this Douglas-fir series. Similarly, Houston (1973) calculated a fire frequency of 20 to 25 years in this vegetation. Pseudotsuqa menziesii/Calamagrostls rubescens fPSME/CARUI This is often the highest elevation of the Douglas-fir series and occurs on moderately dry mountain slopes. The dominant tree species is Douglas-fir both in serai and climax vegetation. Pinegrass, Calamagrostls rubescens decreases after overstory removal by fire or logging because the sites become too dry and warm for this species (Pfister and others 1977). Shrubs are usually sparse in this type. Soils on stands studied were derived from andesite (Table 2). This type has a similar fire regime as the PSME/SYAL habitat type. 24 Pppulus tremuloides/Calamaarostis rubescens fPOTR/CAPTTI The aspen/pinegrass community type occupies, relatively moist benches and slopes irrespective of steepness and aspect. It occurs over a wide range of elevations. The vegetation is simple both in structure and composition. Annuals are never abundant (Mueggler 1988). The extensive root system of. aspen usually creates clones and these below ground parts survive fire and resprout soon after. Because of rare occurrences of conifers this is often a climax community type. Abies lasiocarpa/Calamaarostis rubescens (ABLA/CARin This habitat type is usually located east of the continental divide in Montana but in the GYE it occurs on both sides. It is notably absent from the Wind River and Absaroka ranges (Steele and others 1983). It occupies a relatively warm environment with fertile soils. At climax, dense carpets of Calamagrotis leave little exposed soil. During extended dry periods the grassy understory can carry surface fires under lodgepole pine serai forests in this habitat type (Despain 1990). Soils sampled in the subalpine fir/pinegrass HT were cryoborolls with a very thick dark surface horizon, high water retention and high fertility (Trettin 1986). 25 Abies lasiocarpa/Vaccinium scoparium CABLA/VASO This is the most widespread forest habitat type in the GYE and the northern Rocky Mountains. Subalpine fir is the climax tree species but often this HT is dominated by lodgepole pine because fire commonly occurs before Abies lasiocarpa becomes dominant. Grouse whortleberry is a small shrub that dominates the understory and often forms a continuous cover. This HT occupies cool, relatively dry sites on mid- to upper mountain slopes and benches. Abies lasiocarpa - Pinus albicaulis/ Vaccinium scoparinm fABLA-PIAL/VASC^ This environmental type lies above the ABLA/VASC HT and often reaches to treeline. Whitebark pine and lodgepole pine are the serai tree species while subalpine fir becomes dominant with some Engelmann spruce at climax. Due to the relatively extreme climatic conditions this HT has low productivity and slow stand development. Soil water stress during the growing season is rarely present in this type (Weaver 1990). Whortleberry is the common shrub and other forb and graminoid species are sparser than in forests below. Soils are typically thin and poorly- developed in these cold forested climates. Stand-replacing fires are on long cycles of 200-300 years (Romme 1982). 26 Cover Types Within any of the nine environmental types different serai stages or cover types occur. After disturbance the original vegetation is replaced. Developing vegetation is often partitioned into a series of serai stages or cover types. Cover types for forested stands in Yellowstone National Park have been defined (Despain 1990). Despain7S (1990) cover types partition the four stages of forest development common to many forest ecosystems, including those recovering from fire (Peet 1992) into five types. The first is the establishment phase (LPO) and the second is the tree thinning stage (LPl). The third phase delineates the change in the overstory as subsequent understory trees enter the canopy containing the original postfire cohorts (LP2 and LP3), the mid- and Iate- successional. stages. The last stage (LP4), climax, is the steady-state, as the forest stand reaches a relatively stable composition (Oliver and Larsen 1990, Peet and Christiansen 1987). This pattern of development is found in many Rocky Mountain coniferous forests, aspen stands, boreal forests, as well as others (Peet 1992). The four serai stages sampled in the subalpine fir forests are abbreviated as LPO, LPl, LP2, and LP3 by reference to the serai dominant lodgepole pine. They are described in greater detail in Part II of this study. In subalpine fir forests, where whitebark pine codominates with 27 lodgepole pine in the serai stages, only one cover type was sufficiently widespread to provide three replications — whitebark pine two (WB2). WB2 is similar to LP2 except it occurs in subalpine fir stands near treeline. Part I of this study concentrates on pioneer communities arising after severe fire in mature communities of the nine habitat types. However, stands from pre-fire serai stages (immature stages) in the subalpine fir environment were sampled. Data from these stands have been included in several releve tables to compare species establishing in pioneer communities following serai stages with those establishing after mature types. A summary of each sample stand's code, geographic name, habitat type, and cover type when appropriate, is located in Table 2. This table also indicates which of the stands are sampled from Clark's (1991) original study and which have been replaced or added. Sampling Methods All sample sites, were visited between June to September of 1993. The metal stakes delineating each site were 25 meters apart. Stakes were installed similarly in comparable locations for new study sites. These stakes define the center line of a 25 m. by 4 m. macroplot. The macroplot extends two meters on either side of this line. The plot size examined in this study was within parameters suggested 28 Table 3. Study site stand codes, geographic names, habitat types and cover types. Stand Code Stand Name Habitat Type1 Pre-fire Cover Tvt>e2 Al Crystal Bench DECA/CAREXA2 Lamar DECA/CAREXA3 Canyon West DECA/CAREXBI Bunsen PSME/CARUB2 Waterplant PSME/CARUB3 Blacktail PSME/CARUCl Terrace PSME/SYALC2 Floating Island .PSME/SYALC3 Wraith PSME/SYALDl Bunsen POTR/CARUD2 Lamar POTR/CARUD3 Waterplant West POTR/CARUEl Lamar ARTR/FEIDE2 Frog Rock . ARTR/FEIDE3 Washburn ARTR/FEID 'Fl Indian Creek ABLA/CARU LPOF2 Heart Junction* ABLA/CARU LPOF3 North Indian Creek* ABLA/CARU LPOGl Willow Park North** ABLA/CARU LPlG2 North Roaring Mtn. ABLA/CARU LPlG3 Tuff Creek ABLA/CARU LPlHl Golden Gate ABLA/CARU LP 2H2 Madison ABLA/CARU LP 2H3 Grotto Geyser ABLA/CARU LP2Il South Canyon ABLA/CARU LP312 Southeast Swan ABLA/CARU LP 313 Obsidian Creek** ABLA/CARU LP 3Jl Washburn AGCA/FEIDJ2 Blackball AGCA/FEIDJ3 Specimen AGCA/FEIDKl Fan Creek ABLA/VASC LPOK2 Cygnet Lakes* ABLA/VASC LPOK3 Canyon - Norris ABLA/VASC LPOLI Lulu Pass .ABLA/VASC LPlL2 Willow Park ABLA/VASC ■ LPlL3 Grizzly Trailhead** ABLA/VASC LPlMl West Thumb ABLA/VASC LP 2M2 Lewis Lake ABLA/VASC LP 2M3 Lewis Canyon ABLA/VASC LP2'NI Little Thumb** ABLA/VASC LP 3N2 Canyon - Norris ABLA/VASC LP 3N3 Virginia Falls ABLA/VASC LP 301 Washburn High ABLA-PIAL/VASC WB 202 Observation ABLA-PIAL/VASC WB 203 Washburn Low** ABLA-PIAL/VASC WB 2 29 Table 3— Continued. Stand Code Stand Name Habitat Type1 Pre-fire Cover Tvoe2 Pl Observation East** ABLA-PIAL/VASC ■ WB 3P2 Dunraven High ABLA-PIAL/VASC WB 3P3 West Dunraven ABLA-PIAL/VASC WB 3 * New Transects ** Transects replaced due to logging,etc. 1 Forested habitat types (Steele and others. 1983), grassland/shrubland habitat types (Mueggler and Stewart 1980), and the aspen community type (Mueggler 1988) are: Artemisia tridentata/Festucaidahoensis (ARTR/FEID), Festuca idahoensis/ Agropyroncaninum (ARTR/AGCA), Deschampsia caespitosa/ Carex spp.(DECA/CAREX), Pseudotsuga menziesia/ Symphoricarpos albus (PSME/SYAL), Pseudotsuga menziesia/ Calamagrostisrubescens (PSME/CARU), Populus tremuloides/ Calamagrostis rubescens (PSME/CARU), Abies lasiocarpa/ Calamagrostis rubescens (ABLA/CARU), Abies lasiocarpa/ Vaccinium scoparium (ABLA/VASC), Abies Iasiocarpa-Pinus albicaulis/Vaccinium scoparium (ABLA-PIAL/VASC). 2 Cover types are described by Despain (1990). (MueIler-Dumbois and Ellenberg 1974) for examination of temperate forest undergrowth (50-200 meters squared) and dry grasslands (50-100 meters squared). Each stand was visited once during the growing season. The species list, therefore, is a minimal estimate of species present in postfire pioneer communities. Spring ephemerals were potentially overlooked at sites visited later in the season. Late season flowering plants/ cover abundance was most likely underestimated at sites visited early in the growing season. These effects were minimized because low elevations were sampled first and higher elevation communities were visited later in the season. 30 The data was collected in one growing season, thus, controlling many variables. Completion of sampling in one season had two advantages. First, it removed any analyses of variability between growing seasons due to different seasonal factors affecting plant cover (annual precipitation, length of growing season, etc.). Second, there is no compositional change as a function of time to consider. Thus, time as a factor does not Vary. Two measures of the plant community were made in each 0.01 hectare macroplot. First, floristic composition was measured by species presence in the macroplot. Species presence, although a crude measure for comparison of stands (Greig-Smith 1983), habitat types, or serai stages, is useful in determining in which environments each species is at least minimally successful. Therefore, this measure will aid in determining species success across the environmental gradient. Species presence is simply the identification to the species level of all vascular plants within the macroplot. Second, plant cover was recorded as a measure of species performance. Cover was measured as the vertical projection of the crown area of a species to the ground as a percent of the reference area. In this study, the reference area was a 20 cm by 50 cm quadrat or 0.1 meter squared. Cover is a better measure of plant biomass than density and thus more ecologically significant (Daubenmire 1968). 31 Cover was estimated by sub-sampling the macroplot with twenty quadrats. The quadrats or microplots were placed alternately along either side of the 25 m tape measure stretched between the two terminal metal stakes. The quadrats were located one meter apart on alternate sides of the tape starting at meter three and ending at meter twenty- two. Within each of the twenty quadrats in each sample stand, species cover was recorded using a cover scale. Each species present in a quadrat was assigned a cover class (Table 4) as defined by a range of percent cover (Daubenmire 1959) and refined by Bailey and Poulton (1968). This methodology is easily duplicated and the cover classes are used to minimize sampling error (Daubenmire 1959). Table 4. Cover classes, cover class ranges and midpoints (Bailey.and Poulton 1968). COVER CLASS RANGE OF COVER CLASS MIDPOINTS _________________ (%)______ m 7 95-100 97.56 75-95 85 5 50-75 62.5 4 25-50 37.5 3 5-25 15 2 1-5 3 I 0-1 0.5 Besides vascular plant cover, several other cover categories were recorded. These included bare soil, litter, dead woody residue greater than three inches in diameter 32 (7.5 cm), mosses and hornworts. Woody residue greater than three inches in diameter is referred to as coarse woody debris by fire ecologists and has many important physical, chemical and biological functions (Graham and others 1994). The mean cover for each vascular plant species and ground cover type was calculated by summing the appropriate class mid-points across all quadrats, dividing by twenty, and multiplying by one hundred. Total percent cover for each site was calculated by summing all mean percent cover for individual species and ground cover types. The few taxonomic difficulties were resolved 'as follows. Several sample sites had species that were in a phenological stage such that only the genus could be determined. Non-vascular plants were classified to phylum for bryophyta (mosses) or to genus for the lone identifiable hornwort - Marchantia spp. Two lupines were difficult to separate when not in flower so they have been combined. They were Lupinus argenteus and L. sericeus. For convenience, on forested sites L. argenteus was used and for shrublands/ grasslands L. sericeus was used. Unknown specimens were taken to the Yellowstone National Park Herbarium each evening during the field season to positively identify them with herbarium specimens. Thus, a cumulative knowledge of rare or difficult taxa was acquired during the field studies. Voucher specimens were pressed and dried. A list of plant species found and the sampling site location from which it was Collected is given in Appendix A. These specimens reside at the Montana State University Herbarium (MONT). Final determinations of difficult taxa were completed at the Montana State University Herbarium and were verified by Dr. Jack Rumely, associate curator. Nomenclature of plant species usually follows Hitchcock and Cronquist (1973). Species identified in this study but not included in their treatment of the Pacific Northwest flora follow Dorn (1992). Also, species that Hitchcock and Cronquist present as varieties but now are considered species follow born (1992). Each species identified in the study is listed in Appendix B with its authority. 34 RESULTS One objective is to list and quantify the species present in pioneer communities of nine common environmental types of the northern Rocky Mountains; Table 5 and Table 6 present this information. The headers of Table 5 and Table 6 list environmental types from grasslands/shrublands (E,J,A) through montane forests (C,B,D) to subalpine fir forests (I, N, P). These columns present pioneer species establishing after fire in mature pre-fire communities. The last three columns (F7K7O) present pioneer communities of pre-climax (serai stage) stands burned in three subalpine fir environments (F- ABLA/CARU, K-ABLA/VASC, O-ABLA-PIAL/VASC). The left column in Table 5 lists species present in more than one sample site. The left column in Table 6 lists species present only once in the entire study. Information appended to the species list in Table 5 includes origin (an asterisk indicates exotic species), life-form (graminoids-G, fOrbs-F7 Shrubs-S7 or trees-T) and duration (annual-A7 biennial-B, or perennial-P). Table 5 is broken into sections containing species with particular distributions on the elevational gradient. That is, species that occur in a single environmental type (Table 35 5, Sections C, E, and G) those that occur in a few similar types (Table 5, Sections D and F) or those that occur everywhere in pioneer communities (Table 5, Sections A and B)• There were several with bimodal distributions (Table 5, Section H). Values that occur in the body of the table indicate the constancy of each species in each habitat type. A zero indicates uniform absence, one is one occurrence, two is two occurrences, and three shows that the species was always present in pioneer communities studied in the environmental type. , Data from the pre-fire serai stages is discussed in more detail in Part II of this report. However, since serai stands from the subalpine forest zone share species with sites from the grassland/shrubland and montane environments these data are summarized in the last three columns of Table 5 and Table 6. 36 Table 5. ̂Pioneer communities appearing five years after . severe fire in climax vegetation of nine environmental types along the elevational gradient of the Greater Yellowstone Ecosystem and in several serai stages of subalpine fir environments. Pre-fire Communities Mature______ Seral1 _______Habitat Tvoes2____Species3 Life-form4 E J A C B D I N P F K O Section A: Species commonly present in all environments Potentilla gracilis PF Phleum pratense* PG Campanula rotundifolia PF Penstemon procerus PF Calamagrostis rubescens PG Cirsium scariosum PF Bromus carinatus PG Collomia linearis AF Fragaria virginiana PF Carex praticola PG Agropyron caninum PG Agoseris glauca PF Polygonum douglasii AF Epilobium paniculatum PF Arabis glabra* BF Antennaria microphylla PF Taraxacum officinale* PF Achillea millefolium PF Poa nervosa PG Stipa nelsonii PG Astragalus miser PF Tragopogon dubius* BF Lupinus argenteus/sericeus PF Senecio sefra PF Epilobium angustifolium PF Collinsia parvifIora AF 2 I I 2 . 2 I . d 2 2 3 2 3 2 I a a 2 I I I 2 2 f a 2 I I 2 . 2 b I . I I I 3 3 3 f II I 3 I I I C aI I I 2 I 3 I f II 2 2 2 3 2 3' I f II • 2 3 2 3 3 2 f C 2 I I I . 2 I 23 3 2 3 2 2 2 2 f 2 2 3 2 2 3 2 2 I f b 3I I 2 . I 2 I 2 f I • I I 2 I 2 I I f CI I I 3 3 . I f f I 2 3 I I I • 3 I I f C 2 2 2 3 3 3 3 3 2 I f f 23 3 3 3 3 3 3 2 2 f f 3 2 I • 3 3 2 3 I I f C 33 2 . 2 3 I 2 I f e II 2 . I 2 3 2 f II I I I 2 . I I C d3 2 « , 2 3 2 2 f f 2I I • 3 3 3 I 2 a C • • I 3 3 2 3 3 3 f f 3 « . I 2 I I 2 I 2 f C 2 Section B: Species of weak distribution in all environments Festuca idahoensis Arabis drummondii Androsace septentrionalis Viola adunea Potentilla glandulosa Geranium viscosissimum Aster foliaceus PG 3 3. PF 1 1 . A/BF I 2 . PF . I ,3 PF I . . PF I . . PF I . . 1 . . 2 I . 1 I . I . I . . 1 1 I I . I 2 2 3 I I . . I 2 e I I b a I 2 . I . a . » . e . . . a a . I 3 . 1 37 Table 5— Continued. _______Habitat Types2____Species3 Life-form4 E J A C B D I N P F K O Agoseris aurantiaca PF Aster ascendens PF Perideridia gairdneri PF Epilobium glaberrimum PF Descurainia richardsonii AF Arabis holboellii PF Cerastium arvense PF Gayophytum diffusum AF Carex hoodii PG I . . . I . I . . . 1 1 I . . . 1 3 I . . . I . . I . 2 . .. 2 . . 2 . . 3 . 1 1 . . I . . . I . . I . 2 1 • • • G C o 1 ...... . . . l a .2 2 3 a f . . . I b I 3. . . f c I I . . b a I . I . f 2 .. . . c . I Section C: Species primarily of grasslands/shrublands Koelaria cristata PG Oxytropis deflexa PF Artemesia cana S Artemesia frigida s Chrysothamnus viscidiflorus S Erigeron gracilis PF Eriogonum umbellatum PF Linum lewisii PF Anemone.spp. PF Saxifraga rhomboidea PF Senecio streptanthifolius PF Stellaria longipes PF Aster occidentalis PF Poa. palustris PG Carex lanuginosa PG Juncus balticus PG Salix geyeriana S Eriophyllum lanatum PF Rumex acetosella* PF Potentilla diversifolia PF Danthonia spicata PG 3 2 .. I I . I I . I I . I I . 1 2 . 2 3 . 1 2 . . 2 . . I I2 . I I . I 1 1 2 . . 3 . . 2 . . 2 . . 2 I . . I . . . I I 1 1 1 e e 2 a . . I Section D: Species primarily of grasslands/shrublands and montane forests Stipa richardsonii PG I IPhlox hoodii PG 2 IAgropyron spicatum PG 2 2Geum triflorum PF IAllium cernuum PF IErythronium grandiflorum PF IErigeron speciosus PF I IPotentilla arguta PFEpilobium ciliatum PF I . I . I . I . I . I . I 2 1 . 2 2 I 3 3 38 Table 5— Continued. Species3 _______Habitat Tvoes2____Life-form4 E J A C B D I N P F K O Artemisia tridentata Thlaspi arvense* Thalictrum occidentals Bromus ciliatus Chenopodium fremontii Trifolium hybridum* Erysimum inconspicuum Symphoricarpos albus Galium boreale Myosotis sylvatica Solidago missouriensis Poa juncifolia Aster hesperius Lithospermum ruderale Aster campestris Astragalus agrestis Trifolium longipes Bromus marginatus S 3 . I 2 I I . dAF I 3 I I I . bPF . I I I .. IPG I 2 2AF • I 2 I . bPF 2 I 2 I . B/PF I . • I I I .S I • I I 2PF 2 I 3 2 2 .PF I I # I I .PF I I . 2 I .PG 3 3 I I I .PF 2 I .PF I I .PF I 2PF I 2 I .PF I I I .PG I I . Section E: Species primarily of montane and serai subalpine fir forests Oryzopsis exigua Lepidium virginicum Geranium bicknellii Viola canadensis Zigadenus venonosus Symphoricarpos oreophilus Corydalis aurea Balsamorhiza sagittata Lactuca seriola Hackelia floribunda Elymus cinereus Melilotus officinalis* Berberis repens Populus tremuloides Pseudotsuga menziesia Geranium richardsonii Linanthus septentrionalis Cynoglossum officinale* Iliamna rivularis Rosa acicularis Aster conspicuus Spiraea betulifolia Cirsium vulgare* Lomatium dissectum PG I « b b .A/PF I • . IAF I IPF I IPF 2S 2A/BF . I aPF • I IAF # 2 IB/PF • I IPG . 2 IAF 2 I aS I 2 aT . I I f .S I IPF I IAF I I aBF I I IS I 2 IS 2 2 2PF 2 2 I bS I 2 I a a .BF I I I b c .PF • • I • • I 39 Table 5— Continued. _______Habitat Types2____Species3 Life-form4 E J A C B D I N P F K O Solidago spp. PF • • ■ • I • • • • b • . Section F: Species common in all forest environments Phacelia franklinii A/BF I I . . IOsmorhiza depauperata PF I I . I . aAster integrifolius PF . I I . . I dFrasera speciosa PF I I . . IGentiana amarella PF I I . . ISolidago multiradiata PF . 2 1 1 . f a IElymus glaucus PG 2 I I I 2 . f a 2Carex geyeri PG I I . 2 . f bFragaria vesca PF I I . I . bPoa interior PG I I I . . IHieracium albiflorum PF I . 2 2 d CTrisetum spicatum PG I 3 3 3 f f 3Nemophila brevifIora AF 2 . . . IGayophytum racemosum AF I I I . e IArnica cordifolia PF 2 2 3 3 f C 3Cirsium arvense* PF I I I . f bCarex rossii PG I 3 3 3 f f 3Capsella bursa-pastoris* AF I I . . bDfaba stenoloba AF • • I I l l • • • Section G: Species common in subalpine fir forests Phleum alpinum PG I . . aCerastium fontanum B/PF I . . eGnaphalium vicosum A/BF . 2 .Aquilegia flavescens PF . I . IGnaphalium spp. A/BF I . . aAntennaria racemosa PF I I . aPinus contorta T I 2 . f fSitanion hystrix PG I . I f C IHieracium gracile PF I . I a IVaccinium scoparium S 1 3 3 d f 2Sedum lanceolatum PF . . I aVaccinium globulare S . . I ePoa scabrella PG 1 . 2 • • 3 Section H : Species of bimodal distribution Carex raynoldsii PG I I . .Senecio canus PF I . I . .Viola nuttallii PF 3 . . 2 . . f ISilene oregana PF b 40 Table 5— Continued. _______Habitat Types2____Species3 Life-form4 E J A C B D I N P F K O Arenaria congesta Danthonia intermedia Draba nemorpsa* Barbarea orthoceras Senecio sphaerocephalus Deschampsia cespitosa Calamagrostis canadensis Phlox longifolia Agrostis scabra PF . I . PG . I . AF . 2 . PF . . I PF . . IPG . . 3 PG . . 2 PF 1 1 . PG 1 . 1 2 2 I I . . I 1 1 I . . I l l f f 2 Section I: Species primarily of serai stages Phacelia hastata Melica bulbosa Aster meritus Arnica parryi Hieracium cynoglossoides Polemonium pulcherrimum - PF PG PF PF PF PF c . . d . . a a . . a I . e I . . 2 1 Pre-fire serai subalpine fir forests: F-ABLA/CARU, LPO cover type K-ABLA/VASC, LPO cover type (a denotes species presence in LPl, b-LP2, c-LPO and LPl, d-LPO and LP2, e-LPl and LP2, f-all three serai stages), O-ABLA-PIAL/VASC, WB2 cover type. 2 Habitat types listed in order of increasing elevation. Grasslands: E-FEID/ARTR, J-FEID/AGCA, A-DECA/CAREX. Douglas fir forests: C-PSME/SYAL, B-PSME/CARU, D-POTR/CARU. Subalpine fir forests: I-ABLA/CARU, N-ABLA/VASC, P-ABLA- PIAL/VASC. 3 Values denote number of occurrences in three sample ' stands of 0.01 hectare. Asterisk (*) represents exotic species. 4 Life-form and duration: A-annual, B-biennial, P- perennial, G-graminoids, F-Forbs, S-Shrubs, and T-trees. 41 Table 6. Pioneer species with only a single occurrence appearing five years after severe fire in climax vegetation of nine environmental types along the elevational gradient of the Greater Yellowstone Ecosystem and in several serai stages of subalpine fir environments. Pre-fire Communities _______Mature______ Seral1 _______ Habitat Types2______Species E J A C B D I N P F K O Astragalus kentrophyta I . . Atriplex nuttallii I . .. Carex vallicola I . . Chaenactis douglasii ' I . . Lewisia rediviva I . . Lappula redowskii I . . Orthocarpus luteus I . . Astragalus adsurgens . i . Astragalus purshii . i . Astragalus vexilliflexis . I . Besseya wyomingensis . I . CameLina microcarpa . I . Castilleja pallescens . I . Delphinium nuttallianum . I . Erysimum cheiranthoides . I . Polygonum bistortoides . I . Lomatium cous . 1 . Senecio integerrimus . I . Taraxacum laevigatum . I . Carex spp. . I . Poa cusickii . i . Stipa lemmonii . I . Potentilla fruticosa . . I Salix wolfii . . i Arnica longifolia . . I Arnica mollis . . I Equisetum arvense . . i Erigeron acris . . I Galium trifidum . . i Gentiana detonsa . . i Mentha arvensis var. canadensis . . I Senecio cymbalarioides . . I Solidago canadensis . . I Veronica serpyllifolia . . I Carex microptera . .. I Carex rostrata . . i Carex sartwellii . . I Luzula campestris . . I Muhlenbergia richardsonis . . I Table 6— Continued. 42 Species ______ Habitat Types2_____ E J A C B D I N P F K O Poa trivialis Linnaea borealis Rubus idaeus Conimitella williamsii Erigeron compositus Galium bifolium Hackelia deflexa Osmorhiza chilensis Phlox multiflora Valeriana dioica Ceanothus velutinus Chrysothamnus nauseosus Sheperdia canadensis Aster pereglans Castilleja miniata Hackelia micrantha Linaria dalmatica Silene latifolia Dactylis glomerata Astragalus eucosmus Allium brevistylum Anemone parvifIora Arenaria Iaterifolia Corallorhiza trifida Senecio pseudaureus Trifolium pratense Trifolium repens Verbascum thapsus Rorippa islandica Poa fendleriana Poa pratensis Abies lasiocarpa Picea engelmannii Ribes lacustre Vaccinium membranaceum Muhlenbergia racemosa Arabis lemonii Arabis lyallii Astragalus alpinus Salix bebbiana Anaphalis margaritacea Deschampsia elongata Poa compressa Stipa occidentalis Arctostyphalos uva-ursi Ribes cereum Microsteris gracilis I ........... I . .I . . . . I . . I . .I . . I . .I . . I . . I . . . . . I .. I . . I . . I . . I . . I . . I . . I . . I . . . I . . I . . I . . I . . . . I . . . . I . . . . I . . . . I . . . . I . . . I . . . . . I .. . . I . . . . . I . . . . 1 . . . . 1 . . . . I . . . . 1 .......... I .......... I .......... I I I a a a a a a 43 Table 6— Continued. Habitat Types2Species E J A C B D I N P F K O Hordeum brachyantherum bCirsium spp. bCrepis atrabarba bSenecio crassulus ISpergularia rubra IArnica latifolia aEpilobium anagallidifolium aPenstemon fruticosus aMarchantia spp. bLupinus lepidus IPhacelia sericea IRumex paucifolius ISibbaldia procumbens IVeronica wormskjoldii IBromus inermis , • • I Serai pre-fire subalpine fir forests: F-ABLA/CARU, LPO cover type; K-ABLA/VASC, LPO cover type; O-ABLA-PIAL/VASC, WB2 cover type. Letters denote serai cover types other than LPO: a-LPl, b-LP2 (Despain 1990). Habitat types are from grasslands/shrublands: E- FEID/ARTR, J-FEID/AGCA, A-DECA/CAREX, Douglas fir forests: C-PSME/SYAL, B-PSME/CARD, D-POTR/CARU, subalpine fir forests: I-ABLA/CARU, N-ABLA/VASC, P-ABLA-PIAL/VASC. Presence is in 0.01 hectare plots (n=3). The mean species richness and its standard deviation of pioneer communities for each of the nine environments was calculated (Table 7). These communities are derived from climax pre-fire vegetation. Because only three sites were sampled for each environmental type, the standard deviation is often greater than the mean. A graphic display of mean species richness of the three ppstfire sample stands in each habitat type of mature pre-fire communities is found in Figure 4. The data presented is from Table 7. 44 The cover comparison of pioneer communities after severe fire in mature communities was condensed by summing cover of species in vegetational classes (i.e. trees, shrubs, herbs, graminoids) and ground cover types (i.e. coarse woody debris, litter, etc.). Each value is the mean for the three stands of each habitat type. The cover has been partitioned among life-forms and types of ground cover. The data is listed in Table 8. Natural regeneration of tree species, seedling density, in the six forested environmental types after severe fire in mature communities was compared (Table 9). All habitat types were sampled three times. Included is data for subalpine fir habitat types that were burned in immature serai stages. Each serai stage (cover type) was sampled three times. Thus, the total sample size for the ABLA/VASC and ABLA/CARU HTs was twelve. The total sample size for ABLA-PIAL/VASC was six. 45 Table 7. Vascular plant species richness of pioneer communities five years after severe fire in mature communities of nine environmental types in the Greater Yellowstone Ecosystem. Habitat types are arranged in order of increasing altitude. Habitat tvne2 Life Form1 TotalTrees Shrubs Forbs Graminoids ARTR/FEID 0.0(0.0) 2.0(1.0) 26.7(12.9) 9.7(1.2) 38.0(11.8)FEID/AGCA 0.0(0.0) I.3(2.3) 25.3( 1.5) 8.0(2.0) 34.7( 0.6)DECA/CARE 0.0(0.0) 2.0(1.0) 20.0( 3.8) 10.7(2.1) 32.7( 4.9)PSME/SYAL 0.3(0.6) 3.0(1.7) 26.3( 8.5) 7.7(2.5) 37.3(11.8)PSME/CARU 0.3(0.6) 4.7(3.5) 28.3( 6.1) 9.0(1.7) 42.3( 7.5)POTR/CARU 0.6(1.2) 2.7(2.3) 26.3( 8.5) 6.0(3.0) 37.0( 9.5)ABLA/CARU 0.3(0.6) 0.3(0.6) 23.0( 2.0) 10.7(4.0) 34.3( 5.0)ABLA/VASC I.3(1.5) I.7(1.2) 11.7( 2.1) 5.7(0.6) 20.3( 3.2)ABLA-PIAL 0.0(0.0) I.3(0.6) 12.7( 2.1) 5.0(1.0) 19.0( 1.0) 1 Mean number of species per 0.01 hectare by habitat type (N=3). The standard deviation of the three sample stands is in parentheses. 2 Habitat types are abbreviations for Artemisia tridentata/ Festuca idahoensis (ARTR/FEID), Festuca idahoensis/ Agropyron caninum (ARTR/AGCA), Deschampsia caespitosa/ Carex species (DECA/CAREX), Pseudotsuga menziesia/ Symphoricarpos albus (PSME/SYAL), Pseudotsuga menziesia/ Calamagrostis rubeseens (PSME/CARU), Populus tremuloides/ Calamagrostis rubeseens (PSME/CARU), Abies lasiocarpa/ Calamagrostis rubeseens (ABLA/CARU), Abies lasiocarpa/ Vaccinium scoparium (ABLA/VASC), Abies Iasiocarpa-Pinus albicaulis/Vaccinium scoparium (ABLA-PIAL). 46 Figure 4. Vascular plant species richness five years after severe fire in mature communities of nine common environments of the Greater Yellowstone Ecosystem. Mean number of species present by life-form and for all species in 0.01 hectare plots (n=3) for each type. Habitat types are arranged in order of increasing elevation: Artemisia tridentata/Festuca idahoensis (ARTR/FEID), Festuca idahoensis/Agropyron caninum (ARTR/AGCA), Deschampsia cespitosa/Carex spp.(DECA/CAREX), Pseudotsuga menziesia/ Symphoricarpos albus (PSME/SYAL), Pseudotsuga menziesia/ Calamagrostis rubescens (PSME/CARU), Populus tremuloides/ Calamagrostis rubescens (PSME/CARU), Abies lasiocarpa/ Calamagrostis rubescens (ABLA/CARU), Abies lasiocarpa/ Vaccinium scoparium (ABLA/VASC), Abies Iasiocarpa-Pinus albicaulis/Vaccinium scoparium (ABLA-PIAL/VASC). 47 Table 8. Mean percent cover and standard deviation (sd) of. life-forms and ground cover in pioneer communities of nine environmental types in the Greater Yellowstone Ecosystem five years after severe fire in mature communities. E J A Habitat type1 C B D I N PCategory1 2 Trees sd 0.00.0 0.0 0.0 0.0 0.0 0.00.0 O'. 0 0.0 0.30.5 0.30.5 0.10.1 0.0 0.0 Shrubs sd 1.2 1.1 2.1 3.6 2.6 2.7 10.2 10.7 6.2 2.1 6.3 6.9 0.0 0.0 7.1 7.5 2.5 3.6 Forbs sd 44.8 18.9 35.3 14.1 20.9 8.5 55.329.0 41.4 21.7 35.4 15.3 34.7 6.6 17.0 4.4 32.2 7.3 Grass sd 43.9 13.0 47.5 19.2 73.6 10.6 9.2 6.3 35.9 3.9 32.0 13.7 30.2 17.4 17.1 11.4 4.6 4.8 Mosses sd 0.0 0.0 0.0 0.0 0.1 0.1 2.2 2.2 5.1 5.9 2.0 2.1 0.7 1.1 10.5 9.3 5.1 . 1.0 Litter. sd 2.7 1.2 2.9 2.4 1.5 1.6 5.1 7.9 7.6 4.9 9.6 10.3 io.3 2.6 30.6 12.8 4.6 3.6 C.W.D. sd 0.0 0.0 0.0 0.0 0.0 0.0 0.8 1.4 2.2 3.9 3.2 3.7 1.0 0.4 2.3 2.0 1.8 0.2 Bare sd 5.5 4.7 11.5 4.8 2.6 3.1 7.0 11.4 4.6 4.1 7.7 8.3 19.2 27.7 16.4 20.3 46.9 2.6 Total sd 87.4 12.5 99.6 3.8 97.0 4.2 89.7 10.1 102.0 19.1 96.7 12.4 96.3 16.7 100.9 5.4 97.4 2.3 1 Habitat types are arranged in order of increasing elevation: E-Artemisia tridentata/Festuca idahoensis, J- Festuca idahoensis/Agropyron caninum, A-Deschampsia cespitosa/Carex species, C-Pseudotsuga menziesia/ Symphoricarpos albus, B-Pseudotsuga menziesia/Calamagrostis rubescens, D-Populus tremuloides/Calamagrostis rubescens, I- Abies lasiocarpa/Calamagrostis rubescens, H-Abies lasiocarpa/Vaecinium scoparium, P-Abies Iasiocarpa-Pinus albicaulis/Vaccinium scoparium. 2 First line of each category is mean percent cover. Grasses include all graminoids, Mosses include hornworts, C.W.D.-coarse woody debris greater than 3 inches in diameter. Bare-bare ground, Total-total mean percent cover of all categories. Second line (sd) is standard deviation. 48 Table 9. Seedling1 density five years after severe fire in mature or serai stage (cover type) forests of six environments (habitat types) in the Greater Yellowstone Ecosystem. Habitat Tvoe3 Cover Tvoe4 Soecies2 Total Substr. Ratio5LPP DF SAF AS ES PSME/SYAL 0.67 0.33 1.00 2:1PSME/CARU 1.67 0.33 2.00 3:0POTR/CARU 0.33 30.33 30.66 3:0ABLA/CARU LPO 1.67 1.67 0:3ABLA/CARU LPl 24.67 24.67 0:3ABLA/CARU LP 2 2.67 2.67 3:0ABLA/CARU LP 3 5.67 5.67 3:0ABLA/CARU All 8.67 8.67 6:6ABLA/VASC LPO 17.33 0.33 17.67 1:2ABLA/VASC LPl 108.33 0.33 108.67 1:2ABLA/VASC LP 2 93.00 0.33 93.33 1:2ABLA/VASC LP 3 3.67 0.33 1.00 5.00 0:3ABLA/VASC All 55.58 O DO Ul 0.08 55.91 3:9ABLA-PIAL WB 2 1.00 1.00 3:0abl a-pial WB 3 0.00 3:0ABLA-PIAL All 0.50 0.50 . 6:0 1 Mean number of seedling's per 100 meters squared (n=3).For all sites of ABLA/CARU and ABLA/VASC, n=12. For all ABLA-PIAL sites , n=6 .. 2 LPP-Iodgepole pine, Pinus contorta, DF-Douglas fir, Pseudotsuga menziesiif AS-aspen, Populus tremuloides, SAF- subalpine fir, Abies lasiocarpa, and ES-Englemann spruce, Picea engelmannii. 3 Habitat abbreviations (Steele and others 1983, Mueggler 1988): Pseudotsuga menziesia/ Symphoricarpos albus (PSME/SYAL), Pseudotsuga menziesia/ Calamagrostis rubescens (PSME/CARU), Populus tremuloides/ Calamagrostis rubescens (PSME/CARU), Abies lasiocarpa/ Calamagrostis rubescens (ABLA/CARU), Abies lasiocarpa/ Vaccinium scoparium (ABLA/VASC), Abies Iasiocarpa-Pinus albicaulis/Vaccinium scoparium (ABLA-PIAL). 4 Pre-fire cover types in subalpine forests increase from early seral/ to mature stand development: LPO, LPl, LP2, LP3. Whitebark pine forests are mid- (WB2) to late (WB3) serai stages (Despain 1990). 5 Ratio of sample stands on soils derived from relatively fertile andesitic (A) or relatively infertile rhyolitic (R) substrate. 49 DISCUSSION Plants Found. There were 262 vascular plant species identified in the 48 stands sampled. These included five tree, twenty-five shrub, 179 herb, and 53 graminoid species. Only one plant species new to Yellowstone National Park was found. This was Silene.Iatifolia Poir., a perennial catchfly, an exotic herb from Europe. While the Montana State University herbarium did not have a collecion from Montana, it has been collected in various parts of Wyoming (Dorn 1992). Two species common in first year postfire communities, probably arising from the seed bank, were rarely found five years after the fires. They were Geranium bicknellii and Dracocephalum parviflorum (Stickney 1990, Anderson and Romme 1993). Of these postfire ephemerals, G. bicknellii was found in only two Douglas fir sample stands and D. parviflorum was not found in any of the communities sampled. \ Pioneer Communities of Manor Environmental Types Fires will continue to occur in all nine environmental types studied. To describe severe fires' initial effects, pioneer communities have been described with a list of colonizing plants and the coverage of each (Tables 5 and 6). 50 The information has also been compiled to characterize and compare pioneer communities, as a whole, by habitat type, across the elevational gradient (Tables 7 and 8). The underlying individual sample stand data is summarized in Appendix C. Plants expected in any environmental type are identified by reading down the appropriate column in Table 5. Those represented by a three occurred in 3 of the 3 sample sites and have a high probability of being found. Those represented by a two have a 66.6% probability of occurrence and those represented by a one have an average probability of 33% of being located in pioneer communities of the environmental types considered. These probabilites have a high variance because the sample size was minimal. Non-ecologists often fear that severely burned sites will revegetate slowly. Table 8 lists the mean percent cover of life-forms and ground cover types of the nine major environmental types. Bare ground cover, five years after fire, was lowest (3%) in the Deschampsia cespitosa/ Carex spp. HT and highest (47%) in the Abies Iasiocarpa-Pinus albicaulis/Vaccinium scoparium habitat type. No other HT exceeded 20% bare ground (Table 8), thus a reasonable amount of vegetative cover appeared in all environments except forests near treeline five years after fire. 51 Species Distribution Along the Environmental Gradient Forty-two species appeared in most pioneer communities following severe fire across the entire environmental gradient (Table. 5, Sections A and B). These "universal" species are 25% of the 166 species occurring more than once in the sample stands. These species are also present in pioneer communities of forest stands that burned in early to mid-seral stages. Species of this distribution were divided into those that were common (Table 5, Section A) and those that were missing in a few habitat types (Table 5, Section B). If more samples in each habitat type had been taken, the species of weak distribution would probably be found to actually occupy all the environmental types. Of the forty- three species occurring across the environmental gradient only four are exotics. A variety of adaptations exist for species to be successful pioneers. No single strategy defines those that were successful across the elevational gradient. Some are present vegetatively in mature communties before severe fire and others are not (Pfister and others 1977, Steele and others 1983). Examples of universal species present in the mature communities of these habitat types that resprout after fire are: Antennaria.microphylla, Fragaria virginiana, Astragalus miser, Lupinus spp., and Achillea millefolium (Table 5). Others are perennials with excellent seed dispersal capabilities, especially by wind: Epilobium } angustifolium, Agoseris glauck, A. aurantiaca, and Taraxacum officinale. Thus, these may colonize from sprouts, the seed bank, or off—site dispersal. Still others are annuals that 1 are rarely present in climax communities, such as Collinsia parviflora, Collomia linearis, Epilobium paniculatum, Descurainia richardsonii, and Polygonum douglassii. These may colonize from the seed bank or by disperal. As expected, there are many species with a much narrower ecological amplitude. Twenty-one species occurred in only the shrubland/grassland environments (Table 5, Section C). Constancies of these species were similar to those of rarer universal species (Table 5, Section B) and lower than those of more common universal species (Table 5, Section A). In apparent contrast with climax dominants (grasses), only six grassland/shrubland specialists are graminoids and most of these (four) ire from the extremely wet DECA/CAREX habitat type. Most of the grassland pioneers are perennial herbs and shrubs from genera such as Artimesia, Chrysothamnus,'Erigeroh, and Eriogonum. A third group of.species pioneered in both grassland and montane forest environments but were absent in subalpine fir environments (Table 5, Section D). Two-thirds of these species are perennial forbs and only one is a shrub. Several species are common pioneers in both grasslands and montane forests, examples are Galium boreale, Poa juncifolia, and Solidago missouriensis. Others species such as Trifolium 52 53 hybridum, Agropyron spicatum, and Phlox hoodii are common in pioneer grasslands but appear rarely in montane forests. These species may be less shade tolerant than those occupying both environments and therefore may have left a poorer seed bank in the forested environments. In this study, there were no examples of species with the opposite distribution, i.e. species commonly pioneering montane forests and uncommon in the grassland/shrubland environments. Some species appear in postfire communities of mature stands in mid- and low elevational environments (Table 5, Section D) or mid-elevational montane forests (Table 5, Section E) and high elevational subalpine fir environments burned in early serai stages. They are absent- in the postfire communities of mature subalpine fir forests. It could be speculated that such species colonize from on-site sources, vegetative sprouting and seed banks, but that they are absent from the more mature subalpine fir forests. Nine of the 25 species listed in Table 5-, Section E, are annuals or biennials. These species rarely occur in climax communities of subalpine fir or montane environments (Steele and others 1983). Five of the species are shrubs common in climax forests, for example, Spirea betulifolia and Rosa acicularis. These can resprout after fire. Several species appeared in both low grasslands and high subalpine fir sites after early serai communities burned. The 54 perennials that can resprout, Danthonia spicata and Potentilla diversifolia are examples. It could be anticipated that the weedy exotics, Cirsium vulgare and Cynoglossum officinale, would be found in other environments if the sample size were larger.. Nineteen species pioneered after fire in all six forest environments (Table 5, Section F). There were three annuals, Nemophila breviflora, Dfaba stenoloba, and Gayophytum racemosum. Most are perennials. All were forbs and graminoids, none were shrubs or trees. Some were more common in subalpine fir. forests than montane forests. Four examples are Arnica cordifolia, Trisetum spicatum, Hieracium albiflorum, and Carex rossii. These all have capabilities to establish either vegetatively or to colonize via the seed bank and seed rain. In contrast, Poa interior, a perennial grass, is an example of a species more common in the montane than in the subalpine fir forest environments. A few species, thirteen, occurred primarily in the subalpine fir environments regardless of pre-fire stand development (Table 5,.Section G). Grouse whortleberry, Vaccinium scoparium and lodgepole pine, Pinus contorta, are examples. Species occurred occasionally in the many (30) postfire stands sampled in these three environments, but never appeared in lower elevational sites. Thus, there is sufficient evidence that these are pioneers restricted to higher elevations. 55 Thirteen species were found to have a bimodal distribution (Table 5, Section H). These appear after fire in both moist grasslands and subalpine fir environments but not on sites occupied by mid-eIevationaI Douglas fir and aspen forests. It is not clear what excludes them from the montane environments. All are perennials, either forbs or graminoids. Most are present in Climax communities of these environments (Steele and others 1983). Thus, all may survive fire and reproduce vegetatively. Some species with bimodal patterns may be artifacts of the small sample size; a larger sample would aid in detecting such species. Six species occurred in several sample stands, but only in pioneer communities derived from serai subalpine fir forest stands (Table 5, Section I). These species also occur in climax subalpine fir environments (Steele and others 1983). Thus, there is little evidence that they require disturbance to maintain their presence in subalpine fir forest environments. One hundred and one species appeared in only one pioneer community (Table 6). The bigleaf sagebrush/Idaho fescue HT had only five singletons. The other grasslands had many species occurring only once. The Idaho fescue/bearded wheatgrass HT had fifteen and the tufted hairgrass/sedge HT had eighteeen. Pioneer communities derived from the subalpine fir forests had very few species of single occurrence; ABLA/CARU and ABLA-PIAL/VASC had three 56 singletons each. The lack of species occurring only once at high elevations parallels overall species richness (Table 7) and may indicate stress due to environment. Species Richness and Cover Along the Environmental Gradient Species richness of postfire pioneer communities is compared among environmental types along the elevational gradient (Table 7 and Figure 4). Similarly, the mean percent cover by life-form and ground cover type is given in Table 8. Six of the environmental types are forests at maturity before fire. Regeneration of tree species along the gradient is included in the discussion (Table 9). Only postfire communities derived from mature pre-fire vegetation in each of the nine environmental types is characterized, except for tree seedling density. Very few tree species were present in any of the nine environmental types, partially because few exist in the arboreal flora of the environments studied. No trees occurred in the three grassland/shrubland postfire sample stands. In montane environments, the Douglas fir habitat types had low seedling densities, no greater than 200 seedlings per hectare. These pioneer communities contained both lodgepole pine and Douglas fir cohorts. The aspen forests had over ten times the seedling density of adjacent Douglas fir environments (3,000 seedlings per hectare), most seedlings were aspen resprouts. This low tree seedling 57 density corresponds to cover and species richness values for the montane forests. The highest diversity in tree species was in the ABLA/VASC HT with a mean of 1.33 tree species/0.01 hectare (Table 7). The most common seedling was the lodgepole pine (Table 9). Values of less than one for tree species richness indicate some of the forested sites had no tree seedlings in the sample site macroplots. Trees, are not expected to dominate in the forest environments so soon after severe fire. There was little cover contributed by trees in the pioneer communities (Table 8). Although tree species diversity and cover was low, some sites in subalpine fir environments had moderate seedling densities. Tree seedling density (Table 9) was highest in the ABLA/VASC HT, 5600 seedlings per hectare, most were lodgepole pine. Very few seedlings appeared in the ABLA- PIAL/VASC HT; there were 500 seedlings per hectare. These harsher timberline environments will be slower to colonize. In a study of lodgepole pine forests two years after the 1988 fires, seedling density varied from 900 to 191,000 seedlings per hectare (Ellis 1993). Thus, the subalpine fir- forests from this study are at the lower end of seedling density variability found by Ellis. For pioneer shrubs, the richest environments were the montane forests (Table 7). The Douglas fir/pinegrass HT had a mean of 5 shrub species/0.01 hectare and 6% shrub cover. 58 Shrub cover was highest, 10.2 %, in the PSME/SYAL HT and second highest, 7.1% in the ABLA/VASC HT (Table 8). These two environmental types each have an understory dominated by shrubs at climax. In the grassIand/shrubland types, the moist tufted hairgrass/sedge HT had the same amount of shrub species as the bigleaf sagebrush/Idaho fescue shrubland. Thus, the physiognomy of climax communities does not necessarily provide a greater species richness within that life-form in pioneer communities after fire. The moist DECA/CAREX HT may be a more diverse environment than the ARTR/FEID HT which, before fire, is dominated by one species of sagebrush. The DECA/CAREX HT also had more shrub cover in the pioneer communities than the ARTR/FEID habitat type. Graminoid richness was equally high in the grasslands and the grassy forests, that is, forest environments dominated by pinegrass in the understory in late-seral stages (Table 7). This data suggests both the PSME/CARU and ABLA/CARU HTs are simply moist grasslands with a forest canopy. However, the grassland/shrubland environments(had higher graminoid cover than any of the grassy forest environments (Table 8). The aspen stands, although described as having pinegrass as the dominant understory component in mature pre-fire communities, had less graminoid richness in pioneer communities than the coniferous forest pioneer 59 communities, except the two high, elevation subalpine fir HTs. Postfire graminoid richness is relatively constant across the entire environmental gradient, from a low of.5.0 graminoids/O.01 hectare in the ABLA-PIAL/VASC HT to a high of 10.67 graminoids per 0.01 hectare in both the ABLA/CARU and DECA/CAREX habitat types (Table 7). The graminoid cover differs across the elevational gradient, with a low of 4.6% in the subalpine fir-whitebark pine/whorttleberry HT and a high of 73.6% in the DECA/CAREX habitat type (Table 8). Species in combination with their cover, rather than richness of graminoids, defines the differences in climax physiognomy between forests and grasslands. Forbs dominate species richness in all environmental types (Table 7). Forbs accounted for over half of the species in pioneer communities in all environments except near treeline in the subalpine fir-whitebark pine environments. ARTR/FEID and FEID/AGCA grasslands had at least six more species per sample stand than the wet tufted hairgrass/sedge habitat type. This higher richness of forbs in the drier grasslands accounted for the higher herbaceous cover. The richest part of the environmental gradient for forb diversity in pioneer communities appeared in the mid­ elevation Douglas fir/ pinegrass and aspen/pinegrass habitat types. Forb cover was highest, 55 percent, in the slightly lower elevational Douglas fir/snowberry habitat type (Table 60 8). At these sites over half of the pioneer vegetation was dominated by forbs. Thus, forb species richness and cover are only loosely related. Differences in diversity become most evident when combining all life-forms into total species richness (Table /7). The diverse pioneer communities of the Douglas fir/pinegrass HT, with 42 species per 0.01 hectare, were over twice as rich as pioneer communities of subalpihe fir- whitebark pine/whortleberry HT that had. a mean of nineteen species/0.01 hectare. The high elevation subalpine fir HTs, ABLA/VASC and ABLA-PIAL/VASC. were clearly the least species rich of the nine environmental types. The lowest seven elevationally - grasslands, montane forests, and the subalpine fir/pinegrass 'HT were relatively similar in total species richness, varying between means of 32 and 42 species per 0.01 hectare. The total mean percent cover of vascular plants in pioneer communities of each of the nine environmental types was highest, 97 percent, in the wet grassland, DECA/CAREX habitat type (Table 8). Recovery from fire is rapid in this type, primarily from vegetative regeneration. The lowest vascular plant cover, 39 percent, was in the pioneer communities of the subalpine fir-whitebark pine/whortleberry habitat type. The low vascular plant cover parallels the lack of species richness in this environmental type. This cool, wet environment with a short growing season (Weaver 61 1990) recovers slowly from disturbance. The three montane forest environments have total mean vascular plant cover, ranging between 74% and 84%, indicating a relationship between high coverage and species richness. While pioneer communities contain many species, this richness can be expected to decrease as stands mature and competition for resources becomes more intense. Many of the annuals and biennials will become rare or disappear as studies at "climax" indicate (Pfister and others 1977, Mueggler and Stewart 1980, Steele and others 1983). The grassland environments have already recovered much of their pre-fire cover, as expected (Antos 1983). The sagebrush/ shrubland pioneer communities will take several more decades to reach climax (Harniss and Murray 1973). Cover of the trees will increase in the forests as the sites develop. CONCLUSIONS. Pioneer communities following fire in nine major environments of the Greater Yellowstone Ecosystem have been described. Forty-two species occurred in most of the nine environmental types along the elevational gradient. This was 25% of the 166 species identified more than once in the study. These species can be expected to pioneer after future fires in the northern Rocky Mountains regardless of the location of the burned area's position along the environmental gradient. There were various patterns of distribution for each of the other 122 species, they all have been described. These species were restricted tb pioneer communities in assorted portions of the environmental gradient after fire. The distribution patterns could not be summarized by methods of colonization after fire. Each pioneer species had varying strategies in various combinations: vegetative sprouting, canopy seed rain, off-site seed disperal, and the seed bank (Stickney 1990). Five years after severe fire, pioneer communities across the elevational gradient differed in recovery response and rates. The wet grasslands recovered quickly? their total vascular plant cover is 97 percent. In contrast, 63 communities in subalpine fir;environments near treeline have vascular plant cover averaging 39 percent. The six habitat types in the grassland and montane forest environments had an average plant cover between 74 and 97 percent. In the three subalpine fir forest environments, at higher elevations, plant cover averaged less than 66% of a site. Thus, there was a trend for cover to decrease with increasing elevation. ( Species richness of pioneer communities five years after severe fire in climax communities along the elevational gradient was highest in the montane forest zone. The highest diversity, a mean of 42 species per 0.01 hectare, was in the Douglas fir/pinegrass habitat type. Mean species richness, from the lowest grasslands/shrublands up to the lowest subalpine fir zone, varied between 32 to 42 species per 0.01 hectare, indicating that the position along the elevational gradient did not exert a great influence on pioneer community richness. The two highest subalpine fir environmental types were distinctively lower in richness, means were 19 and 20 species per 0.01 hectare. These harsher environments have less species capable of establishing after fire. ( 64 Literature Cited Akinsoji, A. 1988. Postfire vegetation dynamics in a sagebrush steppe in southeastern Idaho, USA. Vegetatio: 78: 151-155. Allen, T. F. H. and Hoekstra, T . W. 1992. Toward a unified ecology. Columbia University Press, New York, NY. Antes, J. A., McCune, B. and Bara, C. 1983. The effect of fire on an ungrazed western Montana grassland. American Midland Naturalist 110(2): 354-364. Arno, S. F. 1980. Fire history of the northern Rockies. Journal of Forestry 78(8): 460-465. Arno, S. F., Simmerman, D. G. and Keane, R. E. 1985. Forest succession on four habitat types in western Montana. USDA Forest Service, General Technical Report INT-177. Intermountain Research Station, Ogden, UT. Bailey, A. W. and Poulton, C. E. 1968. Plant communities and environmental relationships in a portion of the Tillamook burn, northwestern Oregon. Ecology 49: 1-13. Bartos, D. L. and Mueggler, W. F. 1981. Early succession in aspen communities following fire in western Wyoming. Journal of Range Management 34(4): 315-317. Clark, D. L. 1991. The effect of fire on Yellowstone ecosystem seed banks. M. S. Thesis, Montana State University, Bozeman, MT. 115 p. Crane, M. F., Habeck,.J. R. and Fischer, W. C. 1983. Early postfire revegetation in a western Montana Douglas-fir forest. USDA Forest Service, Research Paper INT-319. Intermountain Research Station, Ogden, UT. Curtis, J. T. and Macintosh, R. P. 1951. An upland forest continuum in the forest-prairie region of Wisconsin. Ecology 32: 476-496. Dansereau, P. 1968. The continuum concept of vegetation: responses. Botanical Review 34: 253-332. 65 Daubenmire, R. Fi 1952. Forest vegetation of northern Idaho and adjacent Washington, and its bearing on concepts of vegetation classification. Ecological Monographs 22: 301-330. Daubenmire, R. F. 1959. Canopy coverage method of vegetation analysis. Northwest Science 33: 43-64. Daubenmire, R. F. 1966. Vegetation: identification of typal communities. Science 151: 291-298. Daubenmire, R. F. 1968. Plant communities: a textbook of plant synecology. Harper and Row, New York, NY. Dorn, R. D. 1992. Vascular plants of Wyoming. Mountain West Publishing,. Cheyenne, WY. Despain, D.G. 1990. Yellowstone veget