Romme and Walsh: A Comparison of Fire Regimes and Stand Dynamics in Whitebark Pine

A COMPARISON OF FIRE REGIMES AND STAND D YNAMICS IN WHITEBARK PINE (PINUS ALBICAULIS) COMMUNITIES IN THE GREATER YELLOWSTONE ECOSYSTEM + WILLIAM H. ROMME + JAMES R. WALSH DEPARTMENT OF FOREST RANGELAND AND WATERSHED STEWARDSHIP COLORADO STATE UNIVERSITY+ FORT COLLINS

+ INTRODUCTION severity fire regimes have been documented in whitebark pine forests in the Shoshone National Whitebark pine (Pinus albicaulis) is a forest NW of Cody, WY (Morgan and Bunting keystone species of upper subalpine ecosystems 1990), and in NE Yellowstone National Park (Barrett (Tomback et al. 2001), and is especially important in 1994). In Western Montana and Idaho, mixed fire the high-elevation ecosystems of the northern Rocky regimes have been documented in whitebark pine Mountains (Arno and Hoff 1989). Its seeds are an communities in the Bob Marshall Wilderness (Keane essential food source for the endangered grizzly bear et al. 1994), Selway-Bitterroot Wilderness (Brown et (Ursus arctos horribilis), particularly in the autumn, al. 1994), and the West Bighole Range (Murray et al. prior to winter denning (Mattson and Jonkel 1990, 1998). Mattson and Reinhart (1990) found a stand­ Mattson and Reinhart 1990, Mattson et al. 1992). In replacing fire regime on the Mount Washburn the Greater Yellowstone Ecosystem (GYE), Massif, within Yellowstone National Park. biologists have concluded that the fate of grizzlies is intrinsically linked to the health of the whitebark pine It has been proposed that, as a result of 60+ communities found in and around Yellowstone years of fire exclusion, natural fire cycles in seral National Park (YNP) (Mattson and Merrill 2002). whitebark pine communities have been postponed Over the past century, however, whitebark pine has (Arno 1986, Keane 2001, Keane and Arno 2001), and severely declined throughout much of its range as a that both the frequency and spatial extent of modem result of an introduced fungus, white pine blister rust fires are outside of their historic range of variability. ( Cronartium ribicola) (Hoff and Hagle 1 990, Smith As a result, these authors argue that structure in these and Hoffman 2000, McDonald and Hoff 2001 ), stands has been altered, and forest health reduced. native pine beetle (Dendroctonus ponderosae) Restoration efforts, including the use of fire, have infestations (Bartos and Gibson 1990, Kendall and recently been undertaken to attempt to counteract Keane 2001 ), and, perhaps in some locations, whitebark pine decline in western Montana and Idaho successional replacement related to fire exclusion and (Keane and Amo 1996, 2001). While the modem era fire suppression (Arno 2001). of fire suppression may have altered the fire regimes in some whitebark pine communities, it is unlikely The most common historical whitebark pine that all communities throughout its range have fire regimes are "stand-replacement", and "mixed­ declined primarily as a result of fire suppression severity" regimes (Morgan et al. 1994, Arno 2000, activities. Arno (2001, p.83), has postulated that in Arno and Allison-Bunnell 2002). In the GYE, mixed- some areas- particularly were there is an extensive,

Published by Wyoming Scholars Repository, 2003 1 University of Wyoming National Park Service Research Center Annual Report, Vol. 27 [2003], Art. 17

124 continuous subalpine forest, as in Yellowstone • Beartooth Lake, Beartooth Mountains, National Park- large, stand-replacing burns were Shoshone N.F ., WY probably characteristic, corresponding to a stand­ • Avalanche Peak, Absaroka Range, replacement fire regime. Previous research has Yellowstone National Park, WY documented long fire intervals in subalpine forest • Wind River Lake, Absaroka Range, types in YNP (Romme and Despain 1989) and these Shoshone N.F.,WY have historically exhibited stand-replacing crown fue behavior (Romme 1982). In areas such as this, In the summer of 2004, five additional sites will be modern fire suppression will have had little influence sampled: on stand dynamics or fire intervals, and restoration • Atlantic Peak, Wind River Range, would be unnecessary. Shoshone N.F., WY • Union Pass, Wind River Range, Shoshone Our study examines the physical structure N.F.,WY (basal area, density, species composition) and stand­ • Teton Pass, Teton Range, Bridger-Teton age structure of whitebark pine communities N.F./Caribou-Targhee N.F., WY/ID throughout the Greater Yellowstone Ecosystem • Sawtell Peak, Centennial Mountains, (GYE). In addition, this study assesses the historic Caribou-Targhee N.F., ID fire regimes in these whitebark pine communities and • Windy Pass, Gallatin Range, Gallatin N.F., evaluates the influence of surrounding forest types on MT fire severity and extent. These objectives are being accomplished by comparing stand structure, age structure, and fire history in several whitebark pine + METHODS forest communities throughout the Greater Yellowstone Area. Although investigators have The following field methods were used found both stand-replacing and mixed severity fire during the 2003 field season, and will again be regimes in other parts of whitebark pine's range, little employed for sampling during the summer of 2004. data currently exists to quantify the historic fire Each study site was sampled to determine stand regime of white bark pine in the GYE. structure, age structure, and fire history using several techniques. The point-centered-quarter (PCQ) In the summer of 2003, we initiated a study method (Cottam and Curtis 1956, Mueller-Dombois of whitebark pine in the GYE in order to develop a and Ellenberg 197 4) was used to collect stand better understanding of the historic variability of fire structure data. Stand-origin and age-class distribution events and stand dynamics in these communities. data were collected from three tree size-classes The data collected during that period continues to be (canopy, understory, and seedling/sapling) using the analyzed and will be supplemented with additional points established on the PCQ transects. The fire field data collected during the summer of 2004. Our history for each study area is being developed using study addresses two major questions: i) What were fire-scar analysis techniques, based on the collection the historic fire regimes of whitebark pine of fire scar records from both living and dead trees and communities in the Greater Yellowstone Ecosystem?; logs. Communities were sampled either "intensively" and ii) Does within-stand structure and adjacent stand or "extensively". structure and composition influence fire dynamics in Intensive sampling involves: whitebark pine stands? 1) Establishing two 100 m transects, 50 meters apart, through a stand of interest (a 1 ha sampling area) + STUDY AREAS 2) Taking Point-Centered-Quarter measurements every 10 meters to determine physical stand Research sites for this study are located on structure along each transect the high, forested, subalpine mountains in the Greater 3) At each of the points (established at 1Om­ Yellowstone Ecosystem These sites are dominated by intervals along each transect), coring the whitebark pine, although subalpine fir (Abies nearest tree of each species in each of three lasiocarpa), Englemann spruce (Picea engelmannii), size classes to determine age structure. and lodgepole pine (Pinus contorta) are co-dominants on 4) Searching for, and opportunistically many sites. The climate is cold with moist springs sampling, fire scars within a 500 meter and relatively dry summers. A total of eight study radius of the plot center (located between sites will be sampled. Thus far, three stands were the 50 meter marks of the two transects), sampled in the summer of 2003:

2 Romme and Walsh: A Comparison of Fire Regimes and Stand Dynamics in Whitebark Pine

125

thus searching and sampling for fire over an + RESULTS AND DISCUSSION area of .8 square km. Mixed severity fire regimes consist of Extensive sampling was conducted using the same individual fires that vary between nonlethal techniques. However, in extensive sampling sites, the understory burning and stand-replacement burns, age structure protocol only involves coring the both spatially and temporally. If whitebark pine nearest whitebark pine individuals in each size class communities in the GYE experienced a combination at each point to examine age structure, rather than of nonlethal understory burns along with stand­ coring each species present in each size class. replacing fires through time, then evidence of non­ lethal understory bums of small, patchy extents During the 2003 field season, the Wind should be evident in fire-scar records within River Lake site was sampled intensively. During the whitebark pine stands, at least in some locations. Fire summer of 2004, both the Atlantic Peak and Sawtell scars on the edges of stands or patches of trees would Peak sites will also be sampled intensively. All other be insufficient evidence for a mixed-severity regime, sites are being sampled with the extensive protocol. as these fire scars may be recording fires adjacent to stands rather than within them. In contrast, multiple Laboratory methods used to process and scarred individuals in the interior of a stand would analyze data collected in 2003 will also be used in indicate that non-lethal understory bums have 2004. Increment cores and fire-scar wedges were occurred in these communities in the past. In prepared for dating using standard techniques (Stokes addition, evidence for the alteration of this "mixed" and Smiley 1968, Amo and Sneck 1977). All samples fire regime post Euro-American settlement would be were visually cross-dated (Fritts 197 6) and compared provided by a reduction in fire frequency in the tree­ with master chronologies obtained from the ring record. International Tree-Ring Data Bank. Samples that were not cross-dated with accuracy were excluded In sites sampled thus far, we have not found from analysis. After cross-dating, samples were any evidence of a historic fire regime that has examined and dates were assigned to fire scars within included a low-severity understory component. No ring series. Positions of fire scars within the rings trees in any of the three sites in the Beartooth or were also assigned to assess possible seasonality of Absaroka Mountains have had multiple fire scars. past fires. Dates of fire scars are now being compiled Several whitebark pines have been found with single into a composite fire chronology (Dieterich 1980) scars, but there origins can not be confirmed as fire­ induced. Scarring can be caused by a number of The FHX2 program (Grissino-Mayer 1995) agents in the GYE including fire, diseases such as will be used to compute composite fire interval butt and root rots, insects such as mountain pine statistics for all sampling areas. To describe central beetle, mechanical disturbance such as tree-fall, and tendencies in fire interval distribution, both the mean animal damage, i.e., bear scars. Determining the fire interval (MFI) and the W eibull Median cause of single scars on individual trees has been Probability (WMPI) will be computed. Fire regime problematic. As a result, we have developed a key to statistics will be calculated for three distinct time aid in the selection of scarred trees to sample during periods in order to compare fire interval information the 2004 field season. This key includes a checklist to periods of different human activity levels: 1780- of scar features, such as size and shape of scar, 1850 (pre-Euro American settlement), 1851-1920 presence/absence of bark on scar face, multiple scars, (active settlement), 1921-present (modem fire etc., to ensure that scars that are sampled have a high exclusion). Because summary statistics of fire likelihood of being caused by fire. Without this interval data are highly sensitive to these subjective precaution, scars not caused by fire may be sampled, time periods, fire trends will also be graphically and erroneously used to calculate fire statistics for the described as the percentages of sites recording fires in stand. individual years instead of using absolute numbers (Sherriff et al. 2001 ). By doing so, the bias associated The most convincing fire scars that we with disappearing evidence of older fires as old trees found on whitebark pine occurred at the edges of die is reduced. stands, adjacent to subalpine meadows. These too, were individuals bearing single scars. Due to the fact Synthesis and evaluation is being conducted that nearby trees within the stand did not show signs through the assessment of processed field and of scarring, these scars may be recording fires that laboratory data. occurred primarily in the meadow, rather than in the forest stand. Attempts will be made in the 2004 field

Published by Wyoming Scholars Repository, 2003 3 University of Wyoming National Park Service Research Center Annual Report, Vol. 27 [2003], Art. 17

126 season to verify if white bark pine stands elsewhere in the GYE tend to record adjacent meadow fires Our data collected in 2003 indicate that in at without fire spreading throughout the whitebark pine least some locations in the GYE, whitebark pine communities of interest. communities found in a matrix of dense, subalpine forest have little evidence of mixed fire regimes, and We hypothesized that whitebark pine that current stand structure may be the result of communities surrounded by an extensive, continuous infrequent, stand replacing fire events coupled with subalpine forest composed of spruce/fir and subsequent nutcracker seed caching. Additional lodgepole pine are characterized by infrequent, high­ sampling in 2004 will examine a greater variety of intensity crown fires which control stand structure in landscape contexts in which whitebark pine occurs in both the whitebark pine community and the the GYE, and may reveal additional variability in surrounding subalpine forest. Thus far, our sampling historical fire regimes in these communities. has been conducted in these types of extensive, subalpine forest. The lack of evidence for low­ + LITERATURE CITED severity fires occurring in these sites tends to support our idea that infrequent, stand-replacing fires are the Amo, S. F. 1986. Whitebark Pine Cone Crops: A dominant fire type. The age data that has been Diminishing Source of Wildlife Food. collected also supports this hypothesis. The stands Western Journal of Applied Forestry 1:92- sampled thus far have whitebark pine canopy trees 94. that often exceed 350 years of age, with canopy spruce trees that are slightly younger. The average Amo, S. F. 2000. Fire in Western Ecosystems (Ch. age of canopy whitebark pines across the three sites 5). in J. K. Brown and J. K. Smith, editors. sampled in 2003 is 307 years. The Wind River Lake Wildland Fire in Ecosystems: Effects of Fire site was sampled intensively to compare age structure on Flora. USDA Forest Service, General between species in three size classes (Table 1). Technical Report RMRS-GTR-42-volume 2.

Amo, S. F. 2001. Community Types and Natural Table 1. A Comparison of Mean Tree Ages (in years) for Disturbance Processes. Pages 74-88 in D. F. Wind River Lake intensive sampling site Tomback, S. F. Arno, and R. E. Keane, White bark Engelmann Subalpine pine spruce fir editors. Whitebark Pine Communities: Canopy 273.1 212.8 147.2 Ecology and Restoration. Island Press, Understory 224.7 125.6 114.9 Washington, D.C. Sapling 45.2 63.8 59.6 *Size classes are defined for the purpose ofth1s study as: Arno, S. F., and S. Allison-Bunnell. 2002. Flames in a) canopy: dominant trees, at least 2/3 the height of tallest tree in stand; Our Forest: Disaster or Renewal? Island b) understory: those trees that could exploit a canopy Press, Washington, D.C. gap or become dominant with the demise of canopy individuals, 113 to 2/3 of canopy tree Amo, S. F., and R. J. Hoff. 1989. Silvics of height; and c) "sapling": younger trees (perhaps), those that are Whitebark Pine (Pinus albicaulis). USDA smaller than 113 of canopy height but greater than Forest Service, General technical Report 6mm in diameter. INT-253.

Our second hypothesis concerns whitebark Amo, S. F., and K. M. Sneck. 1977. A method for pine communities not surrounded by ·a dense determining frre history in coniferous forests subalpine forest, but rather found in close proximity of the mountain west, General Technical to more frequently burned Douglas-frr forests, Report INT-42. USDA Forest Service shrublands, or grasslands. In communities such as Intermountain Forest and Range Experiment these, we expect that the historical fire regime was Station, Ogden, Utah. dominated by bums of varying frequency and severity, often having small extents. Communities in this type of landscape context, such as in the Wind River Mountains of Wyoming, and the Centennial Mountains of Idaho will be sampled in the 2004 field season. Here, we may find evidence more characteristic of a mixed-severity fire regime. + CONCLUSIONS

4 Romme and Walsh: A Comparison of Fire Regimes and Stand Dynamics in Whitebark Pine

Published by Wyoming Scholars Repository, 2003 5 University of Wyoming National Park Service Research Center Annual Report, Vol. 27 [2003], Art. 17

6