USING VEGETATIVE ANALYSIS TO DETERMINE
THE EXTENT AND FREQUENCY OF AVALANCHES
IN LITTLE COTTONWOOD CANYON, UTAH

by

Michael J. Jenkins, Ph.D.
Associate Professor

and

Elizabeth G. Hebertson
Research Assistant

Department of Forest Resources
Utah State University
Logan, Utah 84322

INTRODUCTION

The Little Cottonwood Canyon road, Highway U-210, connects Salt Lake City with the Alta and Snowbird ski resorts in the Wasatch Mountains of northern Utah. The road is nine miles long and traverses 42 avalanche paths with half the road in avalanche path runout zones. Average daily traffic totals during the ski season are estimated to be nearly 7000 with peaks in winter months reaching over 10,000 (Bowles and Sandahl 1987). It is possible, even in recent years of below average snowfall for avalanches trap or bury vehicles on the highway. This fact was documented on January 15, 1991 when heavy snowfall and avalanches stranded a Utah Transit Authority bus in the neighboring Big Cottonwood Canyon. The resulting traffic jam left other vehicles in paths where they were subsequently buried by other avalanches (Scroggin 1991). It is obvious that such a scenario greatly hampers rescue efforts, and with escalating highway use, is of considerable concern to avalanche forecasters and rescue personnel responsible for public safety.

The success of potential rescue operations requires the development and implementation of plans based on knowledge of the frequency and maximum runout of large avalanche events. Dendrochronology and vegetative analysis may prove useful in augmenting historical records and existing maps to determine past avalanche activity.

Avalanche dating techniques using dendrochronology and vegetative analysis are based on the concept that vegetation growing along the flanks, and within the run-out zone of avalanche paths may experience damage resulting from the impact of snow and debris. The type and extent of damage may vary depending upon the geomorphic situation, the size of the avalanche, and the type of snow (Burrows and Burrows 1976). Trees respond to this damage in a variety of ways making it possible to date avalanche events and determine the frequency of their occurrence. The following discussion summarizes a number of papers that describe the use of dendrochronology and vegetative analysis for performing avalanche studies (Burrows and Burrows 1976, Schroder 1976, Carrara 1979, Bryant et al. 1989, Hansen-Bristow and Birkeland 1989).

Avalanches can cause wounds when rock and debris carried from upslope strike standing trees resulting in scars on the uphill side of stems and branches. Trees form callous tissue around the margin of scars, and gradually add new wood and bark. A wedge cut from either side of the scar allows rings of new wood formed after scarring to be counted, and the date of injury determined.

The force of wind blasts from large, dry powder avalanches, and the direct impact of snow can break the main tree stem, or tear and fracture roots. The shear force of impact can also disrupt the tree's cambium. These types of injuries result in an abrupt reduction of tree growth. Tree-ring analysis may reveal significant growth loss as an atypical series of narrow rings.

Avalanche impact can tilt or uproot trees causing them to produce reaction wood to retain upright growth. In conifers, the reaction wood is produced on the downhill side of the stem. The reaction wood differs from normal wood by its thick, dense cells and dark, reddish-brown color. Deciduous trees produce reaction wood on the uphill side of the tree. The cells are long, and dense with a grey or yellowish color. Reaction wood formation also causes eccentric ring growth resulting in wider than normal rings. Counting rings from the bark inward to the year of initial reaction wood formation provides a reliable date for the avalanche event.

New vertical stems may develop from release of lateral branches or dormant buds on the upper surface in response to tilting or stem breakage. Approximate avalanche dates can be determined from ages obtained from these new stems by counting the annual rings.

The uprooting of trees by avalanche impact creates gaps in the forest canopy. With a reduction of shoot and root competition, residual trees may experience enhanced growth indicated by wider ring widths. Patterns of release may be particularly evident in undamaged trees, or trees bordering the new trimline.

Avalanches commonly strip vegetation from the slope. Aspen have the ability to revegetate a denuded slope by suckering. Run out zones typically have patches of different age-classes of aspen established after individual avalanche events. The year of an avalanche event can be approximated by finding the age of the oldest aspen in each patch. The succession of shrubs and herbaceous plants following the disturbance also serves as an indication of past avalanche activity.

Each method described above has limitations in its ability to provide accurate dates. Soil creep, landslides, and other geomorphic processes can tilt trees resulting in reaction-wood formation and growth suppression. Other agents, including lighting and wind, can break trees. Insects and diseases can cause reduced growth, and site improvement can enhance growth. Finally, reaction wood formation, release, and sprouting can all be delayed leading to misinterpretations. Where avalanche activity cannot be determined from historical records however, vegetative analysis, particularly dendrochronology provides reliable dates for the occurrence of avalanche events (Bryant et al. 1989).

The objective of the study was to determine the extent and frequency of maximum run-out in Little Pine East avalanche path using techniques of dendrochronology and vegetative analysis.

METHODS

Sampling Procedures

Recreation, wildlife, aesthetic, and watershed values in Little Cottonwood Canyon prohibited destructive sampling, and required that sampling be done by extracting increment cores from selected trees to analyze reaction-wood formation and other morphological responses. Transects were run through the terminal portion of the Little Pine East slide path perpendicular to the fall line at one chain (66 feet) intervals, and to the expected absolute maximum extent of the run out zone. Plots were placed along these transects at 2 chain intervals and a subset of the plots was randomly selected for sampling. At each sample point a 10 basal area factor variable plot was established to select the sample trees.

The total sample contained 50 conifers including Engelmann spruce (Picea engelmannii), white fir (Abies concolor), alpine fir (Abies lasiocarpa), and 28 aspen (Populus tremuloides). Up to four increment cores were taken from the uphill and downhill sides of each sample tree. The majority of the aspen sampled had heart rot caused by a fungus that decays the wood tissue in the center portion of the stem making infected trees useless for study. Other aspen cores were twisted making it difficult to see the annual rings. Consequently, only conifer cores were used for tree-ring analysis. Trees observed to have bole damage were sampled by cutting a wedge from the scar, or by taking a core sample using the technique described by Arno and Barrett (1988).

Aspen representing the oldest trees from different age-classes were randomly selected from the west flank, center, and toe of the slide path to determine the year of their establishment. Selected trees were felled with a chain saw and discs cut at ground level. Severely tilted Engelmann spruce and alpine fir with new stems growing from the old bole were observed in several locations in the slide path. From one alpine fir, a disc was cut from the base of two new stems for aging. All increment cores, wedges, and discs were labeled and transported to the Utah State University Avalanche Study Laboratory.

From modified skeleton plots, it was possible to establish approximate dates of avalanche events. To obtain more precise dates, increment cores were cross-dated. Using a dissecting scope, and a Model 4 Digital Display Unit (Fred C. Henson Co. Mission Viejo, CA) with a Tree-Ring Increment Measuring System (Madera

Software, Tucson, AZ) the ring widths of each sample core were measured. Skeleton plots were constructed using computer generated graphs of ring widths for each core. Because sample cores had not been collected from trees in off-path locations, a chronology was developed from a sample of cores with-no evidence of morphological response to avalanche. Growth patterns observed in skeleton plots of core samples were compared with those derived from the master chronology in addition to those of other cores. From this information, dates were assigned to the rings and years of reaction wood formation.

Aspen discs were sanded to reveal the annual growth rings. Beginning with the outermost ring, rings were counted inward to the pith to determine the age of the sample tree. The same procedure was used to determine the age of disc samples cut from the new alpine fir stems. One increment core taken from a new stem had reaction wood on what would have been the underside of the branch surrounded by wood formed following years of release.

Estimating dates of earlier events was more difficult. Only a small number of older trees were still living and available for sampling. With time there is also an increased probability of encountering false and missing rings, and morphological changes in response to damage caused by other agents. From modified skeleton plots, however, dates for earlier events could be approximated.

Skeleton plots derived from graphs of measured ring widths proved useful in substantiating these dates. Although the original purpose of these plots was to cross-date samples taken from different trees, the variability in ring widths made cross-dating difficult. Skeleton plots did, however, reveal patterns of suppressed growth associated with years of reaction wood formation and other morphological responses when compared to the modified skeleton plots (Fig. 1). Abrupt decreases in growth increment in avalanche areas suggesting avalanche activity has been reported in other studies (Carrara 1979). Where comparisons were in close agreement, dates could be adjusted producing a better estimate of the actual date of avalanche occurrence.

A total of 13 avalanches reached maximum run-out in the Little Pine East avalanche path between 189S and 1991. The dates determined for these events along with the maximum extent of runout are shown in Figure 2. Four large Engelmann spruce growing at the bottom margin on the east flank provided the longest record of avalanche activity. Reaction wood was formed in these trees as a result of 1898, 1901, and 1906 avalanche events. Narrow ring patterns produced during the same years that younger trees initiated reaction wood formation (1911-1916, the early 1930's,1935, early to mid 1940's, early 1950's, and 1965) indicated that patterns of reduced growth were probably a consequence of damage caused by avalanche impact. Perhaps the largest avalanche event detected occurred in the early 1920's (1920 or 1921). Trees sampled from the far east flank of the avalanche path, to west of the White Pine trailhead had evidence of reaction wood or reduced growth. Other avalanches more typically ran to the extent of the dashed lines shown in Figure 2. The most frequently recorded event was the 1965 avalanche, where sixteen trees had reaction-wood or reduced growth.

Other vegetative analysis and historical information provided verification for some dates. The age of aspen sampled from different age-classes along the west flank corresponded with dates from the 1920-1921 and 1965 avalanches. The minimum year of establishment for the aspen along the far west flank of the avalanche path was 1924. The aspen sampled from the younger age class was approximately 28 years old suggesting that these trees were established following the 1965 avalanche. The ages determined for aspen growing in the center island and toe of the avalanche path verified the 1951 date. Finally, the Engelmann spruce scar was dated to 1945 indicating some avalanche activity during the mid 1940's.

Historical records of avalanche activity in Little Cottonwood Canyon were used for further verification. The 1983, and the 1965 events were confirmed by historical records (UDOT 1987). Kalatowski (1988) cited accounts of avalanches occurring in Alta in 1898, 1906, and 1911. The agreement between this documentation and tree-ring analysis supports other findings suggesting that morphological responses are useful for dating past avalanche events with reasonable accuracy (Bryant et al. 1989, Shroder 1976, Burrows and Burrows 1976, Hansen-Bristow and Birkland 1989).

Avalanche frequency was calculated by dividing the number of years examined by the total number of avalanches. Calculations determined that large avalanche events occurred in the Little Pine East path once every eight years prior to 1991. The frequency of avalanche events prior to 1960 was once every six years, and decreased to once every 15 years between 1960 and 1991. The increase in avalanche return interval is perhaps a consequence of avalanche control which reduces the likelihood of avalanches running to their maximum extent.

LITERATURE CITED