Gina Collignon  
 
Pride, Prejudice, and Pack Stock:
Monitoring Alpine Meadows for the Future of America

ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
ACKNOWLEDGMENTS

POWER POINT

Abstract
Pack stock have had a historic role in the development of the Sierra Nevadas. Since the early 1800s horses, donkeys, and mules have played a huge role in the exploration of these mountains as well as serving as a widely used and appreciated recreational tool. However, there are growing concerns about the irreversible effects pack stock can have on alpine meadows. In April 2000 the High Sierra Hiker’s Association filed a lawsuit against the US Forest Service for their pack stock guidelines and permit quota system within the Ansel Adams and John Muir wilderness areas. The court mandated that the Forest Service report on the current state of those meadows visited by pack stock and also called for a long term monitoring system. This study aimed to create a simple, representative sampling method to correlate vegetation communities with impact amount and intensity. We hoped to create a sampling method simple enough for the non-expert to perform, but detailed enough to provide invaluable predictive information on pack stock impacts.

Introduction
Opinions over the proper use of wilderness lands are as diverse and contradictory as the many ways people propose to use them. When such highly-coveted and hotly debated areas are protected by a Wilderness act dedicated to establishing, “a National Wilderness Preservation system for the permanent good of the whole people, and for other purposes” (16 U.S. C. 1131-1136), conflicts often arise as to how to utilize such a vaguely defined and protected commodity. When the government claims to set aside lands for the “whole people”, it essentially claims that it will manage the land with the varied interests of cowboys, hippies, baby-toting car campers, ranchers, boy scouts, frustrated office workers, hikers, horse lovers, and zealous field biologists all in mind. While each group has a legal right to utilize wilderness lands, problems arise when the groups’ interests clash. In April 2000 (2004) the High Sierra Hiker’s Association filed a lawsuit against United States Forest Service. They alleged that the Forest Service, through its policy of separate, special permit guidelines for pack stock use in the Inyo and Ansel Adams National Wilderness Areas, directly violated the National Forest Management Act, the Wilderness Act, the National Environmental Policy Act, and the Administrative Procedure Act (2001). The District Court of Northern California ruled in favor of the High Sierra Hiker’s Association only on its claims of misconduct according to the National Environmental Policy Act (2001). The Court ruled that the National Forest had not monitored environmental impacts as required by the NEPA. It mandated that the National Forest Service complete site-specific analysis for pack stock use by 2006.
Those managing publicly owned lands must often feel like a supervisor during recess, holding a kid with each arm as they continue swinging at each other. Neither kid wants to share but both have a right to the basketball court. And, in this case, both kids have a predisposition to sue. The allegations against the Forest Service were made by a non-profit organization of hikers that “feel that the management agencies in the High Sierra are heavily biased in favor of commercial interests such as horse and mule packers, cattle and sheep grazers, and mining companies” (2004). However, as often as these hikers make claims about the detrimental effects pack stock (mainly horses, donkeys, and burros) can have on native vegetation, they also complain about the resultant dung on trails, noise levels, and general misdemeanors against their wilderness experience. Beyond pack stock, this group, “[is] also very concerned about other threats affecting hikers, such as the unfairness of current wilderness quotas and permit systems, and the disruption of the natural quiet caused by increasingly common military training overflights” (2004). In response to the hikers’ lawsuit against the Forest Service, a group of pack stock owners countersued, asserting their rights to the land and the use of it. As strongly as the hiker’s association may represent the evolving trend of environmentally aware wilderness purists, pack stock businesses represent a historic (and often romanticized) way of life, reminiscent of cowboys and early frontiersmen making their living from the land. They represent a powerful legal group to contend with. The hikers’ association cannot demand the total eradication of pack stock use in wilderness areas. Nowhere has the Forest Service ever claimed that it aims to rid wilderness lands of pack use. In its Proposed Action Plan, the Forest Service aims to “balance resource protection in the wilderness including the protection of wilderness character, with the economic realities of small, locally based tourist businesses” (USDA 2004). The Forest Service must somehow create an effective monitoring system to assess the damage pack stock inflict upon alpine vegetation. They must demarcate the line between true biological disturbance and aesthetic discordance. This system must work not to determine whether pack stock have detrimental effects or not, but must determine at what levels pack stock use should be maintained.
Alpine meadows in the Sierra Nevadas have become hot spots of debate because of the many ways people wish to use them. Meadows hold great importance for their “scenic vistas, wildlife habitat, livestock forage, water resources, and wide variety of recreational uses” (Sarr 1995). Furthermore, meadows have, “high amounts of biological diversity, ecosystem function, and aesthetic interest” (Moore et al. 1999). They can produce up to 2 tons of forage per year, act as natural water purifiers, and harbor more than 110 species of plant (Blaney and Moore). Meadows receive high amounts of pack stock impacts primarily because pack stock groups used meadows as camp sites (McClaran and Cole 1993). Many times, horses are allowed to graze overnight. Impacts to meadows affect both the meadows and everyone who wants to use them.
The detrimental effects of pack stock are most prominent on trails and in meadows. The impact of 1,000 pound animals with iron-clad feet often turn trails into veritable sand traps, incurring the wrath of any following hikers due to the resultant dust and dung. Horse impacts on trails produce greater sediment yields than impacts from humans or motorcycles (Wilson and Seney 1994). One might argue, however, that any trail is an inherently impacted area and that pack stock have as much a right to it as any other 1,000 pound hiker. Meadows, however, are sensitive biological areas that can be seriously damaged through heavy use.
Impacts incurred in meadows are can drastically alter a meadow’s defining physical and vegetative characteristics. Meadows form through the slow accumulation of a shallow water table, through the accumulation of fine, textured sediments, and through the development of moisture retaining peat deposits (Sarr 1995). Alpine meadows have only a two month growing season. In riparian ecosystems of arid and semiarid regions (such as found in high elevations in the Sierra Nevada) grazing impacts may intensify in severity due to the meadow’s biological importance and rarity (Cole and Landres 1996). Impacts that alter the water table may be largely irreversible (Cole and Landres 1996). In terms of restoration, it is harder to rebuild a stream than it is to replant a tree.
In order to monitor pack stock impacts in alpine meadows, the Forest Service formed a diverse team of experts competent in assessing all aspects of meadow health. The team included a hydrologist, a botanist, a biologist, a soil scientist and a range conservation specialist. This team then spent a field season trekking throughout the high Sierras. They visited meadows open to pack stock visitation and, based upon the state of the meadow during their visit, rated each meadow for its suitability for pack stock use. The process of evaluation was first written up in paragraph form in which reasons (including meadow attributes, present impact amounts, and each expert’s opinion based on their particular area of knowledge) for a meadow’s evaluation were stated together. The Forest Service then began to create spreadsheets with categorical data. This spreadsheet began to organize knowledge about meadow impacts into a more quantitative form, allowing for any patterns in impact amount or severity to materialize.
Pack stock impacts to meadows can take various forms. They can alter both the physical and vegetative characteristics of a meadow. Vegetative impacts can occur through defoliating vegetation, depositing wastes, trampling, and interacting with wildlife and with visitors (McClaran and Cole 1993). (Olson-Rutz et al. 1996) et. al report that “Pastured horses can impact up to 90% of the forage on pasture with trampling, urination, or defecation”. Selective grazing may influence plant competition and increase erosion (leaf litter helps weaken the force of raindrops on soil) (McClaran and Cole 1993). Cole (1996) notes that “even at low levels, grazing can compromise the naturalness of wilderness mountain meadows”. Dung deposits influence soil nutrient status, water quality, plant vigor, bacterial growth and defoliation patterns. Urination increases nitrogen levels and increases defoliation (McClaran and Cole 1993).
The April 2000 lawsuit against the Forest Service primarily noted the physical impacts that pack stock can have on alpine meadows. Physical impacts are usually more severe than impacts on meadow vegetation and can occur more quickly and with greater severity. Trampling includes walking and rolling and can cause soil compaction and shearing. These actions decrease the space between soil particles and lower water infiltration and oxygen diffusion (McClaran and Cole 1993). Trampling can also cause mechanical damage to plants and eliminate those species that are not structurally adapted to withstand injury (Hagberg 1995). This may increase bare ground cover and erosion. Trails leading to water sources often become highly eroded and may lead to an eventual collapse of that area of a stream bank, etc. All of these impacts lower meadow quality and may eventually lead (through plant species composition change or through drastic changes in hydrology) to a disappearance of the meadow altogether.
Though it seems preferable to have meadow analyses performed by qualified experts, there would be many advantages in creating an alternate monitoring system that non-experts could perform. Experts not only cost the Forest Service more money, they also have tighter schedules. And there are less of them. Hiring a team of experts to trek throughout the high Sierras every year or two becomes an impractical expense. If the same job could be done in a quantifiable, repeatable way by a trained group of back country rangers, for instance, the Forest Service would be in a better position to monitor and compare meadow health throughout the years.
The goal of this study was to create a simple, quantifiable method for assessing meadow vegetation and impact level in alpine meadows of the Sierra Nevada’s. We hoped to not only be able to quantify impact amount, but to also correlate impact severity with meadow vegetation types. This study was performed in conjunction with Thoyre (2004). While Thoyre focused on different types of severity and their occurrence throughout the meadows, this study focused on vegetation. We hoped to define vegetation categories that could aid in classifying meadow types. We then hoped to utilize meadow types in assessing future impact severity. We wanted these classifications to be broad enough to be applicable to all alpine meadows in the Sierra Nevada, yet specific enough to be helpful in predicting impact occurrence. Composite lists of common plant species in Sierra alpine meadows have already been created. There are commonly known species that predominate in these meadows. We hoped to use these lists along with other meadow features such as moisture level, slope, life form, and percent bare ground to create classifications. Through this manner of correlation we aimed to be able to use vegetation type to predict impact severity.

Materials and Methods
This study was performed in the John Muir and Ansel Adams Wilderness areas of the Inyo National Forest. Data was first collected at Duck Meadow (10,500 ft.). Duck meadow is 0.377 acres and is adjacent to Duck Lake. Two years ago it received approximately 120 pack visitations. At this meadow transects began at the SW corner. Aerial maps of the meadow obtained from the Forest Service predefined the boundaries of the meadow. A compass was used to align a 30 m transect tape towards the north. Transects were arranged east –west every five meters. Every five meters north-south along the transect tape a .5 x .5 m quadrat was placed on the ground to the east of the transect line. Sampling in the meadow therefore created a .5 x .5m grid. At each quadrat primary, secondary, and tertiary species and primary, secondary and tertiary life form were noted. Height of primary species, vegetation class, percent bare ground, slope, and the presence of any impact were also recorded. A point sampling method was also used to record the primary species, vegetation type, and height at each point under the five meter interval of the transect tape.
Data collection was also performed at Ram meadow, 0.691 sq ft meadow located at 10,150 ft. Ram meadow is composed of two physically separate sub-meadows. It was chosen for analysis based on its accessible location and relatively small size. It was also chosen due to its high amounts of pack stock visitation.
Data collection methods were modified at Ram Meadow due to observations made at Duck. At Ram meadow, transects again began at the SW corner of each sub-meadow. Aerial maps provided the meadow boundaries. From the beginning of each transect ten meters were paced northwards by one of the biologists. A compass was used to align the biologist in the northern direction. Flagging was used to mark every 10 meter interval. Beginning at 0 meters, a 1 m2 quadrat was placed on the east side of the transect line. Within each quadrat, vegetation type, life form, vegetation height, bare cover class, litter cover class, and disturbance type was assessed and recorded (see Table 1).
Vegetation type was defined by moisture level with xeric as the driest to hydric as the most wet. Xeric was defined as dry soil with no visible moisture below the top layer of soil, mesic was defined as dry soil with a visible moisture level below the first level of soil, wet was defined as standing water below 5 cm, and hydric was defined as standing water above 5 cm. Life forms were based around the most commonly encountered species at alpine meadows and bare, litter, and disturbance cover class levels were based on the Dabmenmeyer scale. As vegetation data was collected, impact data was concurrently collected for analysis in a concurrent study (Thoyre 2004). Impact data was collected using belt transects along the east side of each transect line. Because of the differences in data collection, impact and vegetation data could not be directly correlated.
Multivariate hierarchical clustering within the JMP (SAS Institute Inc.) program was used to create clusters from the Ram data. Vegetation type, life form, litter cover, and disturbance (due to gopher activity) were used as ordering values. Data was converted to a numeric value and labeled as ordinal. Best fit was determined by visual assessment of a canonical plot within the discriminant analysis performed upon resultant clusters. Those clusters with overlapping values in the canonical plot were considered to be at too fine a level. At such instances the number of clusters was lowered. A frequency distribution of the final clusters gave a hypothetical probability value, showing percentage of each vegetation type within the total. A frequency distribution of the clusters was then performed on those quadrats with sod fragmentation and on those quadrats with trail presence. To do this, a separate clustering analysis was performed on the impact quadrats collected through the belt transects. Each quadrat was labeled with its cluster number and all quadrats with no sod fragmentation were eliminated. A distribution analysis on cluster type was then performed on those quadrats with sod fragmentation presence. The same method was performed on those quadrats with trail presence. The estimated probabilities from these impact distributions were tested against total probability of each vegetation type within the hypothetical probability Chi-squared test of the distribution function showed whether the two values were significantly different or not. In order to test whether or not impact distributions per cluster type were greater or lesser than total available area of that vegetation type, the equation

[(frequency of impact per veg type) – (total frequency of that veg type)] / (total frequency of that veg type) was used.

Results
Ram Data was used for all analysis purposes. The multivariate hierarchical clustering of the vegetation data collected at 10 meter intervals at Ram Meadow resulted in five clusters (see Table 2). Qualitatively these clusters can be defined as mesic/shrub, hydric, mesic/sedge with presence of gopher disturbance, wet, and mesic/sedge without the presence of gopher activity. The first cluster is characteristic meadow predominated by both Aster spp. and Vaccinium spp. The second cluster is defined by a dominance of Carex utriculata in pond areas. The third cluster is the one mesic cluster with gopher presence. It is dominated by Carex filifolia and Juncus drumondii. The fourth cluster includes mesic with Muhlenbergia spp. and Carex spp., while wet environments contain a mix of Carex utriculata and other sedge species. The clustering analysis also resulted in a similar 5 clusters when using the impact data collected along belt transects (see Table 2).
A Chi-squared test did not reveal significant differences between the frequency distribution of the vegetation cluster (Fig 4) and the frequency distribution of the trail impact (Fig 5). A Chi-squared test did reveal a significant difference between the frequency of the sod impact (Fig 6) and overall frequency of vegetation clusters. The Chi-squared test revealed a significantly greater (p < 0.0001) amount of sod impact occurrence in wet and mesic/sedge areas.

Discussion
The goal of this study was to investigate fast, efficient, and relatively simple methods capable of characterizing alpine meadows and pack stock impacts within them. In particular, we hoped to define set vegetation categories within the meadows that could be easily identified by a non expert. We hoped that the creation of these categories could aid in impact monitoring and prediction. The first half of this project involved a period of trial and error in which different methods and different levels of sampling were tried in an attempt to characterize Duck meadow. Qualities assessed in the monitoring methods included both accuracy of representation of the meadow and time spent collecting data. We also wanted a method that could be easily taught to a large number of potential data collectors.
Visual assessment and a trial of sampling at Duck meadow was sufficient to determine that the meadow characteristics could be captured at a 10 x 10m (and not 5 x 5 m) grid. While we felt that 10 x 10m were sufficient for Duck and one of the Ram sub-meadows, we also felt that one of the sub-meadows should have been analyzed on a 5 x 5m quadrat level. Sampling quadrat intervals may have to be determined at an individual meadow level. While a smaller sampling grid will always provide the most information, it seems possible to assess optimal grid size at the study site. A smaller grid may give more information, or it may just unnecessarily take more time.
Alpine meadows contain a core group of common plant species. These species are excellent indicators of vegetation communities. However, one of the biggest changes in methods from Duck to Ram meadow included the exclusion of species identity as an indicator. Though dominant plant species may be one of the best ways to determine vegetation community, it was also the hardest data collecting skill to learn. Because of the time and difficulty we had defining plant species, we opted to measure our quadrats by those factors that could easily be measured. Though it is difficult to define grass and sedge species, it is enormously less difficult to define a grass versus a sedge. This not only increases expediency, but also reduces errors in the resultant data due to misclassification. Many of the common plant species are easily identified not only by morphology, but also by surrounding moisture level, etc. So by forgoing species identification but allowing for the collection of data that could identify the most common plants post-data collection, we felt that we had come upon a representative yet simple sampling method. During our second trial of data collecting at Ram Meadow, our sampling methods were successfully taught to two field assistants who had never worked with alpine meadows before. The data from both sub-meadows was collected in a day and a half.
Before beginning the project we presumed that wet areas would be more heavily impacted than drier area. Common sense seems to dictate that wet, squishy areas are less resistant to downward forces than dry area. One study recommends that “trail damage can be minimized by limiting trail use when soils are wet” (Wilson and Seney 1994). The authors note the strong connection between soil moisture condition and “a soil’s ability to bear a moving load” (Wilson and Seney 1994). There was no significant difference between trail impact per vegetation type and overall vegetation frequency. Though the pattern seems to show greater trail impact frequency in mesic her/shrub and mesic/sedge areas, these findings are very insignificant (p = .7492). Low sample size is one of the most probable reasons for the insignificant data. Though there were trails in Ram meadow, they did not traverse across all areas of the meadows. There was a greater sample size for sod impacts than for trail impacts. Sod impacts should a significant increase in wet and in mesic/sedge areas. The high occurrence of sod impact in wet areas agrees with our hypothesis.
Our data seems to indicate that wet areas are more sensitive to impact than drier areas. More importantly, the data used to come to this conclusion was collected in a day and a half by only four people (who also simultaneously sampled for Thoyre 2004). Stronger conclusions about the relationship between vegetation communities and impact severity could be made more easily with data from a larger number of meadows. There are also some direct questions that could be answered in a more experimental manner. For instance, do wet areas receive greater amounts of impact from equal amounts of use, or do these areas receive greater amounts of impact due to greater amounts of use? Many pack stock owners have claimed that horses do not prefer the wet areas of meadow. Is this true? Does anyone actually know where a horse goes late at night? And even if horses do not generally prefer wet areas, would this preference change once meadows starting drying out and wet areas contained the only green vegetation in the area?
We know a few set things. Horse impacts hurt meadow health in known ways. How can we minimize these impacts? Impacts, for whatever reason, center in wet areas. Those meadows with greater amounts of sensitive vegetation are becoming off limits for pack trains. Many pack groups now bring in their own feed. This minimizes the amount a horse needs to graze. There are also growing regulation on where and how horses can be picketed. A picketed horse centers his impact on a very defined area. The longer the horse remains in this area, the more chance for greater impact. People are also beginning to look for other means of carrying cargo. Both llamas and goats are beginning to be used as pack animals. Both lack the weight and iron shoes of a horse and goats have the added benefit of being able to eat unwanted leftovers (besides their ability to follow a hiker untied much like a loyal dog would). Though a horse or a llama or a goat in a meadow can never leave that meadow unimpacted, the same can be said for humans. Even field biologists must make marks in the mud as they look for signs of horse impact. The more we can quantify and cross compare knowledge bases, the more we can effectively monitor the inevitable effects of our presence here on this earth.
We have two definite extremes possible when it comes to thinking about effective land management. There are those who feel that the less interaction we have on the land, the more pure it is. We, as humans, taint the natural world. To preserve wilderness we must minimize/reduce all human impact. As popular as this belief is becoming, it fails to acknowledge our inherent dependence upon the land. We become like Homer Simpson, warped through time by a malfunctioning toaster and trapped in the Jurassic past. Are we afraid to touch anything lest it change the whole course of the earth’s development? And what if we accidentally sit on that fish? Will the world end if one species disappears? No, it will not. However, we must not then fall into the opposite extreme. The earth is not the infinite realm of untouchable vastness that it was once believed to be. Humans have the power to seriously damage our home. While one person with one horse traveling throughout the Sierra Nevadas may make no serious impact on the overall ecosystem, once that person brings 10,000 of her closest friends and then invites them back each weekend, the impact intensifies. It suddenly matters where the horses graze and where and how the people poop. We must learn to integrate practical reality with idealogical tree hugging.
Though most of our lives are framed by either concrete or the cold steel of a car body, we are still very much connected to the land. We need it and we need to maintain it within certain biological parameters. What difference does it make if one species dies? We know now that we are not so different from animals. Our genetic makeup, our embryonic development, and our needs as organic beings still strongly resemble each other. Miners once brought canaries down to the caves with them as a sensitive indicator of air quality. If the canary died, the miners knew that conditions were bad it was time to get out. Plant and animal species are dying around us in an alarming rate. Maybe we should take the hint and try to learn why these organic beings can no longer survive on this earth.
And what of the pack stock? Are they to be unfairly held responsible for human and animal impact in the Sierra Nevadas? Are their impacts any more than those of car campers located at the trail head? Do we castigate and restrict pack stock owners so that the hikers who drove their SUVs up the four lane highway and parked at the paved parking lot can hoist their hundreds of dollars of synthetic gear and enter the woods for the true wilderness experience? In Desert Soitaire (Abbey 1968) Edward Abbey suggests an intriguing possibility. Maybe we should restrict cars altogether from state and federal parks. People can drive to the entrance and hire pack stock to bring them, and all their gear, in. Those who once car-camped can now pack-camp. Multi-laned roads would not be needed, concrete would not be needed, those pack-camping would get a more “rustic” wilderness experience. And since pack stock business would be booming, we can then limit further trail use to those willing to hold all their gear. Given some sacrifices and some paradigm shifts in how we view our relationship to this world, the plan might work. The lion would sleep with the lamb, the pack stock owners and hikers would sing together in joyful celebration around the campfire, and the field biologist would tromp happily through the backwoods, looking for some other type of impact to monitor.

Acknowledgments
This study was supported by grant from the NSF. Many thanks to Eric Berlow, Peter Weisberg, and Chris Lourdie for their guidance. Erin Lutrik and Phil Kudu provided invaluable support. Thank you to everyone who works, plays, and lives at the White Mountain Research Station. Denise and Meagan not only helped collect data, but they proved that our methods could be learned. And Autumn rocks.

About Us Science Fridays Staff Corie Cann: Bio Gina Collignon: References Gina Collignon: Bio Gina Collignon: Main Page Home