percent of the total. The brochure also explained how to 

 find each trailhead, which is rendered somewhat diffi- 

 cult by an intricate network of roads through private 

 ranch land in the Bitterroot Valley. These roads must be 

 traveled to reach the trailheads at the base of the 

 mountains. There also was some very limited informa- 

 tion on a few heavily impacted campsites. 



Distribution of the brochure began in 1974. It was 

 available in boxes on registers at every trailhead (fig. 2) 

 and at the Stevensville Ranger Station, the Forest 

 Supervisor's Office in Hamilton, and the Forest Service 

 Regional Office in Missoula. Italsowasmailedtopeople 

 who wrote to these offices inquiring about visiting the 

 Wilderness. The brochure, revised in 1 977, is still in use. 

 The current version provides less detail on use distribu- 

 tion, indicating only those four trails most heavily used 

 and pointing out that they account for 70 percent of ail 

 use. 



Evaluation 



The Wilderness Management Research Work Unit 

 evaluated this management effort at the District 

 Ranger's request. The evaluation had four objectives: 



1 . To determine if use patterns shifted in the desired 

 way, from heavily used to lightly used trails. 



2. To determine to what extent visitors were ex- 

 posed to the brochure; that is, what proportion obtained 

 it, where, and at what stage in the trip planning process. 



3. To determine how visitors reacted to the bro- 

 chure; that is, whether it influenced choices of trails, if it 

 was considered useful, and what changes were sug- 

 gested. 



4. To determine what factors influenced choices of 

 trailheads and if those factors were unrelated to infor- 

 mation in the brochure. 



The field conditions were the same as they would 

 have been without the study. The Ranger District 

 planned and handled the brochure distribution in the 

 normal way. Trail registers (used as an index of use at 

 each trailhead) were maintained in the usual fashion. 

 Use regulations were essentially unchanged through- 

 out the 1 973-75 period of the study. 



Use Patterns 



Use patterns, by trailheads, for 1 974 and 1 975 (the 

 first years the brochure was distributed) were compared 

 with 1973, a typical use season that served as the 

 "before treatment" base. Trail register data were used 

 as an index of relative use, expressed as a percentage 

 of the total to make it independent of changes in total 

 use from year to year. These data provided an incom- 

 plete measure because some visitors did not register. 

 But the data were assumed to be comparable across the 

 3 years. We checked trail registration rates by direct, 

 unobtrusive observation in the field at five trailheads in 

 1974. 



Comparisons of relative use were made among indi- 

 vidual trails, between all lightly used trails as one group 

 and all heavily used trails as a second group, and among 

 trails grouped into three categories— light, moderate, 

 and heavy use. 



Change from 1973 was classified as to whether or 



not it was in the direction desired. This meant relative 

 use was desired to decrease at heavily used trails and 

 increase at lightly used trails. The moderately used trails 

 did not have a clearly desired change and were not 

 classified. We tested the hypothesis that change was in 

 the desired direction. 



Visitor Survey 



To measure exposure of visitors to the brochure and 

 people's reactions to it, we sent a mail questionnaire to 

 a sample of people making visits to the study area in 

 1 974 and 1 975. The sample list was drawn from regis- 

 tered visitors, and all findings and conclusions apply 

 only to visitors who registered. One person, the one 

 whose name appeared on the permit, wassampled.This 

 person is usually the party leader. For the group be- 

 havior we were concerned with, this is usually the key 

 person. 



This was a systematic sample with a random start. We 

 sorted trail register cards for each trailhead into day use 

 (in and out the same day) and campers (overnight stays). 

 These were ordered by entry dates. Aftera random start, 

 every third camper card and every sixth day user card 

 were chosen for the sample. We sampled campers more 

 heavily because we felt they were more critical: they 

 penetrate the wilderness more deeply, stay longer, and 

 have greater potential for causing impacts. Also, 1973 



Figure 2.— Trail register and brochure dis- 

 penser box (to the right of the trail 

 register) on the Big Creek trait 



4 



