0470-A3
Charlotte Young and Mike Wesner 2
One of the tenets of sustainable forest management in Canada includes the socio-cultural values the public has for forests. Methods to measure economic and environmental values have been considerably improved using the Criteria and Indicator (C&I) framework. However, the measuring of social indicators remains a problem. One social value reported in numerous (C&I) public surveys and stakeholder meetings is that people value the aesthetic qualities of the forest. However, despite the importance of this forest value, specific measurements have not been developed or researched on how the aesthetic quality contributes. This study measured public perceptions of aesthetic quality of industrial forest operations in the natural environment using a quantitative (spatial-chromatic) measure of aesthetics. Similar visual science applications appealing to public perceptions are used with the advertising and military industries' subtle use of conspicuity and camouflage. We chose to test comparable techniques never measured before in the forestry industry for social values of aesthetic beauty.
More specifically, we tested the acceptability of three industrial forest operations, measured under different landscape conditions: two clearcut patterns of checkerboard and irregular patches; and a logging roadway. Each of these landscape intrusions were modulated chromatically (on a 7-step visual scale) to a range of climatic conditions; and to 5 background conditions among boreal forest coniferous trees, with and without a lake to find the publics' position preference influences. Their preference levels from observing oblique aerial photographs taken from a remote airplane were likened to the type of disturbance, chromaticity and location. Each of the public volunteers (N=80) was presented with 84 randomized landscapes from a total of 420 forest images. The 420 images produced a total of 6720 public perception rating-observations. Data interpretation was conducted using a 4-way multifactor design with repeated measures on 3 factors (5 randomized backgrounds X 3 spatial patterns X 7 pattern chromaticities X 4 quadrant locations).
Results show that varying the spatial-chromatic properties of forest industry patterns will significantly influence the public perception of those activities as aesthetically offensive or acceptable. For example spatially, logging roads were rated as being the most acceptable landscape interruption. Irregular harvesting cuts were rated significantly more acceptable than checkerboard cuts from a distance. A discussion from the publics' post-test statements support and expand on additional results. Measuring the impacts of forest industry practices with public expectations has added-value with quantified social research.
Public participation in developing sustainable forest management programs is a recognized pre-determinant for acceptance of national policies. Support for forest management programs and policies are achieved by integrating public values in the content. In Canada three methods assess public perceptions of temperate and boreal forest values. These perceptions are identified through a set of criteria and indicators (C & I), which define cultural, social and spiritual values the public agrees upon by consensus. Public perceptions are also determined through the Social Impact Assessment (SIA) process for timber management and through Certification measures.
These documents define the future process for timber management and acknowledge public perceptions must be considered. However, in all cases no formal process of how to measure public perceptions and aesthetic values of forests is prescribed because of the wide variability found between public interests in managing wilderness regions. No previous measure specifies how or why a scenic quality contributes positively or negatively to human, social, economic, cultural, and spiritual values. Previous consensus ratings of public forests have identified what particular industrial landscape dimensions conflict with public values. Some of these were the visibility of clear-cutting, roads and foreground terrains.
The purpose of this study was to determine whether industrial forest operations distinguished by clear-cut shape, roads and foreground terrains from the public view are really the factors that influence aesthetic judgments positively or negatively in the natural environment. Rigorous experimental controls for chromatic and spatial variability for each of these three forest industry operations were predicted to be the determining source for how land-use impacts on perceived aesthetics. This would allow a more accurate public rating system for industrial operations used in a SIA or as a C&I.
Three forest landscape operations: two clear-cut patterns of checkerboard and irregular patches and a logging roadway were measured with experimental controls for chromaticity (visual colour contrast recognition) and for spatiality (position space of foreground, background, near lakes). These experimental spatial-chromatic controls were chosen to measure industrial forest landscape activities because visual processing effects are common to human perception cross-culturally.
The use of natural spatial-chromatics to measure public preferences of industrial forest landscapes has never been measured before in applied forestry research. However, these are the similar influences used by the advertising industry to design packaging and promotions to stand out and capture the consumers= attention in a commercial environment and to create areas of camouflage in a military environment (e.g., Carter et al. 1981). The three landscape conditions in this paper were chromatically matched to varying climatic conditions and background conditions to determine what the publics=position preferences would be.
The chromaticity of the three industrial targets were modulated (7steps) along red-green axes in Commission Internationale de l=Eclairage (CIE) 1931 chromaticity space. Each of the industrial patterns was presented in five background conditions of oblique aerial photographs of coniferous trees with and without a lake to determine the publics=proximity and position preferences. Each of the 16 observers per background condition (N=80) was presented 84 randomized landscapes from a total of 420 forest images. Data interpretation was conducted using a 4-way multifactor design with repeated measures on 3 factors (5 randomized backgrounds X 3 spatial targets X 7 pattern chromaticities X 4 quadrant locations).
The experiment used 80 undergraduate students from Lakehead University (with an alumni culture mix from 45 countries), ranging in age from 19 to 35 years. All people were tested for colour deficiencies using the Ishibhara colour plate test (24 Plate Edition).
A series of 420 landscape images were derived from a number of elements found in several digitized forest images. These elements were modified and standardized chromatically and spatially and pasted into a quadrant in a standard lake and no-lake background. There were six standard forest operation shapes composed to seven standard chromatic levels in nature. The images were presented on a 17", high resolution RGB monitor (Super Match 17.T). The images were modified using Adobe Photoshop Software (1991). Sample of these images are shown in Fig. 1 and 2 for backgrounds with lakes located near the horizon (Far Lakes) and close to the viewer near the bottom of the scene (Near Lakes) respectively. (See Fig.1 and Fig.2).
Three different forest industry operations served as visual target patterns. The target conditions were a checkerboard clear-cut, an irregular shaped clear-cut, and a logging road. The patterns were arranged for spatial pattern distance, foreground vs. horizon locations and left or right preference effects on public ratings. A total of 12 spatial images were used (ie. 3 Afar@ industrial patterns in quadrant 1 or 2 (6 total) plus 3 Anear@ patterns in quadrant 3 or 4 (6 total). The chromaticity of a forestry operation mid-tone varied by a total of 3 incremental and 3 decremental steps along a red-green axis in colour space (ie. A total of 7 chromatic steps). The highlights and shadows of the patterns and background were kept constant as was their overall luminance. This is because we were only interested in measuring the effects of location and chromaticity of the industrial operations on perceived aesthetics. The reference chromaticities for each of the industrial configurations are plotted in CIE 1931 (x,y) coordinate space along with the range of six realizable chromaticities used in all the scenes (Wyszecki et al. 1982, See Fig. 3).
Each of the forest operation patterns was presented in two background templates. One background simulated an aerial photograph of boreal forest. The second background was the same as the first except there was a lake positioned in one of the 4 quadrants on the CRT. Each public observer participated in only one background condition and rated 84 images. The total number of stimulus conditions was 420 (3 spatial patterns X 7 pattern chromaticities X 4 quadrant locations X 5 randomized background templates). The ratings of the 420 images produced a total of 6,720 observations.
Figure 1: Near Lake images
Figure 1: Examples of landscape images with ANear@Lake backgrounds. The specific forest industry shapes are labelled under each sample scene. CHKBD, RDWY and IRR denote checker-board clear-cut, logging roadway, and irregular clear-cut targets respectively. The chromaticities are shown.
Note: These images are reduced, second-generation colour scans of photographs presented on the SuperMatch 17.T Monitor
Figure 2: Same as figure 1 except target images are withAFar@ lake backgrounds
Figure3: ChromaticityDiagram
Figure 3: CIE 1931 (x,y) chromaticity diagram scale showing the general direction of the chromaticity shifts from green (50g, 25g, 10g) to red (10r, 25r, 50r). AN@ denotes neutrality of the mid-tone chromaticity for logging roadway, checkerboard and irregular clear-cut targets.
Prior to entering the laboratory, the public observers were told that they would rate the beauty of remote wilderness scenes as viewed from a low-flying aircraft. The background templates with and without the lake were presented to all observers (with no industrial intrusions) before they were designated into one of the five background conditions to begin rating the 84 scenes. Thus, only the spatial-chromatic properties of the industrial operation, varies with each presentation (not with the background). The public rated all randomly presented industrial configurations in their background condition during one single session. Interleaved within the session, two backgrounds scenes without forest industry intrusions were presented to see if the backgrounds alone were similarly rated.
The public was allowed to make qualifying statements at the end of the rating session. This allowed the public to describe their strategies for rating the scenes the way they did. Owens (1988) suggested that including a provision for qualifying statements enables researchers to identify values that are consistently important to the public. Also Schroeder (1991) noted that landscape ratings are much more informative in conjunction with an open-ended debriefing of the publics= thoughts and feelings about particular scenes. Following previous researchers recommendations, debriefing all observers as a check of preference frequencies with the rating data was conducted to bring further insight to our findings.
Results show varying spatio-chromatic properties of industrial operations are a common mechanism that significantly influences public perceptions of forest landscape activities as offensive or acceptable. Fig.4 shows the main effects from the analysis.
Overall, spatial patterns were found to have a significant influence on public perceptions for ratings of the three forest harvesting patterns (roadway, irregular, checkerboard; F(1.18, 88.47)=108.19, p<0.01; e=0.589). The scenes with roadways were rated the highest in terms of beauty. The checkerboard clear-cuts were rated less favourably than irregular clear-cuts. These main effects can be seen in fig. 4A. The preferences (means) for the roadways were found to be higher (significantly different) from the irregular and checkerboard means.
A significant chromaticity main effect was also found F(4.96, 372.20)=29.11, p<0.0001). The greenish appearing patterns (towards 50g) were rated higher in terms of beauty than the reddish patterns (towards 50r). The roadway chromaticities had no effect. The trends can be seen with standard error of the means (SEMS). The overlapping SEMs indicate areas where there are small differences between the chromatic levels.
Industrial shapes and their distances from the public observer were found to be an important factor of landscape beauty. Forest operations located in the AFar@ quadrants (Quadrants 1 and 2) were given better ratings overall than the ANear@ operations (Quadrants 3 and 4). Fig 4C shows that the main effect for quadrants was significant [F(1.91, 140.99)=108.56, p<0.001;e=0.636]. The Left versus Right ANear@ were not significantly different (t0.05 = -1.60, p=0.11).
Figure 4: Main Effects of public preferences
Figure 4: Preference ratings for forest industry operations.
Values are measured (in mm) from the bottom of a vertical analog scale that represents a Aleast@ to Amost beautiful@ rating:
The interactions between industrial operations for chromaticity and shape, shape and quadrant, and chromaticity and quadrant are shown in Fig 5, panels A, B and C, respectively. The interaction between clear-cut chromaticity and shape was found to be significant [F(8.86, 664.55)=8.96, p<0.001; e=0.738]. Greenish checkerboard clear-cuts and irregular clear-cut patterns were viewed more favourably than reddish checkerboard and irregular clear-cut patterns in backgrounds with Far/Left, and with foreground lakes. Chromaticity had little effect on roadways (compare solid triangles to squares and circles in Fig. 5A).
The interaction between forest operation shape and quadrant was significant [F(4.38, 328.39)=22.86, p<0.001; e=0.730]. The logging roadway consistently showed the highest rating for all pattern locations. The checkerboard clear-cut consistently showed the lowest rating for all pattern locations. Also, the scenes containing the Far industrial operations were always given higher ratings than the scenes containing the Near industrial operations. Roadways were found to have no significant preference according to distance.
The interaction between industrial-pattern chromaticity and position (Fig. 5C), and the three-way interaction between chromaticity, shape and distance were not significant. Interactions between background and all remaining factors were not significant. Whether the public viewed the industrial operations with or without a lake did not appear to influence overall ratings, although the public did report being more disturbed by viewing a forest industry operation with a lake than without one.
This study demonstrated patterns simulating industrial forest operations have an important influence on public perceptions for aesthetic values of forests. We initially suggested measuring changes in landscapes by common contrast mechanisms shared in human perception may be a good predictor of forest landscape aesthetics. In fact, our findings suggest that all industrial forest dimensions we manipulated in the study (eg. forest industrial patterns, chromaticity, and location) influence perceptions of scenic beauty.
The values measured perceived beauty of forest industry operations were supported by qualifying public comments that were made after they rated the scenes. For example, the public consistently commented about how clear-cut patterns (targets) were a major determining factor in how they rated the forest scenes. Seventy-five percent of the respondents commented on the checkerboard clear-cuts, and of these, 85% gave negative comments stating they looked unnatural and human-made. Sixty-three percent of the public made comments about irregular clear-cut shapes. Of these, 54% indicated that they preferred the irregular shape to the checkerboard clear-cut patches because they looked more natural (ie. a patch from a forest fire or insect infestation). A possible explanation for low rating responses on checkerboard clear-cuts is that even though square patterns are common visual patterns for agriculture farm regions, it is an extremely uncommon sight to see in a forest. The logging roadways appeared to have little negative impact on scenic beauty. Of the 55% of the respondents who commented on roadways,
95% stated they viewed roads as Abelonging to@ the scene. Only 5% of respondents disapproved of roadways.
Figure 5: Mean Rating interactions
Figure 5: Mean ratings of public perception interactions
The mean ratings are measured in mm from the bottom of a vertical analog scale that represents the Aleast@ to Amost beautiful@ rating for forest industry operations (targets).
The importance of chromaticity on scenic beauty was demonstrated by the high ratings for Agreen@ industrial shapes and by the post-rating public comments. Sixty-six percent of public respondents commented on industrial patterns appearing reddish. Of these, 88% disapproved of the patterns. Typical comments were that the reddish patterns appeared to represent a dying forest, recently destroyed by human hand. However, of the 73% of respondents who commented on the greenness of a target, 97% approved of Agreen@ patterns stating they appeared as natural forest re-growth. Only 3% of the public specifically disapproved of the Agreen@ stating forest clear-cut targets had an unnatural and inappropriate look about it.
The importance of spatial distance of industrial operations on public perceptions of forest beauty shows that obstructing background vegetation by placing forest industry patterns in the foreground significantly lowers ratings of scenic beauty. This was also evident in the observers= post-rating comments. Of the 35% of public comments on background (Far) targets, 100% rated the distant industrial forest operations as more acceptable than foreground (Near) industrial operations. For the spatial distance ratings between the Left/Right quadrants for clear-cut targets, both the Near and Far quadrants were consistent with no L/R significant difference.
Our results measuring the four Lakes and No-Lake conditions around forest industry operations show no significant differences in public perception ratings. In post-rating comments, only 23% of the public made comments about lakes. Of these, 15% indicated cuts near the lake were offensive. Interestingly, 7% of the respondents actually described irregular clear-cuts near lakes as having a beautifying effect because they appeared to represent recreational beach sites.
An issue to address with the present study would be that the public observers were a cross-culture of university student volunteers who are not generalized to be representative of the mainstream cross-culture population globally. A suggested future research agenda with this study is to conduct a cross-validation of these findings to other types of scenes and with other groups of public observers. For example, the roads could be examined further by varying their lengths and widths into the forest to match major harvesting operations. In addition, studies should modify the industrial target variables to accommodate the specific recommendations and requirements made from forest policy-management.
Measuring the impacts of forestry land-use with public expectations has added value when social research is quantified for the policy making process. The present study hopefully begins to establish measurable indicators of social values that may be incorporated into future C & I processes to maintain the natural beauty and values expected from forests.
Carter, E., and R. Carter, 1981. Color and conspicuousness, Journal of Optical Society of America, 71: 723-729.
Owens, P.E. 1988. Natural Landscapes, gathering places, and prospect refuges: Characteristics of outdoor places valued by teens. Special Edition: Adolescence and the Environment. Children=s Environment Quarterly, 5: 17-24.
Schroeder, H. 1991. Preference and meaning of Arboretum Landscapes: Combining qualitative and quantitative data. Journal of Environmental Psychology, 11: 231-248.
Wyszecki, G., W.S. Stiles, 1982. Color Science: Concepts and Methods, Quantitative Data and Formulae, Second Edition, John Wiley & Sons Inc., New York.
1 This paper was presented at the Canadian Psychological Association Conference, in Charlottetown, P.E.I. Ms Young is now with Policy, Planning and International Affairs, Canadian Forest Service, Natural Resources Canada, 580 Booth Street, Ottawa.
2 Dept. of Psychology, Lakehead University, Canada. [email protected]