Application of GIS to Measure and Evaluate Landscape Changes

Printer Friendly Format

Page 2 of 2
| Previous |

Application of GIS to Measure and Evaluate Landscape Changes

2.2.2.2 Test of landscape changes
To understand the effects of cutting and forest road building on forest landscape changes, this study used the t-test method proposed by Magurran (1988) to examine the variation of landscape changes in the study area. This method was based on the Shannon diversity index. The equation of t-test is as follows:

where H_i is the Shannon diversity index of landscape i or stage i, m_i is the number of patch types, n_i is the number of patches, and P_j refers to the percentage of either the number of patch type _j to the total number of patches, or the area of patch type _j to the total area. The latter definition of P_j was used in this study. Var H_i is the variance, and its equation and degree of freedom are as follows:

3. Results
The results of using GIS to quantify the landscape structure and eight quantitative indices to analyze the landscape changes were as Table 1. Table 1 depicts the quantitative indices of landscape structure derived from forest type maps of 1988 and 1996. The differences and the effects of cutting and forest road building on forest landscape of 1988 to 1996 were as follows:

Cutting and forest road building resulted in increase of patch numbers (PN) from 104 to 249 and patch richness (PR) from 21 to 30.
All the three landscape diversity indices (SHDI, SIDI, MSIDI) increased from 1988 to 1996, and SHDI had the largest increment among these three indices. The result reveals that the Shannon diversity index performed better in distinguishing landscape diversity.
All the three evenness indices (SHEI, SIEI, MSIEI) decreased from 1988 to 1996, and SHEI had the largest decrement among these three indices. The result reveals that the patches within the landscape tended to distribute more unevenly after cutting and forest road building, but the difference was not significant.
The interspersion and juxtaposition index (IJI), representing landscape configuration, decreased after cutting and forest road building. The difference in IJI is more prominent as compared to the indices of landscape diversity and landscape evenness. This reveals that the disturbance from cutting and forest road building resulted in more uneven distribution of adjacencies for the landscape configuration.

Table 1. Effects of cutting and forest road building on forest landscape from 1988 to 1996. Index 1988 yr. 1996 yr. Difference

To examine whether cutting and forest road building caused significant effects on landscape changes or not, a statistical method was needed to test the extent of landscape changes. Table 2 was obtained using t-test of Shannon diversity index at a=0.05. The results indicate that cutting and forest road building from 1988 to 1996 did not cause significant landscape changes in the study area.

Table 2. T-test results of Shannon diversity index
t value	0.222
Degree of freedom	204
Test result	NS¹⁾
¹⁾NS: not significant at 5% significant level.

4. Discussion
In this study, the forest type data obtained from 1988 and 1996 were used to quantify the landscape structure and to evaluate the effects of cutting and forest road building on landscape changes. The results were obtained based on: (1) the scale of forest type maps was 1/5000; (2) the maps were in vector format rather than raster format; (3) the Pj used for calculating landscape diversity and landscape evenness indices was defined as the percentage of the area of patch type j versus the total area. For different map scales, raster format maps, and defining Pi as the number of patches of type j versus total number of patches, the results need to be further studied and compared.

The quantitative indices of landscape composition can be used to analyze landscape changes. However, the landscape diversity indices had larger differences and more consistent results than the landscape evenness indices, and therefore are more suitable for landscape changes analysis. Among the three landscape diversity indices, the Shannon diversity index is most suitable for analysis.

The landscape structure indices including diversity indices, evenness indices, and interspersion and juxtaposition index were either slightly increased or decreased after cutting and forest road building. This implies that management activities can affect landscape changes. However, the t-test results of Shannon diversity indicate that effects of cutting and forest road building on landscape changes in the study area were not significant.

5. Conclusion
The study was concluded as follows:

Application of GIS and quantitative indices of landscape structure in evaluating the effects of cutting and forest road building on landscape changes is a feasible and effective approach. This approach can be used for landscape planning and monitoring of the study area in the future.
Management activities such as cutting and forest road building increased the numbers and types of patches. They not only directly altered the variety and abundance of the patches, but also indirectly affected the landscape structure properties such as landscape diversity, evenness, and interspersion and juxtaposition. Several indices were compared, and it was found that landscape diversity indices had larger differences and more consistent results, and therefore they were more suitable for landscape changes analysis. In addition, among the three landscape diversity indices, the Shannon diversity index performed better in distinguishing landscape changes.
From the t-test results of Shannon diversity index, cutting and forest road building did not cause significant effects on landscape changes in the study area. Therefore, from landscape perspective, management activities of the study area did not incur significant effects on the forest landscape up until 1996.

Reference

Burrough, P.A., 1986. Principles of Geographic Information System for Earth Resources Assessment. Clarendon Press, Oxford, UK, 193 p.
Cheng, C.C., J.F. Jan, and L.T. Hsu, 1999. Application of fractal dimension in monitoring forest landscape structure and change. Taiwan Journal of Forest Science, 14 (4), pp. 397-407 [in Chinese with English summary].
Cola, L.D., 1989. Fractal analysis of a classified Landsat scene. Photogrammetric Engineering and Remote Sensing, 55 (5), pp. 601-610.
Dunning, J.B., B.J. Danielson, and H.R. Pulliam, 1992. Ecological processes that affect populations in complex landscape. Oikos, 65, pp. 169-175.
Huang, C.C., 1997. Study on landscape monitoring of forest: an illustration of Huisun Experimental Forest Station [dissertation]. National Chung-Hsing University, Taiwan, 161 p. [in Chinese with English summary].
Lam, NS-N., 1990. Description and measurement of Landsat TM image using fractals. Photogrammetric Engineering and Remote Sensing, 56 (2), pp. 187-195.
Li, H., 1990. Spatio-temporal pattern analysis of managed forest landscape: a simulation approach [dissertation]. Oregon State University, Corvallis, 166 p.
Magurran, A.E., 1988. Ecological diversity and its measurement. Princeton University Press, Princeton, NJ, 179 p.
McGarigal, K., and B.J. Marks, 1994. Spatial pattern analysis program for quantifying landscape structure. Dolores (CO): P.O. Box 606. 67 p.
Olsen, E.R., R.D. Ramsey, and D.S. Winn, 1993. A modified fractal dimension as a measure of landscape diversity. Photogrammetric Engineering and Remote Sensing, 59 (10), pp. 1517-1520.
O'Neill, R.V., J.R. Krummel, R.H. Gardner, G. Sugihara, B. Jackson, D.L. DeAngelis, B.T. Milne, M.G. Turner, B. Zygmunt, S.W. Christenser, V.H. Dale, and R.L. Graham, 1988. Indices of landscape pattern. Landscape Ecol 1, pp. 153-162.
Rex, K.D., and G.P. Malason, 1990. The fractal shape of riparian forest patches. Landscape Ecology, 4 (4), pp. 249-258.
Turner, M.G., 1990. Spatial and temporal analysis of landscape patterns. Landscape Ecol 4, pp. 21-30.
Turner, M.G., and R.H. Gardner, 1991. Quantitative methods in landscape ecology. Ecol Studies. Vol. 82. Springer-Verlag, New York, 536 p.
Wang, S.F., 1998. Application of geographic information system and fractal dimension in forest landscape spatial variation monitoring. [MSc thesis]. National Taiwan University, Taiwan. 73 p. [in Chinese with English summary].

Page 2 of 2
| Previous |