0469-B4

Investigation of Erenler Forest Ecosystem in Ankara-Nallıhan (A3) Region of Turkey

Hakan Mete DOGAN ¹ , Musa DOGAN ²

Abstract

This study addressed to investigate some of the forest ecosystem characteristics for some part of Ankara-Nallihan (A3) region of Turkey by using geographic information systems (GIS) and remote sensing (RS) tools. The aims of this study were determined as; (1) visualizing and analyzing plant species diversity and species distribution, (2) explore relationships among species distribution, geomorphology (geology & topography), soil and aspect, and (3) modeling detected relationships. A detailed (1 / 25000 scale) GIS database that includes the latest forest-stand, geological, topographic, and soil maps were created and used. Moreover, plant samples were collected from the pre-determined transect locations by using Global Positioning System (GPS). Shannon Wiener Index was employed to measure species diversity of each transect. Finally, we took the advantage of a LANDSAT-TM image to classify the area according to our aims. Only preliminary results for the years 2001 and 2002 were summarized in this paper.

Introduction

The Earth's genes, species and ecosystems are the consequences of over 3000 million years of evolution. The measure of the variation in genes, species and ecosystems (biological diversity) is valuable because future practical values are unpredictable, because variety is inherently interesting and more attractive, and because our understanding of ecosystems is insufficient to be certain of the impact of removing any component (McNeely 1994). In this respect, wise utilization of forest resources is very crucial.

Recently, sustainable forest management or ecologically sustainable development gained great importance at international level. At this point, relevant, consistent, reliable, authoritative, and internationally compatible statistics and information emerged as one of the basic tools. It is clear that planners, foresters, environmentalists, politicians, and in general all decision-makers can reach constructive decisions only through these statistics. This is the first and important step to conserve biological diversity, flora, and fauna. The impacts of land clearing for agriculture; tourism and settlement; illegal cuttings; forest fires; insects; diseases; droughts; frosts, and construction and exploitation activities on forest ecosystems can only be understood by these statistics and inventories.

A broad capacity-building effort is urgently needed so that countries can monitor forests (Anon. 1992). Governments and institutions should establish and strengthen national assessment and observation systems for forests, forest resources and forest programs. This will require new data systems and statistical modeling, remote sensing, ground surveys and other technological innovations such as geographical information systems.

Geographic data that contain the measurements of three-dimensional space and time have gained great importance for the studies of spatial and temporal relationships in landscapes. Scientific visualization supports the analysis and communication of these kinds of data (Brown et al. 1995). Recent technological changes involving computer cartography and computer graphics has resulted in modern cartographic visualization that is different in both qualitative and quantitative ways. In quantitative terms, a wide range of different cartographic products can be produced much faster and much more cheaply. In qualitative terms, interaction with visual displays greatly increase comprehension in a wide variety of subject areas (Taylor 1994).

GIS, the application of computer graphic visualization in 3-dimensions, is a tool that expands our understanding on a variety of spatial relationships by graphically visualizing all those spatial data (Watson 1992; Habb 1995). GIS needs for visualization contain technological, conceptual, and evaluatory solutions that can be seen in three broad domains namely analysis, illustration, and decision-making (Buttenfield and Ganter 1990). Research tools such as maps, 3-D surface plots and other visual presentations expand our perspective by selectively displaying an array of spatial information for analysis (Watson 1992) and promote awareness through symbolization (Mayoraz et al. 1992).

Both remote sensing and GIS applications have been employed in many disciplines. The discipline areas and the spectrum of applications have been developed and extended for the past 30 years. Moreover, studies have evolved from the basic database establishment to spatial analysis and modeling. The contents of applications also improved from the basic land cover characteristics to species level by the integration of GIS and Remote Sensing.

Materials and methods

The study area, located in 70 km north-west of the Ankara, is called Erenler forest region, and belongs to Nallihan Forest Management District. Erenler forest region contains the forest management series that are called Sarıcal Mountain, Erenler and Kavacik, and these three forest management series covers totally 17 453 hectare (174 530 000 m ² ) area. The elevation changes between 750 and 1500 meters in the area (Figure 1).

Research area consists of ten geological formations (Figure 2). Geological formations are (1) cherty limestone, (2) alluvium, (3) sandstone-mudstone-limestone, (4) muddy limestone, (5) limestone, (6) conglomerate-sandstone-mudstone, (7) andesite-basalt-pyroclastic (volcanic) rocks, (8) andesite-basalt-pyroclastic (volcanic) rocks, (9) alluvial fan, and (10) old alluvium. Volcanic rock classes repeated two times because their ages are different.

Basically, Arc/Info software was used to digitize the paper maps, to correct the digitizing errors, to create topology, to make map projection transformations, to label geographic objects or entities, to determine the attributes, and finally to establish a reasonable GIS database. Some of the database were produced in Arc/View by evaluating the point data that consists of locations of transects (X,Y, and Z format). All produced maps were converted to grid coverages by using a standard grid size (30 x 30 m) in this study. The purpose of this conversion was the preparation of the maps for spatial analysis. The generated polygon and grid coverages were used to visualize, understand, and analyze the distribution of geographic objects and facts by using Arc/View software (especially spatial analysis and grid extensions).

Figure 1. The topography of the study area

Spatial analyses were conducted in three main groups. These analyses can be summarized as (1) analyzing surfaces (histograms by zones), (2) analyzing relationships (correlation), (3) formulating the important relationships (regression). Plant samples including tree, shrub, and herbs were collected from the previously determined transect locations. Transect locations had been selected by analyzing the available GIS database before starting the field studies. Previously determined transect areas were located by using global positioning system (GPS) device in the field. The cover area of species and their numbers were recorded for each transect. International Classification Classes (Unesco 1973) were used in field sampling studies. We also recorded (1) percent area covered by plant species, (2) Braun Blanquet and Sociability of Braun-Blanquet values of species in each transect (Barbour et al. 1987). We defined species diversity according to Shannon-Wiener index on the basis of two factors: species richness, and species evenness. According to this index; the number of species in the community is called species richness, while the relative abundance of species is described as species evenness. This commonly applied measure of species diversity is formulated below (Molles 1999).

Figure 2. The geological formations of the study area.

Correlation analysis provides us an objective, quantitative means to measure the association between a pair of spatial variables. Both the direction and strength of association between the two variables have been determined statistically (McGrew and Monroe 1993). Correlation analyzes among all variables were determined by using both Arc/View and Systat (statistical) software in this study. First, all grid coverages were converted to X, Y, and Z format by using the grid functions in Arc/View, then all these X, Y, and Z values are checked in Microsoft-Excel and aggregated in a single file. Prepared single file was imported to the Systat software and a correlation analysis (Pearson) was conducted. Regression analyses have been applied to the two variables that their correlation was found higher comparing the others.

Finally, we took the advantage of a LANDSAT-TM image to classify the area according to our aims. The LANDSAT-TM image that belongs to 21 August 2000 (Path/Row:178/32) has been obtained as raw image. We used histogram equalization method for clearance of the image. After getting a clear image, geometric corrections that comprised the process of being the satellite image fit on the real world coordinates have been proceeded. Geometric corrections have been arranged in two stages. At the first stage, the image has been rectified and fit into the city boundaries at Turkey database. Second, ground control points, determined by GPS during the field works, were used in order to get a precise rectification. Both supervised and unsupervised classification methods were used. We also produced Normalized Difference Vegetative Index (NDVI) maps by using the image. All classified images have still been evaluated to develop a model.

Results

When ve analyzed surfaces, the results of `histogram by zone' supplied valuable information between different variation of variable pairs. Moreover, they give an idea how geomorphology effects land cover and species distribution. Strong correlations have been detected between the variables such as; species composition and geomorphology. We only give the results supervised classification here, because we still conduct the modeling studies. We identified five basic land classes of the area (Figure 3). Those can be summarized as forest, agriculture, rangeland, fallow/barren, and degraded forest. After getting main classes we investigated each classes independently by using all available database.

Forest class mainly consist of black pine (Pinus nigra Arn.). Some juniperus species such as Juniperus communis L., Juniperus excelsa Bieb., Juniperus oxycedrus L. subsp. oxycedrus constituted other important forest species in the area. Quercus species (Quercus cerris L. and Quercus pubescens Willd.) have been found scattered throughout the area and their stands are mostly sparse.

Figure 3. Supervised Classification of LANDSAT-TM Image

We excluded the areas that were stated as agriculture, fallow/barren, and degraded forest, and focused on the forest and rangeland only. We determined 236 species belong to 45 family in the study area. Number of species recognized in each family stated in Table 1. According to the Table 1; Leguminosae, Compositae, Labiatae, Rosaceae, and Cruciferae families have more species comparing the others. Although there was only one species determined for the families such as Chenopodiaceae this species was scattered throughout the area, and it was collected repeatedly from the transects. Therefore, one species in a family does not mean that it is found in one place only.

Table 1. Number of Species Recognized in Each Family

Conclusion and discussion

The developed database and conducted analyses supplied valuable information. Produced thematic maps showed variations of plant characteristics very well throughout the area. These visual tools make enable us to understand the nature and characteristics of land cover and species distributions. Not only the visualization but also spatial analyses gave us a chance to evaluate the created database in a sound way. Histogram by zones function of Arc/View, for example, depicted some of the hidden relationships between the variables incredibly. Moreover, grid functions of Arc/View make possible many GIS operations easy. Some of the functions of GIS operations such as; creating surfaces, conversion of polygon themes to grid themes, and transformation of grid themes into X-Y-Z format makes possible to conduct further statistical analyzes. We did not put those results here because, they could be meaningless without completing the model that we are still working on it.

Conducted correlation and regression analyses indicated the relationships between variables, and formulated them. Although, results showed that some correlation coefficient values are low, strong correlation (0.87 Pearson correlation) detected between species richness and geomorphology. Modeling studies are still proceeding and we believe in that a suitable model will be developed very soon.

References

1- Anonymous, 1992. Agenda 21, The UN Conference on Environment and Development (UNCED). Rio De Janerio, http://www.igc.apc.org/habitat/agenda21/index.html

2- Barbour, M. G., J. H. Burk, W. D. Pitts. 1987. "Terrestrial Plant Ecology." Benjamin Cumming Publishing Company, Inc. California

3- Brown, W. M., M. Astley, T. Baker, H. Mitasova. 1995. GRASS as an Integrated GIS and Visualization System for Spatio- Temporal Modeling. Twelfth International Symposium on Computer- Assisted Cartography, Charlotte, North Carolina Volume: 4 pp. 89-99.

4- Buttenfield, B. P. and J. H. Ganter. 1990. AVisualization and GIS: What Should We See? What Might We Miss? In Proceedings of the 4th International Symposium on Spatial Data Handling, K. Brassel and H. Kishimoto, Zurich, Switzerland, vol 1, no:2 pp. 307-316

5- Habb, F. J. 1995. Computer Generation and Rendering of Terrain. Master`s of Science Thesis, University of Neveda, Las Vegas, Nv.

6- Mayoraz, R., C.E. Mann, and A. Parriaux. 1992. Three-Modeling of Complex Geologic Structures; New Development Tools for Creating 3-D Volumes. Computer Modeling of Geographic Surfaces. AAPG Computer Applications in Geology. 1. American Association of Petrolium Geologist. Tulsa, Ok. pp. 261-269.

7- McGrew, J. C. Jr., C. B. Monroe. 1993. Statistical Problem solving in geography. Wm. C. Brown Publishers, Dubuque.

8- McNeely, Jeffrey A. 1994. Critical Issues in the Implementation of the Convention on Biological Diversity. In Widening Perspectives on Biodiversity, IUCN, The World Conservation Union, International Academy of the Environment, Geneva Switzerland.

9- Molles, M. C. 1999. Ecology, Concepts and Applications. WCB/McGraw-Hill, Boston

10- Taylor, D. R. F. 1994. Cartographic Visualization and Spatial Data Handling. Sixth International Symposium on Spatial Data Handling, Edinburgh, Scotland, Vol:1, No:2, pp 16-28.

11- Unesco, 1973. International Classification and Mapping of Vegetation. Ecology and Conservation 6. Unesco, Paris

12- Watson, D.F. 1992. Contouring. A Guide to the Analysis and Display of Spatial Data. Pergamon Press. Oxford, England.

¹ Head of Geographic Information Systems and Remote Sensing Department in General Directorate of Agricultural Research<
Address: Tunus Cad. No: 15/2, Kavaklidere 06680 Ankara/TURKEY.
Telephone: (+90 312) 327 01 50,
e-mail: [email protected],
Web address: www.tagem.gov.tr

² Professor in Middle East Technical University
Address: Middle East Technical University, Biology Department, Ankara/TURKEY.
Telephone: (+90 312) 210 51 58,
e-mail: [email protected]
Web address: www.darwin.bio.metu.edu.tr