DOI: 10.1007/s00267-002-2471-7
GIS-Based Support Tool System for Decision-Making Regarding Local Forest Protection: Illustrations from Orissa, India MADELENE OSTWALD* Earth Sciences Centre, Physical Geography Go¨teborg University Box 460 405 30 Go¨teborg, Sweden
surrounding forest resources, and population mix. Methods used were digitizing information for the systems’ different layers, analyses of satellite information, field work, gathering of local information, and the application of five risk/priority analyses: erosion, ecological and institutional sustainability, conflict, and degree of dependency.
ABSTRACT / A support tool system comprising risk and priority analyses was illustrated in a geographical information system environment and also tested with data from two forest protection areas for comparison of the system output. The system is recommended as a management monitoring tool for areas where village forest protection at a local level is taking place. The geographical area in the eastern part of India is subject to scarcity of forest resources and is representative in the context of widespread occurrence of local forest protection. Data used were topography, hectares protected, population census, distance to forest and other villages, degree of forest regeneration, presence of plantations, age of protection,
Questions asked were how the different analyses should be interpreted and how the system could be kept updated. The results show that the system needs resource-demanding and field assistance to be kept dynamic. The system is also dependent on the interpretations of the analyses. The limits or levels of assistance for forest management depend on the resources available. The system illustrates how a tool can be utilized for decisions regarding input of resources. It can further be very useful in defining and comparing different areas in order to detect areas in need of assistance and the type of help needed.
In the wake of local forest protection in many countries of the developing world (Wolvekamp 1999), new demands from management and new monitoring mechanisms are needed (Poffenberger and others 1997). The complexity and dynamics in the local process require tools other than simple timber calculations, harvest cycles, and species growth. One such promising tool is a dynamic support tool system that can include parameters of both ecological and social characters with the possibility to analyse risk or priority. Poffenberger and others (1997) have developed a manual geographical information system (GIS) mapping method based on topographical maps and several plastic sheets of information on demarcation, water sources, grazing areas, plantations, and forested area. This method is recommended for local forest rangers working together with forest protection groups. However, the method is not feasible for larger areas (a whole state) or for a larger number of forest protection areas, since the amount of data will be unmanageable.
Today assistance to these local forest protection groups is not systemized or done in an objective manner. This diversity of assistance programs is a consequence of rapid changes within separate protection areas as well as growth and variety of local forest protection in large geographical areas such as Orissa, India. The recognition of local forest protection is also fairly young, which explains the lack of existing systems for decisions based on relevant and dynamic information. On the basis of these facts, this paper suggests a system that produces output that can aid decision making for local forest protection areas. The objective of this paper is to illustrate a computerized support tool system based on GIS. It includes ten different parameters used to produce priority or risk analyses over two test areas, both subject to local forest protection but with slightly different character. The specific questions posed are: How can these risk and priority applications be interpreted? How is the system kept dynamic and up-dated?
KEY WORDS: Support tool system; GIS; Local forest control; Risk/priority analyses; India
Background
*Email:
[email protected]
Environmental Management Vol. 30, No. 1, pp. 35– 45
Local forest protection has, as in this illustration from Orissa, India, developed on a local basis among ©
2002 Springer-Verlag New York Inc.
36
M. Ostwald
people living in or next to decreased forest resources. The forests are mostly state owned and have started to decay due to overexploitation from an increasing population using forest products for daily use and from growing timber for the pulp industry. The protection strategy is the nonuse of woody material, allowing the users to collect only dry and dead wood and nontimber products such as mushrooms, fruits, and climbers. Committees are usually formed and a patrol system is developed among the participants. The total area under local forest protection in Orissa is hard to define, since only the communities that are registered through the Forest Department (through Joint Forest Management) are counted. Poffenberger and others (1996a) state that figures from late 1980s estimated 3000 – 4000 village committees protecting 10% of the forest in Orissa. Orissa Forest Department stated in 1996 that there were 2600 villages involved in the protection of a little bit more than 6% of the forest (Orissa Forest Department 1996). Forest Survey of India (1999) calculated in 1999 that Orissa had 3704 Joint Forest Management villages protecting close to 10% of the forest while the true number of the local protection initiatives remains uncertain. A new focus on forestry matters in tropical and dry tropical forest management is natural regeneration. This concept is moving away from the idea of plantations that have been the main solution to deforestation by actors in ministries and donor positions during the 1970s and 1980s. The potential that lies within natural regeneration has recently been emphasized (e.g., Persson 1997) as being cost effective; it promotes biodiversity and implies a holistic view of land management that includes nontimber forest products and grazing, agriculture, and water management. The regeneration process is also rapid in the tropical forests compared to temperate forests. The prerequisite for regeneration is that the rootstock be left in the forest. The new view of natural regeneration also includes a change of focus from timber towards a more process- or ecology-oriented forestry and its management. There are several differences in the management structure and its effect between the local or participatory forest management and that of conventional management, as is demonstrated in Table 1. To make risk or priority analyses, a decision or choice between alternatives is needed (Eastman and others, 1993). For the suggested system presented here, the decisions are to be made by the forest managers (partly or fully) as to which forest protection areas require an input of resources or help as compared to other areas. The use of GIS in forest management decisions has given benefits to forest organizations. The
Table 1. Differences between conventional and local forest managementa Conventional forest management
Local and participatory forest management
Centralised management Revenue oriented Production motives Single products/species Large working plans Target oriented Controlling people Department bureaucracy
Decentralised management Resource oriented Sustainability Multiple products/species Micro planning Process oriented Facilitating people People’s institutions
a
Modified from Singh and others (1997)
amount of resources available for inputs is one variable limiting the action. The heterogeneity among the local forest protection groups is important. One area is not like another in terms of physical and social features. The needs of input therefore differ from case to case. Variables used for this application are taken from two forest areas under local protection in the Ranapur area in Orissa, India. The two areas are used to illustrate the input required for the system and to illustrate the output from the analyses. The coastal area of Orissa is marginalized as a combination of high population and low forest resources as compared to the interior of Orissa. The two illustrations are also representative in the context of extensive development of local forest protection that has taken place in eastern India during the last few decades. The variables include: topography, size of protected forest area, size of population in the community, distance to natural forest, distance to other villages, degree of regeneration in the protected forest, other forest resources, age of protection, status of surrounding natural forests, and heterogeneity versus homogeneity of the population. The risk or priority applications considered are: erosion risk analysis, ecological sustainablity analysis, institutional sustainability analysis, conflict risk analysis and degree of dependency analysis.
Study Area The location of the study area is characterized by two mountain ridges, belonging to the northeastern part of the Eastern Ghats, and plains with paddy fields (Figure 1). These two ridges are 1–5 km wide, rising approximately 500 m above the plains and are the main areas for natural forest, and also the site for this study. The soils in the area are mainly Ferric Luvisols with some Eutric Nitosols (FAO-UNESCO) or Alfisols (Soil Taxonomy).
Forest Protection in India
Figure 1. Map of study area (top) shows the location of Orissa. (Bottom) Study area indicating villages and forest areas associated with protection: 1, Dhani Hill forest area (Panch Mouza); 2, Jiripada forest area.
Orissa is located in the tropical belt and falls under the influence of the southwest monsoon. Eighty percent of the total rainfall occurs during the monsoon season from the middle of June to the beginning of October. Annual precipitation in the area varies from 1058 to 1565 mm. The maximum and minimum monthly mean temperatures are 30.6 and 23.7°C, respectively with maximum between May and October and minimum in January (Government of Orissa 1988). Of the state’s total area, 36% is classified as forestland and 50% of this is dense forest [crown cover (c.c.) ⬎40%), 37% open forest (c.c. 10%– 40%), and 12% shrub (c.c. ⬍10%) (Orissa Forest Department 1996).
37
The forest in the area is classified as northern tropical dry deciduous forest (Padhi 1994). Common tree species in the region are Madhuca indica, Emblica officinalis, Acacia nilotica, Dalbergia paniculata, Aegle marmelos, Bahunia purpurea, Diospyrus melanoxylon, Bridelia retusa, and Butea monosperma (Kant and others 1991). At the study site, 250 different plant species are estimated to be present (Vasundhara 1997). According to the elders in the area, prior to 1967 the mountain area contained dense diversified forest and a number of wild animals. In a biodiversity inventory made by a local nongovernmental organization, it is stated that with the growing population after independence (1947), forest on the mountain was intensively used for local consumption and for the commercial pulp industry (Vasundhara 1997). The villagers claim that about 30 years ago the forest started to change due to overuse, and by 1982, most of the mountain was a barren hill. Panch Mouza (five villages) and Jiripada are the two areas used to illustrate the system. The two forest areas are adjacent to each other, located on the southern slope of the Eastern Ghats ridge, with the protecting villages on the plain below. Before the protection efforts, most of the forest areas were a barren hill with young saplings. Even though the two forest areas show similarities in vegetation and user history, several significant biophysical differences are found as a consequence of the time and effort of the local population’s forest protection (Ostwald 1998). The sociocultural setting is also different between the two areas. Since 1986 Panch Mouza has been protected by five villages with a total of 1242 people. This area has also gained a lot of attention from researchers (Johansson 1996, Vasundara 1997, Ostwald and Baral 1999, Ostwald 1999) and the forest department (through the signing of Joint Forest Management) and has received a state award for environmental activity (Prakruti Mitra Award). The Jiripada forest area has been protected since 1996 by the single village, Jiripada, with a population of 600, and has received very little outside attention.
Methods The System One principal behind a GIS-based system is to be able to handle large amounts of data, as compared to the small amount that can easily be done with maps and tables on a local basis. Another principal is to be able to put into the system first-order data that can be used in other applications to produce second-order information, i.e., new information. Yet another principal is the power of first-order information converted to second
38
M. Ostwald
order or new information that can be used to develop indices. These indices can finally be used for decisionmaking. As will be illustrated here, a GIS-based system is dynamic in the sense that new or changed information is easy to include (Burrough 1998). The diversity of sources and variation of criteria is also important (Carver 1991). The suggested information used in this illustration is taken from physical, empirical, and monitored data, local knowledge, and official statistics. The Layers The raster-based software IDRISI (Eastman 1997) is used in this GIS work. Due to the user-friendly layout of IDRISI and the not too advanced computer hardware as well as the academic development of the program, users do not require extensive education about the system or great financial resources for software or hardware. The layers that are presented and used here are those partly shown to be of importance in local forest protection in Orissa. The layers used are also available for this illustration. Several other parameters are of importance and are discussed later. However, the tool is presented with these layers to show the strength and potential of a support tool system for this kind of natural resource management. No one layer is given a higher importance or value than any other. The first-order data used in the layers could easily be subject to human subjectivity since the choice for input needs to be conducted by a person with possible bias. This could be shown in areas such as hectares protected or age of protection that already are controversial issues between forest owner (Forest Department) and local protectors in Orissa. However, with a growing number of microplans included in the Joint Forest Management, these issues are documented and therefore official information, which can slightly decrease the personal subjectivity. The basis for the different layers and for the two test areas were images from the System Pour l’Observation de la Terre (SPOT) satellite obtained on 29 December 1994 (228 –309), which had earlier been used for the area (Ostwald 1999, Ostwald and others, 1999). The coordinate system used was UTM (45 north) and the resolution was 20 ⫻ 20 m. The methods of producing the different layers are shown in Figure 2. Topography (layer 1). Topography is important when considering soil erosion, soil cover, and vegetation, which are interrelated since vegetation binds the soil cover, which in turn halts the erosion. Although some soil erosion does take place on flat ground, the effects increase with increasing slope, and this affects forest growth (Paton and others 1995). In this illustration, the
topography was based on a map from a Russian contour survey map produced in 1953 (scale 1:200,000 with intervals of 40 m). The map was digitalized with the software program OCAD (Steinegger Software 1997) and later imported to IDRISI as vector files upon the SPOT-based map. The contours were then used to interpolate the inclination of the area where each pixel would produce a slope value in degrees. This means that for every 20 ⫻ 20-m area, the actual mean degree is representative in this layer. In the case of Orissa Forest Department, the national topographical maps from the Survey of India are available. In this paper, the area is within a restricted area (close to military sensitive coastline), hence the use of external material. Hectares protected (layer 2). The size of the forest affects the available material for those that use the resource. Poffenberger and others (1996b) suggest 0.5 ha of forestland per capita as a minimum, depending on how the resource is used and what the status of the forest is. In the case of local protection, too large an area can be unmanageable for the available population to protect. Based on the SPOT image and demarcation from fieldwork, the total area of the two forest protection areas was calculated by IDRISI. The actual hectares are represented in this layer. Population census (layer 3). When dealing with local forest protection, population is of interest in terms of users of forest resources and as key stakeholders. This implies the degree of need of the resource each person or household has. Information about the population census is well documented on the village level by the Orissa State Census Office, which conducts a survey every 10 years (1981, 1991, etc.). During questionnaire fieldwork (Ostwald and Baral 1999) in 1997, the census data and population issue were discussed. The actual number of people is represented in this layer. Distance to forest (layer 4). The distance between households and the forest resource is important (Ko¨ hlin 1998). If the distance from the household to forest resource is too far, dependency can be placed on other resources. From the SPOT-image-based map, the distance to the forest border was calculated in IDRISI for this study. The actual distance in kilometres is represented in this layer. Distance to other villages (layer 5). The more scarce the resource and the larger the groups interested in it, the more conflict-prone it is. Distances to other villages are therefore important (Little 1994). A similar procedure to the one above was used for distance to other villages. The information on location of other villages was recorded through field visits where a global positioning system (GPS) was used. Map material is available at the Survey of India, but there are several decades between
Forest Protection in India
39
Figure 2. Structure of GIS layers for the support tool system. Panch Mouza is represented to the right and Jiripada to the left indicating the data used for the case study.
surveys. The actual distance in kilometres is represented in this layer. Degree of regeneration (layer 6). In the early stage of protection the forest resource has, in most cases, a very low status, or was, as the villagers explained, a bare hill. In tropical dry forest, regeneration is fairly quick if the rootstock is left. With increased regeneration, risk of erosion is reduced while the potential sources of conflict are increased. The degree of regeneration in this particular case was based on earlier studies over the area (Ostwald 1999). The method used involved the same SPOT image mentioned above and forest field surveys. A total of 61 survey plots of 5 m radius randomly laid out over approximately 800 ha of protected forest were used (for more details see Ostwald 1998). To find the degree of regeneration the method included five physical features: number of trees per plot, crown cover, ground vegetation cover, basal area (i.e.,
the surface of timber), and mean height per plot. Each feature was described as less mature or more mature forest, e.g., ⬍25 trees per plot ⫽ less mature forest, ⬎25 trees per plot ⫽ more mature forest. In this way, each visited plot could be considered less or more matured. This was correlated to indices produced from the satellite image, of which the normalized difference vegetation index (NDVI) gave the highest correlation (78%) with a NDVI limit between less or more mature forest at 0.32. When applied to two different locally protected forest areas, one newly introduced to local protection and one that had been protected for 10 years more, it was possible to see a difference. The 10-year-old protected forest had ⬎55% of “more mature” pixels, while the newly protected forest area had ⬍55%. With this layer, there are two possible outcomes—a high degree of regeneration (1) and a low degree of regeneration (2).
40
M. Ostwald
Plantation (layer 7). There is considerable pressure on natural forest, and plantations have become commonplace in Orissa. The effects of the social forestry plantation scheme from 1985 to 1992, the largest in the state, have been studied by Ko¨ hlin (1998), who showed that the plantations have had the effect of decreasing pressure on the natural forest. An available forest plantation can therefore have an effect on the dependency on natural forest. For Orissa the information regarding plantations is, with villagers and/or the local forest ranger. In this layer, the presence of plantations is represented by 2, and 1 represents areas with no plantation. Age of protection (layer 8). As discussed above, the longer a forest has been successfully protected, the more valuable it will become due to regeneration of the resource (Ostwald 1998). The age of protection is therefore important with regard to conflicts. The actual years of protection are represented in this layer. Surrounding forest resources (layer 9). Earlier studies indicate that local forest protection only shifts the pressure on resources to another forest area in the vicinity, i.e., the overall forest status is in a status quo situation (Kant and others 1991, Jonsson and Rai 1994). A study made in this particular area (approx. 35 ⫻ 25 km) based on Normalized Difference Vegetation Index (NDVI) produced by National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) satellite data from 1992 to 1996, indicated that the forest vegetation in general was increasing. This also included locally protected areas of natural forest (Ostwald and others 1999). Even so, the surrounding forest resource does have an impact on dependency on the forest area locally protected. One of the NOAA images (December 1995) was used to estimate the vegetation resource of the surrounding forest. The whole forest area (35 ⫻ 25 km) was investigated to find minimum and maximum vegetation values, which were then divided into three classes—low, medium, and high. All pixels within 2 km from each locally protected border were used to calculate a mean, which indicated whether the surrounding forest had a low, medium, or high vegetation value. A low vegetation value is represented in this layer by 1, medium by 2, and high by 3. Population mix (layer 10). There have been discussions regarding the institutional sustainability and risk of conflict depending on the caste mix of a protection group, where more homogeneous caste groups have been seen as more stable and less prone to conflicts. Tribal groups especially tend to have a dependency on forest resources (Chatterjee and Roy 1995, Sarin 1996,
Yadama and others 1997). Mixed populations include two of the main population groups: general castes, scheduled caste, or scheduled tribes. Mixed populations are represented by 1; homogeneous populations are represented by 2. The Analyses No one analysis is more important or is given a higher weight than any other analysis. The system and the outcome of the analyses are supposed to be flexible. The need or aim of each specific forest area or district is therefore the variable that will give importance or weight to the analyses. Erosion risk analysis. The purpose of seeking low erosion risk is to minimize possible degradation. Factors controlling the effects of erosion are, according to the universal soil loss equation (USLE) rainfall, soil erodibility, slope length, slope gradient, vegetation cover, and management (Wischmeier and Smith 1978). In this particular case the protected forest is situated on a slope with fields and houses below. The slope has, in the past, been bare of trees, which resulted in low permeability of heavy rainfall during the monsoon followed by sheet erosion, which sand-casted the fields below. Since the forest has regenerated, rainfall has started to infiltrate into the forest soil, and this has led to the development of several streams leading the water down the slope and through the fields (Ostwald and Baral 1998). The two parameters used for this illustration and for the scope of this study are topography (layer 1) and degree of regeneration (layer 6), but could include the other parameters in USLE. This analysis could also use a fuzzy-set theory, giving a less rigid result that included some uncertainty (Charnpratheep and others 1997, Mitra and others 1998). However, for the scope of this study the use of actual slope degree varying from 0 to 44, or binary logic, works to illustrate the tool. Ecological sustainability analysis. In its broadest sense, ecological sustainability implies the possibility of a natural resource being sustained while being used by the inhabitants. In this case it is the amount of forest area used by the people protecting it. Based on calculations from empirical data, most protected forest areas in Orissa have a ratio of 0.1–3 ha forest area per household of six persons (Scandiaconsult Natura AB/Asia Forest Network 1998), whereas Poffenberger and others (1996b) suggest 0.5 ha per capita. The ratio is of course dependent on the status of the forest; a more developed forest can support more people than a forest that has just started to regenerate. This means that with growing population or decreasing forest areas ecological sustainability is at risk, since there is the possibility
41
Forest Protection in India
of overuse of the resource, leading to its degradation and resulting in insufficiency for the users. When this happens, managers should invest more effort in protection. This could include changes of sources (e.g., plantations, stoves) or investigations of possible extended forest protection areas. When resources become scarce, there is a higher risk of conflict (discussed below). The parameters used in this illustration are the hectares protected (layer 2) divided by the population census (layer 3). Institutional sustainability analysis. There is a broad variety of local institutions associated with local protection of forest. This depends on sort of inputs for initiating protection, how it is performed, how far away the resource is from the users, how old the protection is (i.e., how far the regeneration has gone) or how homogeneous the protection group is. Several institutional characteristics have been identified as needed in the success of protection; high forest dependency, perception of resource scarcity, geographical proximity to forest, informal of formal rights, presence of indigenous resource management institutions, traditional socioreligious forest values and strong local leadership (Sarin 1996). The parameters used to illustrate this analysis are distance to other villages (layer 5), age of protection (layer 8), and population mix (layer 10). Conflict risk analysis. Conflicts are problematic in forest management, and few managers receive sufficient training in working with interested parties to resolve disputes or feel confident to do so. It is therefore important to tackle the problem early before it gets too big (Higman and others 1999). Conflicts can be internal or external and are dependent on the interest groups, e.g., their power, number, and degree of interest (Little 1994). The parameters used in this analysis are hectares of protected forest (layer 2) and population census (layer 3) (i.e., the result from the ecological sustainability analysis), age of protection (layer 8), and population mix (layer 10). Degree of dependency analysis. Dependency is an important variable for success of the institutional sustainability (Sarin 1996). It has been shown in questionnaire studies from local protection groups that if dependency is low due to high income level, the interest in forest protection issues are low (Ostwald and Baral 1999). Dependency should therefore be considered an important variable for success of local forest protection management. The parameters used in this analysis are distance to forest (layer 4), presence of plantations (layer 7), and surrounding forest resources (layer 9).
Table 2.
Result of the ten different illustrative layers
Topography (degree) Hectares protected (ha) Population census (persons) Distance to forest (km) Distance to other villages (km) Degree of regeneration (high-1, low-2) Plantation (yes-2, no-1) Age of protection (years) Surrounding forest resources (1-low, 2-med., 3-high) Population mix (mixed-1, homogeneous-2)
Panch Mouza
Jiripada
0–48 451 1242 2 3 1
0–44 131 600 1 3 2
2 13 2
1 3 2
1
1
Results The Layers Table 2 presents the results from the applications of the ten layers to the two areas using the proposed methods. This first-order data can be analyzed to produce second-order information, which can be the basis for decision-making. No single layer is more important or should be given a higher weight than any other. The Analyses Erosion risk analysis. For the erosion risk analysis, the topography/slope (layer 1) and degree of regeneration (layer 6) are multiplied. The result is shown in Figure 3, together with an erosion risk map and histograms for the two forest areas. Since Jiripada had a regeneration level of 2, indicating low regeneration, the slope increased to double. Panch Mouza’s slope is given in actual inclination, since the high regeneration level gave a value of 1. According to Young (1972), 15 degrees is the limit between moderately and strong sloping. Thus, if 15 degrees is stated as a limit for erosion risk for this illustration, 15 degrees is the actual risk limit for Panch Mouza, while 7.5 degrees slope is the actual risk for Jiripada. This result shows that Jiripada has 66% of its area at risk while Panch Mouza has 42%. The result makes it possible to evaluate distinct parts of a protected forest area. Ecological sustainability analysis. For the ecological sustainability analysis, the area in hectares protected (layer 2) was divided by population census (layer 3). The result is shown in Table 3; Panch Mouza and Jiripada have 0.4 and 0.5 ha/person, respectively. This means that seen in a larger temporal scale, Panch Mouza will be less sustainable than Jiripada. Seen in the Orissa context of an average of 0.1–3 ha of forest per
42
M. Ostwald
Figure 3. Erosion risk analysis expressed in a map and histograms for the two protected forest areas. Pixel value 15 is used as the limit between risk and non-risk.
Table 3.
Result from ecological and institutional sustainability, conflict risk and degree of dependency
Panch Mouza Jiripada a
Institutional sustainabilityb
Conflict riskc
Dependencyd
0.4 0.5
17 7
5.4 6.5
2 2
Ecological sustainability expressed in available hectare of forest per person.
b c
Ecological sustainabilitya
Institutional sustainability expressed by a relative index indicating need of assistance for low indices.
Conflict risk expressed by a relative index indicating need of assistance for low indices.
d
Degree of dependency expressed by a relative index indicating need of assistance for low indices.
household, none of the areas have a severe level of risk. However, according to Poffenberger and others (1996b) the minimum requirement for forest-dependent populations is 0.5 ha/person. According to this criterion, both the areas have borderline risk. The value should be interpreted as low value– high risk or need of input. Institutional sustainability analysis. For the institutional sustainability analysis, the distance to other villages (layer 5), age of protection (layer 8), and population mix (layer 10) are used and added together. The result is shown in Table 3. The result of the analysis is the index that expresses a relative relationship where areas with low values are in greater need of assistance than those with high values. Since both distance to other villages and population mix is the same for the two areas, the age of the protection is the variable that
gives Panch Mouza a better possibility of sustaining institutionally according to the index. Conflict risk analysis. For the conflict risk analysis, the value of ecological sustainability (layer 2/3) of hectares per person, distance to other villages (layer 5), degree of regeneration (layer 6), and population mix (layer 10) were added together. The result is shown in Table 3. The index expresses a relative relationship where forest areas with low values are in greater need of assistance than those with high values. The two areas are given a very similar index. The conflict risk factor for Panch Mouza is its stage of regeneration, i.e., the better the forest, the higher the conflict risk. This is both internal due to distribution of resources among different groups within the protection group as a whole, and external due to increased value placed upon saw mills, state owners, other villages, etc.. Panch
Forest Protection in India
Mouza also has fewer hectares per person which, due to greater scarcity of resources, causes tension. Degree of dependency analysis. For the degree of dependency analysis, the presence of plantation (layer 7) was added to the status of the surrounding forest (layer 9) and then subtracted by the distance to forest (layer 4). The result is shown in Table 3. The index expresses a relative relationship where forest areas or villages with low values are in greater need of assistance than those with high values. Both areas are given an index of 2 and have the same status of surrounding forest. Panch Mouza has a plantation and is 2 km from the forest compared to Jiripada that has no plantation and is only 1 km from the forest. The results from the five analyses show that a forest manager wanting to invest in local protection areas would have to invest in both areas but focus on different issues in the two areas. In Dhani Hill the focus would be on ecological sustainability and conflict risk, while Jiripada would require help with institutional sustainability and erosion risk. Dependency is, in this result, equal for the two areas.
Discussion Within a state or even a district, the nearby environment may vary, giving these proposed parameters different importance. Is it feasible, for example, that distance to forest stated per km (1, 2, . . . ) is weighted equally with the presence of a plantation (no ⫽ 1, yes ⫽ 2) as in the degree of dependency analysis if used over a large area? If the situation shows that one parameter is of greater importance than another, the weighted parameters should be altered in the system. For instance, if the distance to the natural forest is of higher importance than the presence of a plantation (discussed by Ko¨ hlin 1998) as regards dependency, the distance should be increased by a factor 2 or 3. This might be applicable in areas with high availability of natural forest, while areas with low availability of natural forest show greater importance in the presence of plantation. Another example could be the increased importance of the degree of regeneration. In this case, the low regeneration forest would be given a higher figure, maybe 3 or 4, while the high regeneration forest would be left at 1. This means that the erosion risk analysis would give much higher and more prioritized values to the area with low regeneration. In the case of conflict risk analysis, the higher weighted regeneration would give the area of high regeneration a more prioritized value, more so than the other layers included in the analysis. This way the support tool system can be
43
altered to suit different situations. The dynamic and flexible variables are key issues of this system. The actual values of the analyses, as seen in Table 3, are not comparable in the sense that institutional sustainability, with results of 17 and 7 for the two areas, is more important than ecological sustainability with the results 0.4 and 0.5. In these two illustrations, the relative values from an analysis are not well expressed. However, if applied to 1000 locally protected forest areas, it should be possible to see which areas are most vulnerable. The system could be developed by standardization of the analytical values or a normalization of the output. The question is also which of the five analyses are of greatest importance: erosion risk, ecological sustainability, institutional sustainability, conflict risk, or degree of dependency. One way of dealing with this is to see which locally protected areas are subject to high risk/priority in a majority of the analyses. As illustrated in this paper, the areas showed weaknesses in different indices, two each for the two areas and one with the same index. Preferably the system should develop so that it would be possible to distinguish “slight” need of help from “severe” need of help. Regardless of whether forests are managed by local of national forestry departments or some other body, the available resources are the parameters that must be incorporated in the management system. If resources such as money or manpower were unlimited, assistance would not have to be prioritized. This is seldom the case in natural resource management. Hence, the manager’s resources are the factor that will decide when a local forest management area is at risk and requires intervention and assistance. The system optimizes the use of the available resources. Therefore, this study does not specify at what level an analysis result shows an area at risk. Rather, this system can compare different areas objectively with regards to output. Presented here are ten parameters that have been studied in this particular area of India. Other possible variables (layers) for erosion risk analysis could be the other parameters included in USLE, the soil distribution along the catena, and climatic variables regarding onset and volume of precipitation and temperature. Agricultural activity, local hydrology, and soil could be important parameters for ecological sustainability. Economic parameters could be included in the institutional sustainability, including the economic situation in a protecting village or the relationship to markets for fuel material and nontimber forest products. With respect to finances, the presence of economically valuable trees can also be indicated. This could also be important for conflict risk analysis as well as the legal aspect of local protection, such as a signed and work-
44
M. Ostwald
able Joint Forest Management agreement with the Forest Department, users, and owners rights over forest resources, or legal representation at higher levels. The isolation of an area can also be important in terms of building networks, indicating the power of forest users as an interest group, and could be included in the degree of dependency analysis. Other analyses can be developed, if required, based on usage, aesthetics, culture, religion, or survival. As in Orissa, which was hit by a huge cyclone in October 1999, it could be important to indicate whether an area has been affected by a natural disaster and what effect this has had on the forest, the soil, and the people. Since this paper deals with a system that is not static in its features and where several different parameters are used, the use of fuzzy set application would be desirable. This is one way to deal with the level of uncertainty inherent in the problem (Ghosal 1990). This area should be pursued. The question raised at the beginning of the paper was how this kind of system can be kept dynamic, following the fast and varied changes within local forest protection action. Of the parameters suggested here, three (topography, distance to forest, and other villages) are more or less static, with little need for updating. A population census can, in most places, be updated with national- or state-based survey reports, conducted, as in Orissa’s case, every decade. Local forest personal can collect census information from villages, if needed, in case of drastic and sudden changes. The method of defining the surrounding forest resources can be done off-site with no knowledge of the local situation needed once demarcation of the protected area is completed. This can be done over large geographical areas. Local direct updating has to be done (initially and in case of possible changes such as census information after natural disasters) in the case of hectares protected, degree of regeneration, knowledge of presence of plantation, and age of plantation.
Conclusions By structuring various information data in layers using GIS, an easy, dynamic, and objective tool is provided. In combination with decreasing costs for software and computers, a GIS-based support tool system is a welcome tool for forest managers in areas where local forest protection is taking place. The need for careful local field work and respectful assistance has to be emphasized for the system to be useful and dynamic. Further development of the normalization or interpretation of the indices would strengthen the system even more. However, due to the growth of different forms of
local forest management, the diversity of these local situations, their geographical range, and the limited resources of managers or state owners, the system can be a useful way to improve the decision-making for providing assistance.
Acknowledgments I wish to thank the people of Panch Mouza and Jiripada for their information, assistance and help given during my field work. I also want to thank Maria Lindqvist for digitalizing the topography, Solveig Svensson for the drawings, Kamaruzaman Jusoff, Department of Forestry, University Putra Malaysia in Malaysia, and Marie Eriksson, Sven Lindqvist, Margit Werner, Deliang Chen, and Mira Ovuka, Department of Earth Sciences, Go¨ teborg University in Sweden, for valuable comments and discussions during this work. The anonymous reviewers are also thanked for their constructive comments on the manuscript.
Literature Cited Burrough, P. A. 1998. Principles of geographical information systems. Oxford University Press, New York, 333 pp. Carver, S. J. 1991. Integrating multi-criteria evaluation with geographical information systems. International Journal of Geographical Information System 5:2321–2339. Charnpratheep, K., Qiming-Zhou and B. Garner. 1997. Preliminary landfill site screening using fuzzy geographical information systems. Waste Management Research 15:197–215. Chatterjee, M., and S. B. Roy. 1995. Reflections on gender issues in Joint Forest Management. Inter-India Publication, New Delhi, 203 pp. Eastman, J. R. 1997. IDRISI for Windows v. 2.0. Clark University, Graduate School of Geography, Worcester, Massachusetts, 315 pp. Eastman, J. R., P. A. K. Kyem, J. Toledano, and W. Jin. 1993. GIS and decision making. United Nations Institute for Training and Research, UNITAR Geneva, 112 pp. Forest Survey of India. 1999. State of forest report 1999 — summary. Ministry of Environment and Forests, Dehera Dun, 18 pp. Ghosal, A. 1990. Cybernetic modelling of the ecosystem. Pages 21–29 in A.K. Biswas, T.N. Khoshoo and A. Khosla (eds.), Environmental Modelling for Developing Countries. Tycooly Publishing, London. Government of Orissa. 1988. Climatological data of Orissa 1977–1986. Bhubaneswar, Directorate of Economics and Statistics, 131 pp. Higman, S., S. Bass, N. Judd, J. Meyers, and R. Nussbaum. 1999. The sustainable forestry handbook. Earthscan Publications, London, 289 pp. Johansson, M. 1996. Without the trees—An economic study of the value of informal protection of the forest in Orissa,
Forest Protection in India
India. Bachelor thesis. Unit for Environmental Economics, Department of Economics, Go¨ teborg University, 88 pp. Jonsson, S., and A. Rai. 1994. Forest, people and protection. ISO/Swedforest, SIDA, Bhubaneswar, 54 pp. Kant, S., N. M. Singh, and K. K. Singh. 1991. Community based forest management systems. Bhubaneswar, ISO/ Swedforest, SIDA, 70 pp. Ko¨ hlin, G. 1998. The value of social forestry in Orissa, India. Department of Economy 83, Go¨ teborg University, 133 pp. Little, P. D. 1994. The links between local participation and improved conservation: A review of issues and experiences. Pages 347–372 in D. Western and R.M. Wright (eds.), Natural connections—Perspectives in community-based conservation. Island Press, Washington, DC. Mitra, B., H. D. Scott, J. C. Dixon, and J. M. McKimmey. 1998. Applications of fuzzy logic to the prediction of soil erosion in a large watershed. Geoderma 86:183–209. Orissa Forest Department. 1996. Capacity building for participatory and sustainable management of degraded forests in Orissa. Project document, 10 pp. Ostwald, M. 1999. Estimating degrees of regeneration in tropical dry forest through ground truth data, SPOT and GIS. Applied Geography (Submitted). Ostwald, M., and R. Baral. 1998. Perception, knowledge and attitudes towards forest among people involved in forest protection on Dhani Hill, Orissa, India. Field report. Physical Geography, Go¨ teborg University, 6 pp.
45
Persson, R. 1997. Det skogliga biståandet— utveckling och na¨ rvarande omfattning som fo¨ rva¨ ntad utveckling och planerad fo¨ ra¨ ndringar. Sida, Stockholm. Poffenberger, M., P. Bhattacharya, A. Khare, A. Rai, S. B. Roy, N. Singh, and K. Singh. 1996a. Grassroots forest protection: Eastern Indian experiences. Asia Forest Network Report 7. Poffenberger, M., B. McGean, and A. Khare. 1996b. Communities sustaining India’s forests in the twenty-first century. Pages 17– 48 in M. Poffenberger and B. McGean (eds.) Village voices, forest choices, joint forest management in India. Oxford University Press, New Delhi. Poffenberger, M., A. P. Chatterji, and A. Schwartz. 1997. Field methods manual. Asia Forest Network, 46 pp. Sarin, M. 1996. From conflict to collaboration: Institutional issues in community management. Pages 165–209 in M. Poffenberger and B. McGean (eds.), Village voices, forest choices—joint forest management in India. Oxford University Press, New Delhi. Scandiaconsult AB/Asia Forest Network. 1998. Capacity building for participatory management of degraded forests in Orissa, India—study report. Orissa, 30 pp. Singh, S., A. Datta, A. Bakshi, A. Khare, S. Saigal, and N. Kapoor. 1997. Participatory forest management in West Bengal. World Wide Fund of Nature, India, 42 pp. Steinegger, Software. 1997. OCAD for Windows 95 v. 6.0. Baar, Schweiz, 54 pp.
Ostwald, M., and R. Baral. 1999. Local forest and protection, gender and caste—Dhani Hill, Orissa, India. Geografiska Annaler Serie B 82:115–128.
Vasundhara. 1997. Conserving bio-diversity—A decade’s experience of Dhani Panch Mouja people. Vasundhara, Bhubaneswar, 37 pp.
Ostwald, M. 1998. Effects on soil and trees by local protection in tropical dry natural forests Orissa, India. Proceedings FORESEA Miyazaki, Japan. 566 –574.
Wischmeier, W. H., and D. D. Smith. 1978. Predicting rainfall erosion losses. A guide to conservation planning. Agriculture Handbook No 537. USDA, Washington DC.
Ostwald, M., K. Jusoff, and S. Lindqvist. 1999. Investigating local protection efforts in forest vegetation change in Orissa, India using NOAA AVHRR data. Journal of Tropical Forest Science 12:778 –793.
Wolvekamp, P. 1999. Forest for the future—Local strategies for forest protection, economic welfare and social justice. Zed Books, London, 270 pp.
Padhi, G. S. 1994. Forest resources of Orissa. Bhubaneswar, BITS-Pilani, 294 pp. Paton, T. R., G. S. Humphreys, and P. B. Mitchell. 1995. Soils a new global view. Yale University Press, New Haven, 213 pp.
Yadama, G. N., B. R. Pragada, and R. R. Pragada. 1997. Forest dependent survival strategies of tribal women: Implications for joint forest management in Andhra Pradesh, India. FAO Rap publication 24, 25 pp. Young, A. 1972. Slopes. Longman Group, London, 288 pp.
