Assessment of Restoration Potential in the Catchments of Pavana, Chaskaman and Dimbhe Dams in Northern Western Ghats

In the years 2014-15, Global Forest Watch, under its Small Grants Program provided an opportunity to the Ecological Society, Pune, India to perform ecological surveys and assess the restoration potential in Northern Western Ghats at select dam catchments Pavana, Chaskaman and Dimbhe (Ecological Society, 2014). These catchments are characterised by changes in land use after the construction of dams. The reasons for land use changes are as follows : 1. Ring-roads built around the dam catchment, resulting in housing development for urbanites 2. Displacement of villages on the valley floor due Assessment of Restoration Potential in the Catchments of Pavana, Chaskaman and Dimbhe Dams in Northern Western Ghats


36
JOURNAL OF ECOLOGICAL SOCIETY 2018 major rivers like Krishna, Bhima, and Godavari. Hence water security of Maharashtra and other states like Karnataka, Andhra Pradesh and Telangana depends on the Western Ghats. Due to the above reasons and their criticality, largescale conservation and ecological restoration of Northern Western Ghats has become absolutely necessary.

Objectives
This article documents the results of our study of natural vegetation cover, land use and ecological status of the catchment areas of three dams in the Western Ghats. The methodology employed was to classify the natural vegetation cover and assess the human use of land. The ecological status of the classified natural vegetation cover areas was evaluated to ascertain their restoration potential. The study area was the catchment areas of Pavana, Chaskaman and Dimbhe dams which lie on the eastern side of the Western Ghats, in Pune district.
The "Western Ghats" region is not only globally important for its biodiversity but locally it harbours very special biodiversity. However now it is developing fast and large areas within the region are being converted to second homes and farm houses. Modern poly-house agriculture is also practiced to a large extent. All these changes cause degradation of natural vegetation. As major rivers Krishna, Koyana, Bhima, etc. originate in Western Ghats, it is important to conserve their source region for water security. Restoration of the forests and other natural vegetation types holds an enormous potential for the revival of these rivers. Importance of restoration in the conservation of these areas is explained by Prof. Prakash Gole. In his book, 'The Restoration of Nature', he has highlighted why ecological restoration should receive a strategic importance : "Can natural eco-systems ever be restored to their original form and functions? The answer appears to be a resounding 'no'. Yet 'Restoration', i.e. restoring ecosystems to viable forms and functions is undoubtedly important today." "Restoration of degraded eco-systems to a viable form and substance should therefore, receive the highest priority in any plan of economic and social development. Unfortunately its importance and urgency are not recognized by anyone in India." In this article, we also provide guidelines/recommendations that could be useful in restoration planning in the Northern Western Ghats. These recommendations can also be useful for researchers and restoration practitioners in the Western Ghats in general and dam catchments in particular. Finally, we aimed to develop a method for selecting restoration candidates by prioritizing and choosing from multiple such catchments.

Study Process
The study area i.e. the three catchments lie in the north and west of Pune district (Figure 1 and 2). Table  1 provides the key details for the three catchments.
For the selected three dam catchments, we performed the following process : 1. Satellite images were obtained and the dam catchment region was marked for study with the help of topographic sheets. 2. With the help of a GIS consultant and assigning vegetation class signatures, a mapping of the dam catchment by vegetation class and landusewas prepared (software: ArcGIS). 3. Field trips of the dam catchments were conducted to do ground truthing of vegetation classes. 4. Restoration potential was estimated and restoration guidelines are provided across dam catchments and specific to dam catchments. 5. Developed a method to prioritize and select when one has to choose from multiple such catchments. As part of Step 4 above, we also developed a general-purpose restoration framework and detailed guidelines for restoration in the Western Ghats. This article mainly summarizes work done in Steps 1-3 and 5 with key results from step 4.

JOURNAL OF ECOLOGICAL SOCIETY 2018
Physical and social character of catchments

Pavana
The physical character of the Pavana catchment shows steep slopes and narrow valleys near the crest line of the Sahyadri (Northern Western Ghats). The high drainage density indicates landscape made up of innumerable small sub-basins. The nature of the landscape is young, immature, and dissected.
The valley becomes broad in its middle and lower part. Basalt is the major underlying rock. The soil near the crest-line where heavy rainfall occurs is red while the soil in the valley is typical black cotton soil with silt-loam texture.
The Pavana dam catchment is located close to the cities of Pune, Lonavala and Mumbai. Its natural beauty and proximity to cities make it a favourite location for second homes of urbanites. The hills in the catchment are dotted with such farm-house development projects.
The forest patches are seen on the hill tops of the catchment however they are rapidly getting degraded due to residential construction. Junnar Division Working Plan of the Maharashtra Forest Department mentions that "Various detrimental agencies like fire, grazing, illicit cutting, encroachments, poaching etc. have a deleterious effect on the health and growth of the forest." (Gov. of Maharashtra, Forest Department, 2006) In the valley, agriculture is predominant. Shifting cultivation, though on a small scale, is still practised on the hill slopes. Deviation from the traditional agriculture in the form of floriculture in poly-houses occupies a significant area. This is not a densely populated area. Looking at the number of farm-houses, it is likely that this area has a floating population from the near-by cities of Lonavala, Pune and Mumbai.

Chaskaman
Steep slopes and high drainage density makes the North-western parts of Chaskaman catchment more rugged. Southern slopes of Bhimashankar form part of the Chaskaman catchment. Therefore nature of the landscape is young, immature, and dissected in the upper part of the valley. The middle part of the valley is relatively broad as significant Arala tributary with its sub-catchment joins the river Ghod, creating a terrain more complex.
In the Chaskaman dam catchment, traditional agriculture is observed. It does have human-induced stresses like clearing of mature forest for agriculture on privately-owned area. In-stream agriculture is practiced all over the catchment. Diverting the streams and creating terraces for cultivation alters the character of the stream and the drainage pattern. Modern development is also insignificant.
In spite of hilly and dissected character of the catchment, the number of human settlements is considerably high.

Dimbhe
The River Ghod originates on the Bhimashankar plateau. Here the plateau is extremely rugged as it is criss-crossed by several deep gorges of River Ghod and its tributaries. Thus physiography of the upper part of the Dimbhe catchment is very complex. (See Figure 3 for representative cross-section of Dimbhe catchment) Physiography in the middle and the lower part of Figure 3 : Representative cross section of the Dimbhe catchment and the land use classes-Y axis: Altitude in meters; Scale: 1 cm = 100 meters, X axis: Ground distance in meters Scale: 1 cm = 500 meters, Index of land use/ The Dimbhe catchment has predominantly tribal population in the hilly part. The local population is engaged in agriculture but is also dependent on the forest produce. The forest quality is good since it is adjoining the Bhimashankar wildlife sanctuary. In the valley, the main activity is subsistence agriculture.
It is observed that Bhimashankar shrine has become very popular for religious tourism. This has resulted in commercial and modern development in the catchment area.

Methodology of the Study
We adopted the methodology developed during the previous experience of forest analysis, vegetation assessment, and ecological restoration in the Ecological Society's research project which is described below-

Reconnaissance Survey
A preliminary survey of the catchment area was done torelate to the wider landscape. The main objectives were to understand • the physiography of the catchment : whether its major part is hilly or moderately hilly or has larger proportion of plains, nature of the slopes, etc. • general assessment of the vegetation cover and forest type • presence of Devrai (Sacred Groves) which could be used as a reference ecosystem • water resources and their status (whether they areperennial) etc. • general density of village settlements • area under agriculture, type of agriculture • any other major human interventions.

Acquisition of the Topo-sheets and Google Images
Topographical-maps for the dam catchments were obtained from the Survey of India. The catchment area was identified by taking into account the water-divide. It also helped us in understanding the character of the slope, drainage and extent of forest.
We used the Google image of the catchments along with the topographical-maps. Marking of the water divide to demarcate the dam catchment was a combination of GIS work and close manual observations of topographical-maps. The catchment boundaries delineated on the topographical-maps were superimposed on the Google image.

Mapping Process
The project team worked with a GIS consulting team for the actual mapping of the defined vegetation and land-use classes on the Google image, using ArcGIS software. This identification was done based on the team's domain expertise, observations from the reconnaissance survey, and past projects done by Ecological Society. On an average, each image comprised of about 3000 such polygons.
Polygons where there was uncertain identification of vegetation were added to a "Query List". These polygons were confirmed by ground truthing in the actual field. The landscape is represented as an assemblage of polygons of various vegetation and land-use classes.

Vegetation and Land Use Class Mapping
The landscape is a mosaic ofnatural vegetation, plantation, agriculture and interventions like roads, quarries, other constructions etc. (Figure 4)

Socio-economic conditions
As a foundation for restoration planning, a detailed socio-economic analysis of the region is essential i.e. the people residing in the area, their resource needs, the local economy and its integration with the national economy, and fulfilment of their needs through local natural resources. Secondary data collected on socioeconomic conditions is being analyzed separately and not covered in this article.

Management information from government agencies
A major stakeholder is the government, i.e. departments like Forest Department, Department of Social Forestry and Water Resources Department (formerly, the irrigation department). To find out their land use and utilization and management of natural resources in the catchment, the project team visited the respective departments and got an understanding of their policies and working.
Forest resources of these catchments are managed by the Forest Department. Its role is to protect, conserve and enhance the forest resource. Pavana and Dimbhe contain a significant area under forest department's ownership (~21% and ~16% respectively per our estimates), while Chaskaman has relatively less area under forest department (estimated to be around 5-6%).
Forests are cut mainly for timber and fire-wood by local people. The department of Social Forestry came into existence to help locals and fulfil their needs by energy plantation on their own land, thereby reducing stress on the government-owned forest. This Department does not own any land by itself. Methods of operation of Social Forestry Department are summarized in online supplementary data.
The Water Resource Department is another major stakeholder as water resources from the catchment area are managed by this department by storing water behind the dam and redistributing it through canals.
To better understand the government stakeholders, the following were studied : We tried to relate the secondary data from the Forest Department and Water Resource Department to understand resource distribution among local communities and people outside the catchment.

Results
The analysis and interpretation of the Land-use classes in each catchment is given below. Figure 5 shows the vegetation and land use map of Pavana catchment. Table 2 shows area statement by vegetation and landuse and Table 3 shows vegetation composition.

Pavana : Observations and Comments
• The western part of the catchment, near crest-line, has high rainfall (2500 mm). Semi-evergreen forest is observed in this area having composition of Memecylon-Actinodaphane-Syzygium associated with Olea dioica, Atlantia racemosa, Glochidion ellpticum, etc. • Moist deciduous type of forest was observed in the zone with the rainfall between 1500 mm and 2500 mm which includes Terminalia sps.-Bombax ceiba, Ficus racemosa, associated with Meyna laxiflora, Albizzia chinensis, etc. • Mature Forest is mainly observed at the mountain tops, base of vertical cliffs, slopes and ravines, near the crest-line. • In zone III, with rainfall less than 1500 mm, dry deciduous forest type was observed with Acacia sps., Bombax ceiba, Boswellia serrata, Cassine glauca, etc. • A Sacred Grove is present at the base of Fort Tung (18º 39.862" N, 73º 27.447" E) and is named after the goddess Waghjai. Its area is 0.34 acres. It is surrounded by agricultural land. Dense patch of unique vegetation of Pandanus sp. is an indicator of high ground water level which was validated by the flowing perennial well. Old growth trees like Caryota urens, Syzygium cuminii, Mangifera indica, Olea dioica, Ficus racemosa, Atlantia racemosa associated with climbers like Elaeagnus conferta, Dalbergia horrida, Tylophora dalzelii, Jasminum malabaricum are found in this sacred grove. • The Pavana catchment also consists of a few more sacred groves. The Ajivali sacred grove is a known example for its commercial management by local community. • The Pavana catchment has lot of modern development, having resorts, farm houses and spas. Road construction through thick forest near village Dudhivare pass has fragmented the forest, and resulted in soil erosion. (Figure 6) • Dwarf canopy forests with composition of Memecylon-Syzygium-Actinodaphane associated with Canthium dicoccum, Careya arborea, Macaranga peltata, etc. which is a typical montane forest composition near the crest line was also observed in Pavana catchment. • Dense Shrubbery was observed with intermittent trees on gradual slopes found in all rainfall zones. As the rainfall zone changes, tree species in the dense shrubbery also changes. E.g. Memecylon, Actinodaphane in high rainfall zone area and Terminalia, Bombax in low and medium rainfall zone. • At places Karvi patches dotted with trees like Olea dioica, Memecylon umbellatum near crest-line were observed on the steep slopes. • Abundant shifting cultivation patches with good shrubby growth were observed on the higher slopes. • In dwarf canopy forest, the composition is more or less similar to dense shrubbery with increased tree density. • Dense shrubbery patches are termed as regenerating patches because vegetation structure is in regenerating state due to presence of root stocks. These areas have a good potential for restoration as these are moderately degraded areas compared to scrublands • Observed intermittent coppiced trees of Dillenia pentagyna, Terminalia elliptica, Syzygium cuminii, Carrisa congesta in land patches used for Rab (It is an age-old practice in which farmers burn the piece of land before cultivation). Farmers also cut standing trees on the plot where they want to cultivate. These patches indicate earlier canopy forest.   for the city-dwellers. Another activity of modern agriculture is floriculture in polyhouses. The spread of polyhouses is significant. (Figure 7) • We saw some abandoned stone quarries in the catchment. e.g. near Thakursai village adjacent to the dam wall. In the quarries, there is no substratum, but completely exposed rock. Such areas are difficult to restore ecologically as there is no soil left. However such areas, if protected and restored, will provide refuge to a variety of biodiversity. (Figure 8) (Bradshaw et. al., 1980) Chaskaman : Vegetation Classes and Composition Figure 9 shows the vegetation and land use map of Chaskaman catchment.   Hence, natural vegetation could be observed from Mandoshi onwards in the western part of the catchment. • Nilgiri plantation was also observed near Dehane village. • Modern development has not yet emerged as a significant threat to the landscape, but this may change in future. At a few places, we saw roads being constructed right on the hill for proposed projects. (Figure 10) Figure 11 shows the vegetation and land use map of Dimbhe catchment. Table 6 shows areastatement by vegetation and landuse and Table 7 shows vegetation composition.

Dimbhe : Observations and comments
• Natural vegetation classes comprises of 64.96% of the total catchment area. • Mature forest is higher as compared to Chaskaman in Dimbhe • Open area is maximum (19.39%) of all the vegetation classes indicating high level of degradation.  • The Dimbhe catchment receives rainfall of less than 2000 mm, hence it is unable to support species of evergreen forest. In the Western part of the catchment, semi-evergreen forest is present. The second important forest type seen here is the moist deciduous forest. • The Sacred grove (Devarai) near village Phalode is devoted to the deity named Marideo. It has old growth trees like Ficus amplissima (Piparani), Terminalia chebula (Hirada) associated with Syzygium cuminii (Jambhul), Atalantia racemosa (Makadlimbu), Mangifera indica (Amba), Casearia graveolens (Bokhada), Actinodaphane angustifolia (Pisa), etc. The Devrai is located near the main road and surrounded by agricultural land with a few settlements. Therefore, it has become isolated and lost connection with other forest patches. Prominent edge-effect could be observed with weeds and hardy species like Lantana camara (Tantani), Crotalaria pallida (Khulkhula). • The Ahupe sacred grove is fragmented by road. It has many old growth trees and lianas. It is an ideal habitat for Ratufa Indica, the Indian giant squirrel which is a key-stone species and endemic to the Western Ghats. • Community conserved Hirada trees are present in Phalode and in many other villages in the catchment area. Collection and sale of Hirada fruits is one of the income sources for the locals in this catchment. • Mature forest was also observed at the end of the west side of the catchment which includes area of Kondhwal and Ahupe located near the crest-line. • Good forest patches were seen near 1 st order stream niches which are almost inaccessible for the humans and have higher availability of moisture. • The composition of dense shrubbery in this catchment is more diverse, with relatively larger number of species. Please see vegetation composition in tabular format given in Table 7. • The opportunity to improve the status of dense shrubbery to mature forest by restoration is significant.

2018
JOURNAL OF ECOLOGICAL SOCIETY

Catchment Restoration Potential and Recommendations
This study defines Restoration Potential of a dam catchment as the total area of land in the catchment underthe four main natural vegetation classes (Mature Forest, Dense Shrubbery, Scrub, Open land) and plantation. These classes can benefit from protection and restoration and lead to an improvement in nature's services and biodiversity. (Urbanska, 1997) The overall region can be seen as a mosaic of habitats. Finding restoration potential in a specific part of it requires us to evaluate forest patches, corridors that connect patches, and matrices, which are arrays of various land uses and habitats around a forest patch, dotted with vegetation (Forman et. al. 1996). The ecological quality of the matrix (and it's variation over seasons) are important when restoring overall ecological health of a fragmented landscape (Jules et. al. 2003).
The central theme of our restoration recommendations is to protect good forest patches, restore degraded patches, connect patches using new or existing corridors where possible, and protect matrices (as they provide nature's services to surrounding forest patches).

Pavana : Restoration Potential and Recommendations
Restoration potential of the Pavana catchment is 68.96 sq. km. Please refer to the Figure 5 when reviewing the recommendations below.
1. In the high and middle rainfall zone, we observed a lot of scrub. Ideally, there should be good forest cover, considering the rainfall in this area. Hence the large extent of scrub land here is unacceptable and indicates a high degree of degradation. We recommend this area (patches of scrub) as a top contender for restoration. 2. A lot of modern development including polyhouses is observed in the midst of mature forest. Such development and roads leading to them can cause further forest erosion, introduction of waste in natural ecosystems, and long term impact on nature's services. Such projects should not be allowed in mature forest areasin future. For already existing projects, the owners should be involved in landscape level restoration programs. The local councils and state government should consider restrictions on further development in/near mature forest. 3. In-stream agriculture is relatively high. It should be discouraged or sustainable alternatives should be provided.   . Similarly, owners of the land where modern development exists can be integrated in the restoration program. With proper protection, the patches of scrub and dense shrubbery adjacent to mature forest would automatically upgrade so as to form a continuous, large stretch of mature forest. 5. Similar condition is observed to the west of Shilimb, where there are large stretches of dense shrubbery in the midst of mature forest. Similar program to increase the extent of mature forest should be considered here.

Chaskaman : Restoration Potential and Recommendations
Restoration potential of the Chaskaman catchment is 185.45 sq. km. Please refer to Figure 9 when reviewing the recommendations below.
1. In the high rainfall zone, towards the west, near crest-line of Western Ghats, village settlements are few. Mature forest areas are surrounded by dense shrubbery area. With simply more protection mainly from human interference, these areas can be restored to a continuous stretch of mature forest. Some of the area within this zone belongs to forest department and may be effectively protected by preventing grazing and agriculture. 2. In the mid rainfall area of the catchment, dense shrubbery is present to the north of the reservoir. However, it is absent to the south of the reservoir. Logically, with same rainfall, even this area should support dense shrubbery vegetation. To make this possible, restoration programs should be devised including trading of land. The land owners can be given a substitute land in open areas identified in this study to practice agriculture. of tribals and other locals as well as their aspirations should be taken into account when planning restoration. 5. Regarding private owners asserting land rights by clearing mature forest on their own land, we suggest a land trade program be worked out with the owners. Owners can be offered subsidy or an alternative piece of land in open vegetation areas in the catchment. 6. Dense Shrubbery throughout the catchment should be provided protection.Protection can be provided by a. Forest department if its within their reserved forest area b. Land trade programs to bring more land under reserved forest, subject to socio-economic feasibility c. By convincing private landowners for more protection and restoration. Suggested ideas could be i. Private sanctuaries ii. Landowner-conservator's collective (Pathak, 2015) iii. Model Eco-sensitive Zones (Pathak, 2015) Dimbhe : Restoration Potential and Recommendations Restoration potential of the Dimbhe catchment is 177.99 sq. km. Please refer to the Figure 11 when reviewing the recommendations below.
1. To the west of the catchment, between Walunjwadi and Kondhwal, there are fragmented patches of mature forest and dense shrubbery. This area belongs to Bhimashankar Wildlife Sanctuary. A restoration program can be implemented by Forest Department to create continuous green cover. 2. There is a large extent of dense shrubbery throughout this catchment which should be provided protection. There is a need to educate the local community to know the benefits of protecting this vegentation type. 3. Continuity of mature forest and dense shrubbery areas should be a top priority. Both, currently are highly fragmented. Restoration programs should focus on creating corridors connecting these patches of dense shrubbery to increase available habitat for various kinds of biodiversity. 4. Agriculture in the hilly areas with steep slopes is difficult. Often it is subsistence agriculture, as is the case here, to the north-east of the catchment in the vicinity of villages Chinchawadi, Jinjrewadi, Shindewadi, Ghotmalwadi etc. A substi-tute land in the open areas could be offered to the owners and hilly areas could be provided complete protection for natural vegetation to grow. 5. Drainage density of this catchment is high indicating a large number of streams. However, compared to these number of streams, riparian vegetation is negligible. Creation of riparian zone along these streams will add ecological value to the catchment.

Recommendations Applicable Across Catchments
A few additional recommendations and ideas that could be applicable to any of the three catchments are given below :

Restoration Prioritization across Catchments : A framework
The study also attempts to address the possibility that a funding agency, an NGO, or a government department may be faced with the decision of selecting a given dam catchment for restoration from among multiple such candidates. This could happen due to budget and time constraints, interest in relative restoration potential, and so on.
The framework below tries to answer the above question, on an experimental basis.Itis based on our limited analysis and not on any focused or empirical studies. It is an intuitive framework based on our general experience with restoration in the Western Ghats.
Disclaimer : For selecting a given catchment among multiple such candidates for restoration, several aspects are pertinent. We only cover a subset of these below. Moreover, most of the scoring methods we provide below are simple and could be refined. Further analysis and enhancement of this framework is needed.

Extent of vegetation cover
For our purposes, this is the percentageof land in the catchment under the four natural vegetation classes: mature forest, dense shrubbery, scrub, and open land. The higher the extent, the more the potential for protecting vegetation and improving natural forest. On the other hand, it could be argued that the lower this number, the more the need for restoration, since this area is eroded and is in need of restoration. While both these arguments are true, we recommend the "higher" extent as a driver for determining potential, simply because the more the vegetation, more the nature's services that can be counted upon from this area.

Overall amount of fragmentation of natural vegetation
Fragmentation is the effect of land use changes due to human intervention of all types, gradual natural processes including land erosion, climatic conditions over the years, or calamities (e.g. landslides or diseases on plants). The more the fragmentation of vegetation, the longer the time, and more the effort and costs that will be needed to restore a landscape.
Fragmentation could be measured using sophisticated GIS analysis techniques (e.g. statistical analysis of vegetation class polygons in an average unit area of a landscape), however we could not attempt this during the course ofthis project.
For now, we have a simple scoring scheme, based on a subjective assessment of the LULC maps, as below : Fragmentation score for our three dam catchments are given in Table 8.

Conduciveness of natural conditions
This is an important dimension for restoration and we can think of sub-dimensions as below.

a. Rainfall
The more the rainfall, more is the water availability 56 JOURNAL OF ECOLOGICAL SOCIETY 2018 in a catchment on an annual basis. Nature's response to higher rainfall is faster growth of forests. The scoring for scale for rainfall should not be absolute, but should be determined in a relative way, among the candidate catchments. In our case, all three catchments are in the Western Ghats, but Pavana gets relatively more rainfall while Dimbhe and Chaskaman get relative less. The relative scoring is provided in Table 8.

b. Drainage Density
The higher the drainage density, the more the length of drainage paths in a unit geographic area. Hence higher drainage density will mean relatively more water availability, making a catchment more conducive for restoration. More drainage density also results in higher habitat diversity and supports diverse platforms.This too should be a relative measure among the available candidates with some kind of indexation applied.
Since the drainage densities of the three catchments are close to each other, we assign the same and arbitrary score of 3 to all. If any of them was to have a significantly lower drainage density, say < 1 or between 1-2, a lower score like 1 or 2 could be assigned.

c. Soil Quality
Richer the soil quality in a given landscape, higher will be chances of success of restoration since less effort will need to be spent in enriching the soil. Soil Quality measures could be elaborate and will need detailed sampling and analysis methodologies. Also it may not be entirely correct or easy to assign a single soil quality measure across the landscape. Rather, a basket of soil measures may need to be worked out. In the absence of all this, we have assigned subjective scores to the soil quality in the three catchments based on our general observations.

d. Shape of catchment
The more elongated a catchment, the higher the probability that multiple, varying "rainfall instances" will happen across the length of the catchment. This means more rain will be retained and drainage will be gradual. On the other hand, the more circular a catchment, the faster will be runoffout of the catchment. Thus the shape of the catchment has an impact on the extent of moisture retention, which in turn affects restoration potential. Scoring across catchments on this parameter will be relative. A rigorous quantitative model for shapes can be done using measurement of long and short axes, further spatial analysis, etc. For now, simple scores for our three catchments, based on our assessment of their shape is provided in Table 8.

Conduciveness of human conditions and interference
Restoration programs cannot be successful without the involvement of stakeholders. It may be important to develop a variety of measures to express how conducive the general human conditions in a catchment are, for restoration. Below are two sample measures.
a. Degree of human land use We have simplified this measure to take the sum total of percent land under all human use. This could be fine-tuned further to consider the varying impact each of them may have on restoration efforts. E.g. Roads bisect forests and affect habitat development, Village settlements and modern dwellings may be sources of waste that choke up natural ecosystem flows. All these impact ecosystem health and hence restoration success. For now, we will simply add up all human land use and assign a score as 4 (<25%), 3 (25-50%), 2 (50-75%) and 1 (75-100%). Extent of human land use and scores for our three dam catchments are given in Table 8 b. Vicinity to cities Particularly in a densely populated, developing country like India, the more the vicinity of a forest to a major metro area, higher the chances that human land use in future may suffer from "urban effects", including urbanites buying land, developers building farm house schemes, large companies choosing these areas for townships and private hill stations, tourism, and so on.
Thus simply being near a city can affect the potential for restoration. The farther an area from a city, relatively less will be the urban interference. This could also be modelled quantitatively though actual distances and road connectivity to nearby metros. For now, we observe that Pavana is very near to Mumbai, Pune and Lonavala, while Dimbhe and Chaskaman have only Pune as a nearby city, but not as close as in case of Pavana. The scoring is given in Table 8.

Composite Scores of the Three Catchments
After developing such basic measures, we could consider arriving at a composite score to provide a more definitive answer as to how to prioritize across catchments based on their restoration potential. For this, various weighted average or other such formulae could be considered. For now, we will do a simple addition of the above scores.
Thus, in this case, the simple model developed above does not provide a distinctly attractive candidate for restoration. Rather all three catchments score close to each other and have similar restoration potential.
In such a case, a decision could be made based on the total sq. km. of restoration potential in the catchment. In our case, the highest potential is with the Chaskaman catchment, of 185.45 sq. km.
In conclusion, our study of vegetation cover, land use and ecological status of the Pavana, Chaskaman, and Dimbhe catchments points to the significant restoration potential of such regions and how restoration can contribute to the larger objective of conserving the Western Ghats biodiversity hotspot. A landscape ecological approach, with restoration potential as a basis for discussion, can help develop a constructive dialog between all the stakeholders i.e. government departments, local communities, ecological and socio-economic experts and NGOs. We hope this study and similar such studies provide useful data for such dialogs and prioritization of restoration.