Livelihood Dependence, Traditional Knowledge Conservation ...

Jayita Bit & Sarmila Banerjee/ Arthaniti 12 (1-2)/2013/1

Livelihood Dependence, Traditional Knowledge Conservation and Sustainable

Forest Management – A Study on Birbhum Forest Division, West Bengal1

Jayita Bit2

Department of Economics, University of Calcutta

[email protected]

Sarmila Banerjee Department of Economics,University of Calcutta

[email protected]

Submitted: 27.8.14; Accepted: 9.2.15

Abstract.

This paper attempts to verify whether the present forest management policies have ensured biodiversity

conservation and in turn livelihood protection for the forest dependent people. For this purpose a household survey

was conducted in the Birbhum district of West Bengal, India and various socio-economic determinants of livelihood

dependence on forests have been identified using simple OLS regression. LOGIT regression, both regular and

ordered, are applied to assess the marginal influences of different ‘cause factors’ on enhancing the probability of

conservation of traditional knowledge within the forest community. Results strongly suggested a cultural transition

among the local people who were once used to self-sustained forest dependent livelihood, but due to forest

degradation and non-availability of necessary resources/ services were gradually exposed to the outer world and

enhanced participation in general economic development related activities. The economic use of forest dominates

the newly afforested areas as against the concern for protecting their biodiversity leading to comprehensive

ecological balance.

Keywords: Joint Forest Management (JFM), Livelihood Dependence, Cultural Migration, Flow

of Traditional Knowledge, Ordered LOGIT regression, Birbhum

JEL Classification: Q23, Q57, C25

1The authors wish to thank Debabrata Biswas, the Chief Conservator of Forest (WB) for his guidance throughout the

field survey, the participants of the Research Scholars’ Workshop held in the University of Calcutta during July

2014 and Indrila Guha of Vidyasagar College for Women for her insightful suggestions on an earlier draft. The

comments of the anonymous journal referees add to both focus and clarity of presentation. Of course the usual

disclaimer applies. 2 Corresponding Author. 56A, B. T. Road, Kolkata – 700050; Tele-fax: 033-2546-5949.


1. Introduction

Forests play a vital role in sustaining the life supporting systems of a country's environment and

the quality of life of its people. From time immemorial, the aboriginal people were the original

forest inhabitant whose livelihoods were based on hunting and gathering for the supply of food,

fodder, fuel‐wood and even cosmetics (wild gems) from the woods (Ghosal, 2011). Tagore

(1915) pointed out that “in India it was in the forests that our civilization had its birth”. Vedic

literature also indicates that forests were held in high esteem and ‘ashrams’ (hermitages) of the

sages existed in these forests. Thus forests always helped to create an integrated and harmonious

coexistence of man and nature through their various social and cultural uses. Even in our present

times, as noted in the State of Environment Report (2009), nearly 0.7 billion rural population of

India out of the total population of 1.21 billion directly depend on climate-sensitive sectors like

agriculture, forests, fisheries and other natural resources such as water, biodiversity, mangroves,

coastal zones, grasslands, etc. for their subsistence and livelihood. The share of forest dwellers

alone exceeds 100 million and 54 percent of them belong to the tribal communities (Sankaran et

al., 2002). Their living and livelihood are so intricately woven with the forest and its product

variety through customs and culture that they act as a niche for the tribal communities, who are

not only extracting forest resources but are also trying to maintain its ecological balance through

different rituals and practices at the same time (De, 2012). The destruction of forests through

land acquisition and conversion becomes a serious concern only when the influx of outsiders

raises the speed of deforestation beyond the carrying capacity of the system. A number of studies

have also observed greater extraction tendency of the forest resources from relatively well-off

people and the settlers compared to the poorer and native ones (Pandey, 2010). So poverty alone

cannot be blamed for forest degradation; it is a combination of ignorance, greed, power and

wealth that results in deforestation, especially in developing countries. The aborigines and

adivasis are essentially a part of the forest ecosystem and promotion of economic development at

the cost of forest degradation raises the vulnerability of their daily livelihood (Padel and Das,

2010). However, it cannot be denied that certain demographic, cultural and environmental

changes may alter the configuration of institutions and values that characterize traditional

resource-dependent communities and in turn decrease their willingness for forest conservation

(Robson and Nayak, 2010). So, sustainable forest management attempts to establish a balance


between the global eco-system and the directly forest-dependent community at the local level;

this should ensure not only the protection of forest canopy cover but its richness in terms of

biodiversity should also be protected to guarantee essential livelihood support, which deserves

equal attention, if not more. So, now there is a need to understand the economic and social issues

related to forestry at the grass-root level faced by the forest dependent community at the micro

level by observing the impact of Joint Forest Management (JFM), the present forest management

approach on local livelihoods. Thus, in this paper we are interested in exploring issues related to

(a) the relationship of the nature of forest biodiversity with the pattern of forest dependent

livelihood, given control for different socio-economic characteristics of the users and (b) the

intergenerational flow of traditional knowledge about forest resource and its impact on the

observed pattern of forest dependence. Since all these issues deserve in-depth investigation, a

micro-level study is appropriate here as it creates the opportunity to look at the micro-level

interactions among individuals. And to collect fair and detail information from local people

(especially if they are indigenous) through such interactions, the interviewer must be an insider

in terms of ethnicity, community or locality. Also the possibility of interview at grass-root level

is easy if it is conducted in vernacular. So this paper is basically the interpretation of a primary

survey conducted with individual stakeholders in the Birbhum Forest Division of West Bengal

(WB), India. Given this background the rest of the paper will be organized as follows: section 2

will describe the features of our field area, section 3 will talk about the survey design and

questionnaire and section 4 will present the household level analysis. Finally, section 5 will

conclude the paper by extending an overall discussion.

2. Birbhum Forest Division

To document the pattern of livelihood dependence in afforested land, a field survey was

conducted during September- November, 2012 in the district of Birbhum, West Bengal. This part

of the state is bounded on the North by the Santhal Parganas of the state of Jharkhand and in the

West and South by the districts of Murshidabad and Burdwan of West Bengal. It is separated

from Burdwan by the Ajoy River (Figure 1) (Debnath & Mondal, 2014). This district is notable

for its undulating topography and its cultural heritage which is unique and is somewhat different

from that of the other districts of the state. In fact, in the native language of the region the

meaning of the term ‘bir’ is forest, so, Birbhum is basically a land of forest. Though at present


the district is predominantly agricultural, 3.51% of total land area is reckoned as forest land

(Human Development Report, 2009). Forestry is one of the main industries in this district.

Among 425 Small and Micro Enterprises and Artisan units present over here, 145 depend on

forest products like timber and wood pulp. There are 13 wood and wood based furniture units

and 132 paper and paper product units existing in this district (Directorate of M & SSE, 2013).

From the administrative perspective of the Department of Forestry, Government of West Bengal,

this district is divided into 7 ranges (Bolpur, Md. Bazar, Rajnagar, Rampurhat, Sainthia,

Dubrajpur and Suri) containing 19 beat offices. Among them, 6 beats under 5 ranges (Map 1)

were randomly selected for the present study to cover 10 locales with the presence of active

Forest Protection Committees (FPCs), keeping in mind the variations in topographical properties,

cultural heritage as well as nature of forest biodiversity (Table 1) . Ranges with limited forest

cover were excluded from the sample.

Figure 1: Location of the Study Area

Source: Debnath & Mondal, 2014

3. Survey Design and Questionnaire

Historically the forests of Birbhum remain covered with Sal trees3. However, our survey found

dominant presence of Eucalyptus and Sonajhuri with Sal trees mainly present at jahir sthan4

3 The forest areas that remained dense at the time of introduction of JFM generally have not lost their density till

now. However, the less dense or scarce forest patches are afforested with the monoculture plantation of non-native

species.


Map 1: Surveyed Ranges of Birbhum Forest Division

Source: Aranya Bhawan, Government of West Bengal, 2012

4 Jahir Sthan is a sacred place of worship in traditional tribal culture; the tribals worship trees like Sal.


under most of the sampled FPCs. The study region was classified into three zones on the basis of

forest types (nature of forest biodiversity). This is because forest uses are governed by available

tree species and the more the variations among tree/animal species, the more the dependence

should be. From recent literatures like Rojas (2012) and McDermott (2012) it has also become

evident that dominance of traditional species increase forest biodiversity while monoculture

plantation of non-native eucalyptus or acacia species decreases the chance of existence of other

species. It is also found in our study that though the alien species are an efficient means for quick

afforestation and for rapid production of natural and renewable fiber for a growing industrialized

economy, they are associated with demolition of ecosystem health and biodiversity and loss of

the traditional forest dependent practices of local communities. Accordingly, all the FPCs are

divided into three zones: FPCs under natural forest type, those under mixed type of forest and the

rest under monoculture plantation forest type as shown in Table 1.

Table 1: Surveyed Forest Areas in Birbhum: Location and Characteristics

Source: DFO, Birbhum

In all, ten FPCs have been surveyed maintaining variations with respect to the percentage of

tribal members, proximity to township, quality of FPC performance, etc. From each of the

locations, ten households were randomly selected containing at least one family person as FPC

FPC Mouza Beat Range Nature of

Forest

Dominant

species Kheledanga

Adibasi Kheledanga Bolpur Bolpur

Monoculture

plantation

Eucalyptus,

Sonajhuri

Lohagodd Benuria Bolpur Bolpur Monoculture

plantation

Eucalyptus,

Sonajhuri

Jamboni Murgaboni Illambazar Bolpur Natural Sal

Bonvilla Ramnagar Illambazar Bolpur Natural Sal

Laxmipur Laxmipur Illambazar Bolpur Mixed

Sal,

Eucalyptus,

Sonajhuri

Bonsuli Usardihi Illambazar Bolpur Natural Sal

Pachiara

Chandrapur-

Srichandrapur-

Bodaguri

Hetampur Dubrajpur Monoculture

plantation

Eucalyptus,

Sonajhuri

Sultanpur Sultanpur L.N.Pur Rampurhat Mixed

Sal,

Eucalyptus,

Sonajhuri

Asna-

Sundarkhele

Asna,

Sundarkhele Aligarh Rajnagar Natural Sal

Jethia-Rampur Jethia Mallarpur Md. Bazar Monoculture

plantation

Eucalyptus,

Sonajhuri


member and to isolate the impact of FPC participation on forest dependent livelihood,

conservation of traditional knowledge and sustainable forest management, similar information

was collected from ten households not containing any FPC member. Information pertaining to

personal details of household members, condition of the nearby forest-area, general awareness

regarding the role of forest, role of FPC and the flow of traditional knowledge was collected

from personal interviews made to a total of 204 households. To facilitate the ease of data

collection all the questionnaires have been developed in a bilingual mode. Finally data collected

was interpreted and analyzed by utilizing the direct, first-hand experience gathered from the

field.

4. Household Level Analysis

4.1 Salient Features of the Sample

Nearly 63 percent of the sampled households belong to the schedule tribe with around 49 percent

illiterate household head and 50 percent holding the Below Poverty Line (BPL) card. They are

mostly engaged in traditional vocations with 56 percent directly engaged in forestry and

agriculture based activities. Table 2 presents the member characteristics of the sampled FPCs in

terms of the tribal dominance in the membership pattern as is obtained from the official records

of the Forest Department and compares that with our sample properties to ensure

representativeness of the latter. From the table it appears that at least in four out of ten FPCs the

Schedule Tribe (ST) respondents are over represented. Though the overall presence of ST

members in the FPCs is around 40 percent in our selected sample, combining both the FPC and

non-FPC households the corresponding figure is around 63 percent. There was no official data

available on the education status of the households according to the forest ranges/ beats and

therefore no direct comparison of sample composition with the relevant population averages was

possible here. The following analysis is reported by keeping this limitation in mind.

4.2 Livelihood Dependence on Forest

The views of the local villagers (both FPC members and non-members) regarding forest

dependence and present management practices were widely divergent, depending upon their

culture, educational status, occupational pattern, etc. It is noticed that though nearly 90 percent of

the sampled households depend on forest for their daily needs (Table 3), figure 2 shows that out


of this nearly 50 percent uses only fuel-wood, 11 percent uses only non-fuel wood related

products and only 29 percent uses both.

Table 2: Sample Coverage of ST Members of the Selected FPCs

FPC

Record of the Forest Department Sample Composition

No. of

member S.T. % ST

Sample

member

ST

member % ST

Kheledanga Adibasi 52 52 100.00 17 17 100.00

Lohagodd 26 0 0.00 10 0 0.00

Laxmipur 74 20 27.03 11 5 45.45

Bonsuli-Beloan 299 33 11.04 10 10 100.00

Jamboni 47 47 100.00 10 10 100.00

Bonvilla 51 30 58.82 10 8 80.00

Pachiara 65 0 0.00 10 0 0.00

Sultanpur 55 7 12.73 10 1 10.00

Asna-Sundarkhele 84 41 48.81 11 10 90.91

Jethia-Rampur 145 125 86.21 11 9 81.82

Total 898 355 39.53 110 70 63.64

Source: Department of Forestry, GoWB & Primary Survey, 2012

Table 3:Livelihood Dependence of Respondent's Family on Forest

Nature of Forest FPC Names Percentage of Positive Response

Natural Forest

Bonsuli-Beloan 95.00

Jamboni 100.00

Bonvilla 95.00

Asna-Sundarkhele 100.00

Total 97.50

Mixed Forest

Laxmipur 100.00

Sultanpur 75.00

Total 87.50

Monoculture Plantation

Kheledanga adibasi 100.00

Lohagodd 65.00

Pachiara 73.91

Jethia-Rampur 95.24

Total 83.54

Total 89.71

Source: Primary Survey, 2012


Figure 2: Pattern of Usage of NTFP


Using the information collected on the nature of day-to-day dependence on forest bio-diversity,

an estimation of average dependence is also obtained in terms of the types of dependence (Table

4). Six broad types of dependence are identified as (a) using for fuel woods, (b) making of Sal

plates, (c) consumption of vegetables and other food items, (d) consumption of fruits, (e) using

for medicinal plants and (f) other usages including collection of materials useful for household

chores, using forest as grazing land, etc. From the survey responses it is found that nearly 78

percent households are relying on forest as a source of fuel wood, but less than 4 percent depends

on forest for collection of medicinal plants. The households depending on forest for at most one

factor is identified as low-dependent ones and those with dependence on at least 5 types or more

are marked as highly dependent. Those in-between are moderately dependent. It is apparent from

Table 5 that the majority of the surveyed are low dependent (65.69 percent) and the relative

share of the households under monoculture plantation is the highest in this category while that

under natural forest is the lowest. On an average only 7.35 percent have high dependence on

forest produces. Thus, there is a relation between nature of forest and livelihood dependence.

Venn Diagram

Total Sample size = 204

FW

NFW

(78 %)

(40 %)

ND 23 (11%)

FW: Collects fuel wood, NFW: Collects NTFPs other than fuel wood,

ND: Non-dependent on forest products for living.

60

29 %

22

11 %

99

49 %


Table 4: Nature of Biodiversity-linked Livelihood Practices at Birbhum

Sl.

No.

Type of

Livelihood

practices

% of local

people

involved

Pattern of usage Depended Species

1 Using as fuel

wood 77.94 Personal consumption Dry leaves and twigs of any plants

2 Making of Sal

Plates 31.37

Stitching plates from

leaves of sal trees for

domestic use and also

for sale

Sal trees (Shorea robusta)

3 Consumption of

Vegetables 20.1

For personal

consumption and also

for sale

Varieties of mushroom (35), bon-alu (2), kapu

alu (1), mahua flower (3)

4 Consumption of

fruits 14.22

Personal

Consumption

Mahua (22), piyal (5), khejur (1), mango (3),

jamun (3), jackfruit (1)

5 Using as

medicines 3.92 Personal consumption

Joripat (4), horitoki (14), kalmegh (45), satmul

(19), got (7), kundri (1), boincha (2), mugar dal

(3), bohora (2), vela (10), wild garlic (1), rahim

chhal (5), arjun chhal (2), thankuni (3), ananta

mul (3), baranga (1), horek kolai (2), kanaklata

(2), chirchira (1), chorbori (1), etc.

6

Other (making

broom, talai,

chatai, etc.)

5.39 Personal consumption Sar pata (2), kuchi grass (3), khejur pata (8)


Table 5: Degree of Forest Dependence (in percentage)

Nature of Forest FPC Names Low Medium High

Natural Forest

Bonsuli-Beloan 65 20 15

Jamboni 20 80 0

Bonvilla 10 70 20

Asna-Sundarkhele 40 50 10

Total 33.75 55.00 11.25

Mixed Forest

Laxmipur 50 31.82 18.18

Sultanpur 70 20 10

Total 60.00 25.91 14.09

Monoculture Plantation

Kheledanga adibasi 100 0 0

Lohagodd 100 0 0

Pachiara 100 0 0

Jethia-Rampur 100 0 0

Total 100.00 0.00 0.00

Total 65.69 26.96 7.35



Due to a long period of deforested state followed by monoculture plantation of alien species,

forest resources have turned out to be unsuitable and inadequate to guarantee sustainable

livelihood. So, the once forest-dependent villagers in these areas have been forced to reduce the

extent of forest dependence in their day to day living. Branches of Sal trees, which are auspicious

augury for tribal marriages and cremation ceremony, are nowadays available in markets to meet

such demand. Local people have even failed to recall the names of the extinct species as they

have been lost long. Thus the prolonged deforestation followed by monoculture plantation has

acted as push factor from inside the system which has increased the dependence of local people

on external economy. To isolate the marginal effect of these socio-economic factors on

livelihood dependence across different types of forestry, a livelihood dependency index has been

constructed and its relation with certain socio-economic factors is obtained by running simple

least square regression analyses. The construction of forest dependence index (dependent

variable) and its related independent variables is briefly described below.

The dependent variable Y, defined as the forest dependence index (FDI) is meant to capture the

extent of forest dependent livelihood in terms of the pattern of dependence. Since in all, six

different types of possible uses of forest has been identified, the maximum possible score of

dependence would be 6 and the minimum would be 0. Following the method of construction of

Human Development Index, the FDI can be proposed as

100x

reMaximumSco

eActualScor. This is a multi-

dimensional index5 and is constructed by combining responses to a set of questions related to

different aspects of forest dependence of household, viz., (i) both the type and variety of non-

timber forest produce (NTFP) collection, (ii) use of medicinal plants, (iii) use of grazing land and

(iv) direct financial dependence on forest. These variables are given equal weights for the

construction of the concerned index6.

The independent variables, Xs, are defined to represent certain socio-economic factors related to

family, society and the economy:

5 The minimum value of FDI is 0 and the maximum value is 60. The average value is 24.66 and the standard

deviation is 19.31. 6 FDI has also been constructed using Principal Component Analysis (PCA) where unequal weights are

endogenously determined for the same set of responses. The scatter plot suggested almost a near perfect correlation

between the Y-variables constructed by following these two alternative methods. Since PCA did not suggest any

significant improvement, we have taken the simple average based index of FDI as our dependent variable.


Family Factor:

• SIZE: Size of the family, which is expected to have a positive association with FDI;

Social factors:

• TRIBE: a dummy variable is defined to assume value 1 if ST and 0 otherwise, which is

expected to have a positive influence with FDI;

• MEMBER: a dummy is defined to assume value 1 for the households containing at least

one FPC member and 0 otherwise and the variable is expected to have positive influence

on FDI;

Economic factors:

• BPL: a dummy is defined to assume value 1 if the household holds BPL card and 0

otherwise. The expected sign of influence is positive;

• ASSET: an index of households assets is constructed by collating information on various

types of assets like van, cycle, bike, radio, fan, television, etc. owned by it and the

expected influence of this variable on FDI is negative;

• LVSTCK: an index of household’s livestock holding is constructed by collating

information on different types of livestock like buffalo, cow, goat, pig, poultry animal,

etc. owned by it and the expected influence of this variable on FDI is negative;

• OCCUP: a dummy is defined to assume value 1 for agriculture and forestry related

occupations and 0 otherwise with an expected positive sign;

Since the original forest dwellers were not much interested in the employment opportunities

created outside the forest niche, the extent of forest dependence is likely to have a negative

correspondence with the extent of participation in the Mahatma Gandhi National Rural

Employment Guarantee Scheme (NREGA).

• NREGA: Here for each household the three year’s average annual job-day data have

been collated from the official website7.

The study variable FDI has been regressed on all these independent variables to isolate marginal

influence of all factors and to study the possible variations of these causal relations across

7 Out of 204 households surveyed the matching works for 190 cased and for the remaining 14 cases the village

average is taken as a proxy estimate.


different types of forestlands, two forest-type dummies have been introduced additionally by

treating natural forest as the reference category.

• DMIXD: a dummy is defined to assume value 1 for mixed forest and 0 otherwise;

• DMONO: a dummy is defined to assume value 1 for monoculture forest and 0 otherwise;

The regression results are reported in Table 6. From Table 6 it is seen that initially, when no

distinction is made in terms of the forest-type, TRIBE and NREGA appear to be the only two

significant influences on FDI, the former with expected sign and the latter with an opposite sign.

So, in regression 2 the two forest-type dummies are incorporated and interestingly both of them

became statistically significant with predicted sign and lesser is the biodiversity component

weaker is the extent of forest dependence. However, besides these two dummy variables no other

explanatory factor was statistically significant, though the adjusted R2 increased substantially,

from 0.16 to 0.41, indicating the possible presence of multi-collinearity in the dataset. So, three

separate regressions were run on three different types of forestlands and reported as regression 3

(for natural forest), regression 4 (for mixed forest) and regression 5 (for monoculture plantation

forest).

Table 6: OLS Regression Results of Forest Dependence Index

X-Variables Y-variable: FDI

Reg_1 Reg_2 Reg_3

(Natural)

Reg_4

(Mixed)

Reg_5

(Mono)

SIZE 0.20 0.60 -- -- --

TRIBE 10.54*** 3.86 16.68*** -10.72 4.37**

MEMBER 0.28 0.93 -- -- 2.76*

BPL -2.48 1.90 6.57* -- --

ASSET -0.03 -0.07 -- -0.44** 0.11**

LVSTCK 0.04 0.06 -- 0.53** 0.05

OCCUP 2.71 1.62 5.22 -- --

NREGA 0.20*** 0.12 -- 0.44* 0.15**

DMIXD . -8.98*** . . .

DMONO . -23.72*** . . .

2R

0.16 0.41 0.12 0.17 0.33

DF 195 193 76 37 76

F 6.00*** 14.84*** 4.53** 3.04* 9.06*** Note: *** Significant at 1% level ** Significant at 5% level * Significant at 10% level

Source: Authors’ Calculations


For natural forest TRIBE is a very important determinant of FDI followed by economic need

(BPL). For mixed forests three important factors are ASSET, LVSTCK and NREGA while the

former with a negative sign suggests same kind of connection between livelihood dependence on

forest and economic status, i.e., greater dependence for the economically weaker section, as was

suggested by BPL in case of natural forest. For LVSTCK also FDI increases with the increase in

the demand for grazing land. The most interesting relation is indicated by the statistically

significant positive influence of NREGA on FDI. We have included NREGA as a variable to

capture the impact of general development opportunities available from outside the forest based

system on the pattern of forest dependence. Though a negative relation is expected in this case,

contrary to our expectation, here the coefficient is positive as well as statistically significant.

Before making any attempt to explain this result it would be better to note the situation

prevailing in case of monoculture forests, i.e., regression 5. In this case the variables turned out

to be statistically significant as explanatory factors are MEMBER, TRIBE, ASSET and NREGA

with all imparting favorable influence. Moreover, this regression has strongest F among all

forest-type specific regressions and the adjusted R2 is also quite high (0.33). Barring the

exception of TRIBE, all other regressors collectively indicate a newly emerging politico-

economic scenario. Those who are relatively affluent (with access to better ASSET) are enjoying

more secured social position, becoming the members of the FPC, utilizing benefits of greater

man-days in NREGA and accessing forest resources more in their daily living. So, instead of

forming a support for the poorer section, under monoculture plantation in the newly afforested

areas, forest development has become a parallel agenda of general development and forest

resources are getting treated at par with any other economic resource. This tendency is of deep

concern to us as livelihood support is an in-situ experience whereas economic resource is an ex-

situ entity that can always be separated from the whole system, transformed into monetized units

and transacted in the market. If forest resources are perceived as economic resources and forest

development is treated as any other development agenda, then the marketable part of forest

resource would enjoy priority and protection of biodiversity would be of very little importance.

In fact, to be effective support of livelihood the conservation of biodiversity is necessary but not

sufficient. For it to be useful the traditional knowledge about the use of this diverse resource

needs to be transmitted across generations. An attempt has been made to construct an index of

retention of this traditional knowledge that is going to be reported in the following sub-section.


4.3 Flow of Traditional Knowledge

Traditional Knowledge (TK) can be defined as “the knowledge base acquired by indigenous and

local people over hundreds of years through direct experience and contact with the

environment.” It is the product of keen observation, patience, experimentation, and long-term

relationships with plants and animals (http://npdc.usda.gov). Its importance for the protection of

biodiversity and the achievement of sustainable development is slowly being recognized

internationally and the Convention on Biological Diversity of 1992 has acknowledged these

contributions of traditional knowledge and therefore included laws pertaining to its access and

use (Mazzocchi, 2006).

Many studies like Folke (2004), Beamer (2009) and Johannes (1993) have shown several

innovative ways to incorporate traditional knowledge into modern practices of ecosystem

management. Some have solely focused on the integration of TK into forest management. Rist et

al. (2010) have discussed the merit of combining TK with scientific data to achieve improvement

in forest management in BRT Wildlife Sanctuary of South India where such an application

helped managing the mistletoe infection in Amla tree (with serious bio-diversity and livelihood

impacts) more efficiently. Butler (2010) also discussed how the tribes in Suriname, Brazil and

Colombia are combining their traditional knowledge of the rainforest with western technology to

conserve forests and maintain ties with their history and cultural traditions, which include

profound knowledge of the forest ecosystem and medicinal plants. Thakali and Lesko (1998)

have described some important contributions that several American Indian tribes have recently

made, applying their traditional knowledge to the management of forest resources in the United

States. Furthermore, Charnley et al. (2008) have paid attention to the ecological knowledge of

three local groups who inhabit in the Pacific Northwest region and have pointed out that

integration of traditional and local ecological knowledge into forest biodiversity conservation is

most likely to be successful if the knowledge holders are directly engaged with forest managers

and western scientists in on-the-ground projects in which interaction and knowledge sharing

would be facilitated.

Although all these works have become successful in stitching the local traditional knowledge

with the modern practices of managing ecosystems, there is lack of any methodological

intervention from accounting perspective to assess the quantitative significance of such


knowledge conservation. Rarely documented, traditional knowledge is built on a history; gained

through many generations of human beings teaching their children practical techniques that

underscored this crucial human-environmental relationship upon which culture and life itself

depended (http://npdc.usda.gov). So to check whether traditional knowledge has been conserved

or not, one needs to look into its flow from one generation to the next. Construction of index for

traditional knowledge is apt for such purpose and would be a unique addition to existing

methodological approaches. Here the analysis has been carried out at two stages: (i) first we have

considered the case of complete knowledge transfer and defined the traditional knowledge flow

index TKNWI as a dummy variable and (ii) next we have defined another variable TTK

representing the extent of knowledge transmission as an ordered categorical variable to

accommodate the possibility of complete as well as partial knowledge transmission.

The dependent variable defined as the traditional knowledge flow index (TKNWI) is constructed

with variables representing the transmission of traditional knowledge within the household

across generations. Three specific questions were asked to each interviewee:

(a) Whether they have acquired this knowledge from their ancestors,

(b) Whether they are passing it on to their next generation and

(c) Whether the posterity is taking any interest in this information.

If all the answers are in the affirmative then the traditional knowledge is considered to be

conserved and conservation is thought to be inadequate when at least one answer is negative,

indicating break in the chain of intergenerational knowledge transmission.

TKNWI: a dummy variable is defined to assume a value 1 for knowledge conservation and 0

otherwise.

This is our dependent variable and, since it is binary by construction we may identify different

independent variables which are likely to influence the probability of TKNWI conservation, i.e.,

[ ]1== TKNWIPPi and a LOGIT regression can be run to find out the corresponding odd-ratios

and marginal effects8.

8 The odd-ratio with respect to the explanatory variable jX can be defined as ( )ji

jij

P

POR

−=

1and the corresponding

marginal effect is ( )[ ]jijijji

jiPP

X

P−= 1β

δ

δ. In this case unlike OLS regression the marginal effect is contingent on

both i-th observation and j-th regressor, indicating presence of a built-in non-linearity.


The probable causal influences are defined to capture the following dimensions: (i) forest-based

livelihood dependence, (ii) general awareness related to alternative land use pattern of forestland,

(iii) improvement in forest status after implementation of JFM and (iv) perception of the

household regarding different community forest based rituals. Consequently variables are

defined as:

PFDI: This is the predicted value of FDI obtained from the previous exercise. Since the

dependence on forest for livelihood (FDI) and the conservation of traditional knowledge

(TKNWI) are jointly endogenous, to avoid the problem of endogeneity we have replaced FDI

with its predicted value (PFDI), which served as a suitable instrumental variable in this context.

AWRI: An index of awareness regarding the felt importance of forest conservation has been

constructed on the basis of binary response on different dimensions of the land use related

problems faced by forestland and collated into a single measure by following the methodology of

human development index.

JFM: It is an index based on the pattern of change in forest conditions that has been experienced

since the formation of FPC. The constituent indicators of the index are: (a) area under forest, (b)

number of tree species, (c) tree on private land, (d) number of ponds within the forest and in the

fringe area, (e) availability of NTFP, (f) time to be spent for the collection of fuel, fodder and

leaf litter and (g) protection against flooding/landslides. In each case the improvement has been

assigned a score 1 and 0 otherwise and the individual scores have been aggregated and

normalized to produce the required index.

RITUL: This index is based upon the changes occurring in the importance of traditional forest-

based rituals like (a) collective hunting of animals, (b) collective gathering of fruits, fuel, fodder,

medicinal plants and other NTFP, (c) rituals related to the grazing of domestic animals. For

decrease in the importance of each of the rituals we have assigned value 0 while value 1 is

assigned otherwise. Then using the same methodology the forest ritual index is constructed.


Finally, a LOGIT regression is run and the marginal effects (evaluated at mean) are reported in

Table 7. To give control to the nature of biodiversity, the two forest-type dummies defined

earlier, viz., DMIXD and DMONO have been incorporated.

;6543210 iiiiii uDMONODMIXDRITULJFMAWRIPFDITKNWI +++++++= βββββββ

where ui follows a logistic distribution: i

i

x

x

ie

euf

βα

βα

+

+

+=

1)( ; this gives the probability equation

−=

i

ii

P

PTKNWI

1ln ;

Table 7: LOGIT Regression Results

X- Variables

LOGIT (Marginal Effect)

Y: TKNW

All Natural Mixed Monoculture

PFDI 0.005 -0.008 0.011 0.007

AWRI 0.004** 0.003 -0.002 0.017**

JFM 0.004** -0.001 0.005 0.009**

RITUL 0.002 0.006** 0.004 0.000

DMIXD 0.136 -- -- --

DMONO 0.109 -- -- --

Pseudo R2 0.05 0.07 0.11 0.18

No. Of Obs. 204 80 42 82

LR Chi2 14.05** 7.52 6.53 20.52***

Note: *** Significant at 1% level ** Significant at 5% level * Significant at 10% level


When all responses are considered together and the forest-type dummies are incorporated, the

marginal effect of only two factors, AWRI and JFM, appear to be statistically significant. This

indicates that the success of the joint forest management program generates interest among the

local people to conserve the traditional knowledge about uses and abuses of biodiversity. For

deeper investigation we have run the same regression separately on three different types of forest

areas. The results turn out to be quite interesting.


Nowhere PFDI, i.e., the livelihood dependence on forest plays any statistically significant role in

conserving traditional knowledge. Effectively, TKNWI and PFDI behave like two disjoint

factors. In case of natural forest the only important variable having significant marginal effect is

RITUL. For mixed forestry no definite relationship has been suggested by our field observations

and the strongest results are obtained for monoculture plantation where the success of JFM and

general awareness about forest land use pattern appeared to be the most important. This shows

that the overall pattern is also dominated by the monoculture forests and what adds to this

discomfort is the fact that all respondents from this area admitted that they spent the proceeds of

timber sale obtained as FPC-members in personal consumption and not to promote any

community purpose. So, the incentive scheme adopted by the JFM is merely an economic

incentive that reduces forest resource into an ordinary economic resource, which places a high

weight on use the value of a product or service and generally fails to assess the existence value in

a non-market frame.

To explore this issue further, we have made a distinction between complete transmission vis-à-

vis partial transmission of traditional knowledge and defined another study variable TTK

(transmission of traditional knowledge), which is an ordinal (ordered) variable assuming values

1, 2, 3 & 4, according to the extent of knowledge transmission. If answer to all the relevant

questions are negative or only one is affirmative, there is no evidence of knowledge

transmission. So, here the assigned value of TTK is 1 (poor). If answer to each question is YES,

then there is perfect knowledge transmission, and the value of TTK is 4 (excellent). In between

there are two possibilities: (a) answer to the first two queries is affirmative and that of the last

one is negative and (b) answer to the first one is negative but that of the following two are

positive. In the former case the knowledge is flowing from the past to the present generation and

in the latter case it is passing from the present to the future generation. In terms of traditional

knowledge conservation the first case is more potent as here the future generation may gain

motivation at a future date in acquiring traditional knowledge and may make the transmission

process complete whereas the second case does not have any chance to lead to complete

knowledge transmission even at any future date. So, a score of 3 is assigned to the first case


(good) and that of 2 (moderate) is assigned to the other one where the traditional knowledge is

acquired from any formal exogenous source and not necessarily from within the community.

Thus, TTK is an ordered categorical variable on which an ordered LOGIT can be applied to

understand the prospect and probability of TK conservation across different types of forestlands.

In ordered LOGIT, an underlying score is estimated as a linear function of the independent

variables and a set of cut-points (threshold parameters). The probability of observing outcome i

corresponds to the probability that the estimated linear function, plus random error, is within the

range of the cut-points estimated for the outcome and is specified as:

),.....Pr()Pr( 111 ijkjkjiijj uxxpioutcome αββα ≤+++<=== − where uj has a logistic

distribution and there are k number of possible outcomes in all.

)exp(1

1

)exp(1

1

1 βαβα jijiij

xxp

+−+−

+−+=

−

.

To be consistent with the latent variable interpretation of this discrete ordering, α0 is taken as -∞

and αk is taken as +∞. Since we have 4 possible outcomes, there will be 3 cut-points9.

Summation of Pi’s would be unity for all j’s. Our prediction of y is simply the outcome with the

highest probability (Woolridge 2002).

Table 8: Ordered LOGIT Regression Results

X- Variables

OLOGIT (Odd-Ratio)

Y: TTK

All Natural Mixed Monoculture

PFDI 0.980 0.985 0.957 1.011

AWRI 1.019*** 1.020* 1.015 1.032**

JFM 1.012** 0.995 0.997 1.042***

RITUL 1.000 1.018** 0.998 1.004

DMIXD 1.092 -- -- --

DMONO 0.485 -- -- --

CUT_1 0.348 0.590 -1.445 3.786

9 In fact, the standard binary LOGIT model has a single cut-point.


CUT_2 0.929 1.231 -0.679 4.318

CUT_3 1.481 1.834 0.051 4.808

Pseudo R2 0.031 0.039 0.034 0.115

No. Of Obs. 204 80 42 82

LR Chi2 15.31** 7.6* 3.57 22.05***

Note: *** Significant at 1% level ** Significant at 5% level * Significant at 10% level


Table 8 reports the ordered LOGIT results for overall Birbhum Forest Division along with the

areas with Natural forest, mixed forest and Monoculture plantation and the respective cut-points

for different levels of TTK. Similar to the case of TKNWI, in case of TTK also the overall

pattern of knowledge transmission probability is mostly matching with that of the Monoculture

plantation or the newly afforested areas. The statistically significant predictors are also not much

different, with a few notable variations: (a) in case of Natural forests the AWRI variable has

gained statistical importance in influencing the probability of TKK favorably, (b) the LR(χ2) of

this regression becomes statistically significant at less than 10% level and (c) all the regressions

improved statistically, indicating a distinct improvement in the predicted outcome.

The threshold levels reported in Table-8 deserves some special attention. No distinct pattern is

observed for the mixed forest area; though there is striking difference between the cut-points of

natural vis-à-vis monoculture forests. For all categories the cut-off values are much higher in

case of monoculture areas indicating a generally weaker possibility of traditional knowledge

transmission there. It is important to note here that while the direction of the effect of xk on the

probabilities P(y=1|x) and P(y=4|x) is unambiguously determined by the sign of βk, the sign of

βk does not always determine the direction of the effect for the intermediate outcomes (y = 2 & 3

here). Table 9 (a) & (b) report the observed frequency and the predicted probability of most

likely outcome of TTK for different types of forests. Though it is found that across all forest

groups there are a significant number of observations (almost 25 percent of the total) that fall

under category ‘moderate’ and ‘good’ when the frequency table is derived on the basis of sample

responses [Table 9(a)]. However, none of these observations with intermediate values have

retained their status in the model based predicted values [reported in Table 9(b)]. There a


tendency towards polarization is noted. Compared to the sample based estimation the model

based prediction shows much better prospect for knowledge transmission (0.49 and 0.89

respectively) and though the sample observations did not suggest much difference between

natural and monoculture forests in this regard, the model based predicted difference is 0.31.

Thus, it is expected that the newly afforested areas would be 31 percent weaker than natural

forest in conserving traditional knowledge.

Table 9(a): Sample frequency and predicted frequency of the level of traditional knowledge

conservation across forest types

Values Tabulate OLOGIT

Trad_Know All Natural Mixed Monoculture All Natural Mixed Monoculture

Poor 54 19 8 27 22 10 4 35

Moderate 24 10 6 8 0 0 0 0

Good 26 11 7 8 0 0 0 0

Excellent 100 40 21 39 182 70 38 47

Total 204 80 42 82 204 80 42 82

Source: Primary Survey, 2012 (Authors’ Calculations)

Table 9(b): Sample proportion and model based prediction of the probability of conservation of

traditional knowledge across forest types

Trad_Know Tabulate OLOGIT

All Natural Mixed Monoculture All Natural Mixed Monoculture

Poor 0.26 0.24 0.19 0.33 0.11 0.13 0.10 0.43

Moderate 0.12 0.13 0.14 0.10 0.00 0.00 0.00 0.00

Good 0.13 0.14 0.17 0.10 0.00 0.00 0.00 0.00

Excellent 0.49 0.50 0.50 0.48 0.89 0.88 0.90 0.57

Total 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

Source: Primary Survey, 2012 (Authors’ Calculations)


5. Discussion

An important implication of this study is the presence of cultural migration, which is pulling the

forest dwelling tribal people out of their natural habitat and community living and forcing them

to be a part of the outside monetized economy governed by the market incentives. The tribals

now take money as the only compensation to do any work instead of once practiced barter

system. Getting formal education from mainstream institutions, they have started socializing

outside their own communities and mould themselves for a better career in the outside world.

Money and jewelries have also cribbed as dowry into their marriage system in place of country

liquor and domestic animals and all legal issues are settled in courts or at police station as an

alternative for solution through the intervention of morol10

of a community. This cultural

transition gains momentum through the targeted socio-economic programs on education, health,

poverty alleviation to further intensify the pull effect. Even if someone has the information of

kabiraji11

flora/ fauna, there is no more reliance on such traditional line of herbal treatment due

to enhanced health awareness among local people and availability of modern treatment facilities.

It is noticed that though 89 percent of the sampled households depend on forest for their daily

needs, only 4 percent of them collect medicinal plants from it. So the indicator for livelihood

dependence on forest is not found to be significant anywhere when regressed for

intergenerational transmission of traditional knowledge. We have incorporated participation in

general development programs (like 100-days work in MGNREGA, etc.) in our regression

analysis and statistically significant results have been obtained to indicate the eradication of

difference between livelihood development policies in in-situ environment vis-à-vis economic

development policies in ex-situ environment. This transformation dampens the local households’

involvement and commitment towards forest protection irrespective of their level of awareness

regarding the long-term consequences.

In fact, the pull factor is closely connected with the push factor which, in its turn is the

consequence of rampant deforestation and rapid afforestation through monoculture plantation of

exotic species. There is a definite change in the nature of demand for forest products from

10

Morol is the head of a village/community, who has a final say on any judgment decisions. 11

Kabiraji is a traditional practice of Ayurveda in India.


outside world and this change is connected with the technological development. Nowadays

strong wooden poles are not that much demanded by the railways, mining and quarrying

industries, construction sector and others; instead the demand for paper has gone up manifold.

This has resulted in stagnant demand for many traditional timber products like sal (Shorea

robusta) which supply wood with discontinued fibers and there is vibrant claim for short and

uniform fiber trees like eucalyptus, sonajhuri, etc (Nanko, et al., 2005). This indicates a serious

compromise with biodiversity and creating a push effect for the traditionally forest dependent

people to culturally migrate to the outside world to satisfy their daily needs.

The local level practices are needed to be stitched with the national policies incorporating the

effects of modern development agenda. Pandey (retrieved from

http://www.infinityfoundation.com/mandala/t_es/t_es_pande_conserve.htm) has also indicated

the necessity of embedding local and traditional knowledge into the doctrines of formal science

to understand their implications for sustainable forest management. In the absence of such

theorization, it is impossible to preserve the traditional wisdom in a generalized framework and

enhance its applicability.

References

Beamer, E. (2009): Thoughts on Traditional Ecological Knowledge, Values, and Salmon

Management, Presented at PSGB Ecosystem Conference, February 10, 2009, Seattle,

WA

Butler, R. (2010): Protecting their Rainforest: Amazon Tribes Embrace Technology to Save

Land, Culture. [Online] http://www.geoplace.com//ME2/dirmod.asp?sid=.....

(Accessed 10th

June 2011)

Charnley, S., Fischer, A. P. and Jones, E. T. (2008): Traditional and Local Ecological

Knowledge about Forest Biodiversity in the Pacific Northwest, General Technical

Report, PNW-GTR-751, April 2008, United States Department of Agriculture


De, U. K. (2012): Livelihood, Dependence on Forest and Its Degradation: Evidence from

Meghalaya, Environment and Natural Resources Research 2 (3), 97 – 114

Debnath, G. C. & Mondal, P. (2014): Effects of Ecogeomorphological Parameters on

Environment: Case Study of Birbhum District, West Bengal, International Journal of

Innovative Research and Studies 3 (3), 655 – 667

Directorate of M & SSE. (2013). [Online] http://www.mssewb.org/htm/key_personnel_1.html

(Accessed 24th June 2013)

Folke, C. (2004): Traditional knowledge in social–ecological systems, Ecology and Society 9(3):

7. [Online] URL: http://www.ecologyandsociety.org/vol9/iss3/art7/ (Accessed 10th

January, 2015)

Ghosal S. (2011): Pre‐Colonial and Colonial Forest Culture in the Presidency of Bengal, Journal

of Studies and Research in Human Geography 5 (1), 107‐116

Human Development Report (2009). [Online]

http://wbplan.gov.in/htm/humandev/docs/DHDR_Birbhum/birbhum_main.htm

(Accessed 15th June 2012)

Johannes, R.E. (1993): Integrating Traditional Ecological Knowledge and Management with

Environmental Impact Assessment, Traditional Ecological Knowledge Concepts and

Cases, edited by Julian T. Inglis, International Program on Traditional Ecological

Knowledge and International Development Research Centre. Chapter 4, 33 – 40

Mazzocchi, F. (2006): Western science and traditional knowledge: Despite their variations,

different forms of knowledge can learn from each other, Viewpoint, EMBO Reports,

7 (5)

McDermott, C. L. (2012): Plantations and Communities: Key Controversies and Trends in

Certification Standards, FSC Certified Plantations and Local Communities workshop:

Overview paper, Forest Stewardship Council

Nanko, H., Button, A. and Hillman, D. (2005): The World of Market Pulp, Appleton, WI, USA:

WOMP, LLC, 107–109


Padel, F. and Das, S. (2010): Out of This Earth: East India Adivasis and The Aluminium Cartel,

Orient Blackswan Pvt Ltd

Pandey, D. N.: Traditional Knowledge Systems for Biodiversity Conservation. [Online]

http://www.infinityfoundation.com/mandala/t_es/t_es_pande_conserve.htm

(Accessed 9th August 2014)

Pandey, R. (2010): Heterogeneity in Household Characteristics, Forest Resource Utilization and

Sustainability in Hills of Uttaranchal: A Case Study, Silva Lusitana 18 (1), 75 – 84

Rist, L., Shaanker, R. U., Gulland, EJ. M., Ghazoul, J, (2010): Combining Traditional Ecological

Knowledge and Conventional Scientific Data in Forest Management, Traditional

Knowledge Bulletin, June

Robson, J. P. and Nayak, P. K. (2010): Rural Out-Migration and Resource-Dependent

Communities in Mexico and India, Population Environment 32, 263 – 284

Rojas, I. (2012): What's wrong with tree plantations? [Online]

http://print.news.mongabay.com/2012/0921-plantations-commentary-rojas.html.

(Accessed 4th March 2014)

Sankaran, K., Sinha, S. and Madhav, R. (2002): Forests Background Note, WIEGO Law Pilot

Project on the Informal Economy. [Online]

http://wiego.org/sites/wiego.org/files/resources/files/fow_background_note.pdf

(Accessed 8th

May 2012)

State of Environment Report, India. (2009): Environmental Information System (ENVIS),

Ministry of Environment & Forests Government of India. [Online]

http://www.moef.gov.in, http://envfor.nic.in. (Accessed 10th March 2014)

Tagore, R. (1915): Sadhana – The Realization of Life, The Macmillan Co., New York

Thakali, R. and Lesko, L. (1998): Wisdom of the Ages: Traditional Knowledge and Forest

Ecosystems. [Online]

http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb5163026.pdf,

(Accessed 1st June 2011)


Traditional Ecological Knowledge: An Important Facet of Natural Resources Conservation,

Traditional Ecological Knowledge - Technical Note 1, National Plant Data Center,

[Online] http://npdc.usda.gov (Accessed 9th

January, 2015)

Woolridge, J.M. (2002): Economic Analysis of Cross Section and Panel Data, The MIT Press,

USA

Livelihood Dependence, Traditional Knowledge Conservation ...

Documents

Transcript of Livelihood Dependence, Traditional Knowledge Conservation ...