Socio-economic contributions and determinants of locals engagement with Boswellia papyrifera in lowland woodlands in Burie Zuria district, Ethiopia

Abstract

Ethiopia's lowland woodlands are comprised of the major gum and resin-producing genera Acacia, Boswellia, and Commiphora. Boswellia papyrifera is primarily found in the degraded drylands, Burie Zuria district is the existence of the species; however, limited information is available on the woodland socio-economic contribution and determinants of locals' use of woodlands for their livelihoods, particularly Boswellia papyrifera. So to fill this gap, the study was conducted to examine the locals' socio-economic benefits from the woodland and to identify the determinants of locals' engagement in the collection of the benefits for better wise use and conservation policy implementation. The data was collected from household interviews, focus group discussions, and key informants. Species' socio-economic benefits were analyzed through descriptive statistics, whereas determinants of local's willingness to use woodlands were analyzed through an econometric model. The dependent variable used was annual income gained from the woodland and explanatory variables taken were gender, marital status, livestock number, family size, land holding size, age of the respondent, education level, wealth status, perception of local communities towards forest cover change, and distance from woodlands to the household's residence. Based on the results, the total annual income derived from the woodland by the sampled households' was a mean of 1759.45 USD. The result shows the woodland species shares 14.37 % of the total annual income, but because of legal investors Boswellia papyrifera had no contribution to this share. However, the local communities use woodlands for their livelihoods due to explanatory variables, and the usage percentage shows significant differences. The multiple linear regression results showed that the model was significant at a 5 % probability level, and 95.9 % of the variation was due to the explanatory variables. Age of respondent, family size, total land holding size, and education level showed positive and statistically significant relations to annual income derived from the woodland while other remaining variables showed non-significant. The result concludes that the species contributes significantly to local communities' livelihoods and engagement in gathering timber and non-timber products. The study recommended that awareness about the use of non-timber products in addition to forest products be created among the locals, and policies related to woodland sustainability and conservation should also be implemented.

1. Introduction

In Africa, Comberutem-Terminalia trees have numerous economic and ecological benefits. Gums and resins give significantly different purposes to the national economy and neighborhood workers of many people in Africa [1]. In the dry land areas of Sub-Saharan Africa, oleo-gum resins like frankincense, myrrh, gum Arabic, and opopanax are produced by indigenous woody species like the genera Boswellia, Commiphora, and Acacia [1,2]. Sub-Saharan African countries, including Ethiopia, Sudan, and Eritrea, get a lot of their national economies from the sale of oleo-gum resin products worldwide [3,4]. According to Ref. [2], in 2014, US$500 million was traded worldwide.

The species Boswellia papyrifera (B. papyrifera) is mainly found in the degraded drylands of Ethiopia within an altitudinal range of 950–1800 m a.s.l. with an annual rainfall of less than 900 mm and an average temperature between 20 and 27 °C [5,6]. It provides several economic and ecological benefits to the country [7]. Many regions of the nation utilize this species for family furniture, walls, meds, apiculture, soil and water preservation, creature feed, fumigation, and different purposes, as well as adapting to the impacts of environmental change [1,7,8]. In addition, we can see an economic use; between 1995 and 1999 the country exported 7728 metric tons of frankincense [7]. Also, in 2014, the country exported about 7,900 tons of frankincense with a value of US$8.8 million. This makes the country a significant frankincense producer and supplier to the world market and shows how much the species contributes to the country's economy [9].

The species is scattered throughout the whole nation of Ethiopia such as Tigray, Amhara, Oromia, Benshangul Gumz, and the Afar regions [4]. The Amhara region was the second-most dominant area of the species existence, whereas the Tigray region was placed first [5]. Dryland areas of the Amhara region, where B. Papyrifera occurs in Metema, Jawi, Adi-Arkay, Abergelle, Debre Elias, Womberma, and Burie Zuria districts, while the species exists in this area, its socio-economic importance is not studied in the whole district.

In Ethiopia, a lot of research was done on the species' socio-economy importance but except for [6], all reported that the species contributes to the local communities' livelihoods through employment creation, and fodder for livestock, agricultural implements, medicinal purposes, and house construction. Besides, a lot of studies were conducted on the benefits of B. papyrifera-dominated woodlands to the local communities but most studies didn't address the determinant of locals' engagement in the collection of forest and NTF products from B. papyrifera-dominated woodlands. Burie Zuria district is one of the existing areas of the species however limited information is available on B. papyrifera dominated woodland socio-economic contribution and determinants' of locals to use woodlands for their livelihoods, particularly on B. papyrifera. Therefore, this study tries to examine the local communities' benefits from the woodland and the determinants of locals' engagement in the collection of forest and non-timber forest (NTF) products from the woodland because a better understanding of locals benefits from the woodland and determinants of household's engagement to collect forest and NTF products, suggest appropriate usage policy, to guarantee their sustainable use and conservation strategies. In general, the objectives of the study were:

• ➢To assess the local communities' socio-economic benefits getting from the woodland
• ➢To quantify the contribution of major forest and NTF products to forest-related income
• ➢To identify the determinants of locals' engagement in the collection of forest and NTF products.

2. Materials and methods

2.1. Description of the study area

The study was conducted in Abay Gauge, Burie Zuria district of West Gojjam, Amhara National Regional State, Ethiopia (Fig. 1). Located at 10° 15Ꞌ 2ꞋꞋ and10° 42Ꞌ 29ꞋꞋ north latitude, and 36° 52Ꞌ 1ꞋꞋ and 37° 7Ꞌ 9ꞋꞋ east longitude, about 411 km North West of Addis Ababa. The district agro-climatic or traditional thermal zones are Weyna-dega and Kola. The mean annual temperature ranges from 14°C to 24°C, and the mean annual rainfall ranges from 1386 to 1757 mm [10].

Fig. 1.

Map of the study area.

This study was conducted in the kola (dry land) parts of the district: Boko-tabo kebele (Fig. 1). The kebele relative location is between 10° 27Ꞌ 0ꞋꞋ to 10° 67Ꞌ 14ꞋꞋ north latitude and 36° 52Ꞌ 0ꞋꞋ to 37⁰ 02Ꞌ 11ꞋꞋ east longitude [11]. The vegetation cover of the study area was 32 % of the total area. The land use pattern in the kebele consisted of about 66.5 % cultivated land, 27 % natural forest, 2.5 % plantation forest and 5 % are others. The area has different species of trees, shrubs and vegetation types. The most dominants are Acacia Commiphora and Combretum-Terminalia woodland such as Anogeissus leiocarpus, Acacia seiberiana, Lannea schimperi, Apodytes dimidiata, Acacia senegal, Ziziphus mucronata, Erythroxylum fischeri, Diospyros abyssinica and B. papyrifera [11]. Farming system of the community in the study kebele was mixed farming system in which both agricultural crop production and animal rearing are combined [11]. The maximum hectares cover in permanent crops such as teff, maize, pepper, Boleke, Sesame, Masho, and the types of animals reared in the area are cattle, sheep, goat, chicken and donkey [11]. Majority of their economies are come from the sale of their crop products and animals [11]. The communities had no separate communal grazing lands whereas they used natural forest areas for these purpose [11].

2.2. Study design

2.2.1. Sampling design and sampling size determination

In this study, a multi-stage sampling procedure was used. Of the all districts of the West Gojjam zone, just Burie Zuria district and Womberma district are the presence of the most important and jeopardized species (B. papyrifera). Because of the two districts' agricultural Office information, in the Burie Zuria district, the impacts of anthropogenic aggravation and livelihood dependence on forests were greater than Womberma district so in the first stage, the Burie Zuria district was selected purposively.

In the subsequent stage, Boko-Tabo kebele was chosen purposively from 18 kebeles of the Burie Zuria district: because of the presence of the species and the neighborhood local area's livelihood base. In the third stage, six key informants and six focus group discussions (FGDs) were chosen through the snowballed sampling technique. In the fourth stage, from the entire (10) villages of the kebele six villages were chosen purposively, to stay away from misjudge and underestimate, the selection criteria was closeness to the woodland (three towns are closest and three towns are farthest). In the fifth stage, family heads from the chosen villages were chosen by a simple random sampling method. The number of individual family respondents from the chosen villages was determined based on Yamane's formula at a 95 % degree of certainty and 0.05 % level of variability with an 8 % level of perceptions (because of cash limitations and time shortage)

$$\mathbf{n}=\frac{\mathit{N}}{\mathbf{1}+\mathit{N}(\mathit{e})\mathbf{2}}$$

$$\mathbf{n}=\frac{\mathbf{429}}{\mathbf{1}+\mathbf{429}(\mathbf{0.08})\mathbf{2}}$$

$$\mathrm{n}=114.34\sim 114$$

where, n = the required total sample size of respondents.

N = total households in all sampled villages

e = precision level is 8 %

1 = the probability of an event occurring.

Through Yamane's formula, 114 numbers of individual households were used for interviews from overall households (429 numbers) of each village. The total household numbers of Boko-village, Kilo-Village, Jakot-Village, Kerkero-Village, Wendevo-Village and Merkato-Village were 45, 56, 49, 64, 87, and 218 respectively. Finally, sample households were selected from the above villages for interview by using a proportionate sampling technique (Table 1).

Table 1.

Distribution of the sample household heads in the selected Villages.

Villages	Total number of households	Percentage of sample respondents in each Village (%)	Number of sample respondents in each village
Boko-village	45	10.5	12
Jakot- Village	49	11.4	13
Kilo- Village	56	13.1	15
Kerkero- Village	64	14.9	17
Merkato- Village	128	29.8	34
Wendevo- Village	87	20.3	23
Total	429	100	114

Source: Boko-tabo kebele agricultural office, 2023.

2.2.2. Data collection method

The data was collected from January 25 to February 28, 2023. The data were collected from field observation, household interviews, key informant interviews, and FGDs.

2.2.2.1. Household interview

For the household survey, both structured and semi-structured questionnaires were completed. The point of the survey was to collect both qualitative and quantitative data [5]. The questionnaires were comprised of such major issues socio-demographic characteristics (sex, age, family size, wealth status, education level, and marital status), their assets such as land, crop products and livestock, livelihood activities, kinds and quantities of products extracted from the woodlands and their perceptions towards the woodland species cover change. The questionnaires were written in an English language but it was interpreted into Amharic language. Three local enumerators were recruited from the perspective of sample villages. All of them were fluent speakers respective with the local language and they were trained ability for data collection procedures, interview techniques, and the detailed contents of the questionnaire training.

2.2.2.2. Key informants

To get accurate information for each objective of the study six key informants were used. Both structured and semi-structured interviews were undertaken. A major points raised for key informants include several issues related to the livelihood of local communities with forest relationship; to get information on the amount of oleo-gum resins collected, the percentage share of the income generated in the forests for household subsistence annually, other relevant socio-economic information, and general information about the forest [12].

2.2.2.3. FGDs

Six FGDs range from 6 to 8 members, and a total of 40 members were involved; such members include village elders, young females, and males, village leaders, and formal village committees. Major points raised for discussion were regarding livelihood activities; information on the amount of oleo-gum resins collected the percentage share of the income generated in the forests for household subsistence annually, type of forest and NTF products they used for their livelihoods, and other relevant socio-economic information and general information about the forest [12]. The information obtained from FGDs was used to check and conifer the data collected from the household's interview.

2.2.2.4. Calculation of income

Income is the return from labor and capital that the household owns, uses in its production and income-generating activities (self-employed or company) or sells on the market. According to Ref. [13], the total income so the household was determined by different fields of activity as the sum of household income from these activities. Total household income was calculated as the net income of the household sample in seven main categories [14]: crop income, forest-related income, business income, livestock income, wage income, other income, and non-forest environmental income. However, the only definitions used in this study for these categories are given below:

Annual total household income = ∑ (Crop income + forestry income + commercial income + livestock income + wage income + other income).

Y = ∑ni = 1 Xi Eq1

where Y is total annual household income and Xi is income from source i.

Crop Income: The sum of the yield value of the various crops grown in the household, minus production costs. The total crop yield was calculated as follows:

YC = ∑nn = I [CiPi − Ki], Eq2

where YC is the total crop yield, Ci is the yield of crop i, Pi is the market price of crop i, and Ki is the production cost of harvest i.

Forest-related income includes the sale or consumption of plant or animal products harvested from or grown in the natural forest on designated forest land, as well as payment for forest environmental services.

Yf = ∑n = en [FiPi − Ki ], Eq3

where Yf is the total forest revenue, Fi is the collected amount of product i, Pi is the market price of forest product i and Ki is the production cost of forest product i.

Business income includes financial income from business, but not income from domestic agricultural or forestry production and processing. Livestock income includes only income from the sale or use of livestock.

Yl = ∑n = in [NiPi − Ki] + ∑nn = I [QiPi − Ki], Eq4

where Yl is the total amount of cattle income, Ni is the number of animals belonging to class i, Qi is the amount of animal product obtained from livestock i, Pi is the market price of livestock i, and Ki is the financial cost of keeping livestock i, such as wages shepherds medicine and fodder.

Wage income includes money from all paid work and other income includes the daily wages of workers and others.

2.2.2.5. Wealth status

The value of all the valuable assets that an individual, group of people, business, or nation owns is measured as wealth [15]. It is possible to envision compiling a list of all assets (cash and tangible), valuing each one according to market standards, depreciating it, and adding up the totals. Debts can be handled in the same way, and net assets can be calculated by deducting the debts from the assets [15]. The term “wealth” is used in this study to refer to financial resources in the form of assets and liabilities. In order to analyze wealth and income, the population is divided into three levels using the demographic and health survey wealth index score and the World Bank poverty headcount thresholds:

• 1.According to this study, people who make less than $1.90 per day are considered as poor.
• 2.In this study, people making between $1.90 and $5.50 per day are classified as medium.
• 3.The World Bank defines rich people as those who earn more than $5.50 per day.

Demographic and health survey data is weighted by household size, and all household members are assigned to one of the three income levels based on the percentage of the population that lives at each level, assuming that the ordering of households based on wealth was similar to the ordering of households based on income.

2.2.3. Data analysis

Socio-economic data was summarized, discussed, and calculated by using descriptive statistics (range, mean, standard deviation, maximum, minimum, and percentage) through tables, bar graphs, and pie charts. The data was performed using MS Excel 2010 and IBMSPSS version 27 software.

Determinants’ of locals' engagement in the collection of forest and non-timber forest (NTF) products were analyzed through a multiple linear regression model using IBMSPSS version 27 software.

2.2.4. Econometrics model

Before deciding to use a regression model, however, the data set was checked carefully. As a result, the data clearly showed that the entire household gained at least a minimum benefit from the woodlands to their livelihood but their usage amounts vary. So the dependent variable in this research was annual income gained from the woodland for their subsistence however unit of measurement of fuel wood, grazing land, farming utility, handy craft, medicine, and wild edible fruit are different so to overcome this difference, all of them were changed to birr. Explanatory variables taken in this study were the sex of the respondent, marital status, livestock number, family size, land holding size, age of the respondent, education level, wealth status, perception of local communities towards forest cover change, and distance from woodlands to the household's residence.

The following econometric model was employed to know the determinants:

$$\mathrm{Y}={\mathrm{B}}_0+{\mathrm{B}}_{\mathrm{i}}{\mathrm{X}}_{\mathrm{i}}+\pounds$$

where: Y = total annual income derived from woodland for their subsistence;

B₀ = a constant term.

B_i = a vector of estimated coefficient of the explanatory variables.

$\mathrm{\epsilon }$ = the stochastic disturbance term

3. Results and discussion

3.1. Demographic and socioeconomic profiles of sample households

The socioeconomic characteristics of the respondents are summarized inTable 2. From randomly taken sample respondents 85.1 % were male-headed. 39.5 % of the respondents in the study area were between the ages of 30–45 followed by the age of 45–60 (28.9 %). Out of the total sampled respondents (n = 114), 91.2 % were married, whereas the remaining 8.7 % were single, divorced, and widows. Of the total sampled respondents (n = 114), 42.1 % were writers and readers followed by illiterates (36.8 %), and the rest were attending primary and secondary school. Most of the sampled respondents (75.4 %) were under medium wealth status, and the average family size of sampled households was 5.35.

Table 2.

Demographic and socioeconomic characteristics.

Variables items		Frequency	%
Gender	Male	97	85.1
	Female	17	14.9
Age	15–30	13	11.4
	30–45	45	39.5
	45–60	33	28.9
	>60	23	20.2
Marital status	Married	104	91.2
	Single	3	2.6
	Divorced	4	3.5
	Widowed	3	2.6
Education level	Only write and read	48	42.1
	1–8	16	14.0
	9–12	8	7.0
	Illiterate	42	36.8
Wealth status	Rich	21	18.4
	Medium	86	75.4
	Poor	7	6.1
	Total	114	100.0

Source: own field survey, 2023.

Land for crop production and livestock production were important fixed assets of the households in the study area. All respondents were native to the area and on average the household owned 1.92 ha of land. The source of land for the majority of sampled households (36 %) was their landholding, 25.4 % were their landholding and rent, 25.4 % were own holding and inheritance, 8.8 % were own holding, inheritance and rent, 9 % were only rent, 6.1 were only inheritance, and 0.9 % were Government concede. Out of the total, 93 % of sampled households used their land holdings for crop production and 7 % were used for agroforestry purposes.

Households in the study area rely on a wide range of economic activities related to crop production. A major source of income identified from the household survey were crop production, livestock production, forest products, and off-farm activities (pity trade and honey production) and they share 49.4 %, 35.1 %, 14.36 %, and 1.12 % of their livelihood, respectively. The average distance from the forest to the respondent's residence is 2.5 km (they were between 1.5 and 3.5 km). The respondent had an average no of 10.35 tropical livestock units (ranging from 3.18 to 23.96).

3.2. Socio-economic contribution of B. papyrifera-dominated woodland to local households’ livelihood

All sampled households used forest and NTF products for their annual consumption. This study was in line with [16]. The total households' incomes in the study area were the summations of all incomes they gained from crop production, livestock production, and forest-related and off-farm activities. Of the total mean annual income; forest-based incomes contribute 14.36 %. Other researcher results in Ethiopia: Liban by Ref. [17], Hammer district by Ref. [18], Yayo district by Ref. [19], and Liban by Ref. [12] were 32 %, 21.4 %, 44.7 %, and 22 % respectively. These researchers placed the contribution of forest and NTF products as second whereas in our study forest and NTF products were contributed next to crop and livestock productions (placed as third). The lower income shares in this study compared to the above-mentioned studies were due to three main reasons. These are non-engagement of local communities in gum and resin extraction, due to a lack of awareness locals could not use woodlands for honey production, and they couldn't sell forest and NTF products or they used forest and NTF products only for annual subsistence.

The maximum total annual income of the sampled respondent was 48493.9 USD with a mean of 12048.1 USD. The total annual income derived from forest and NTF products by sampled households ranged from 732.7 USD to 2825.4 USD with a mean of 1759.45 USD. This study was comparable with the work of [19], who reported that the total annual income derived from NTFPs was between 0 and 1083.41 USD and a mean of 277.36 USD. However in our study, according to FGDs, households, and DAs survey results, B. papyrifera had no contribution to the annual income of the locals. The main reason was frankincense production was done by legal investors and due to a lack of awareness locals did not search socio-cultural values of the species. According to the sampled result, these investors had no contribution to the local communities by any means because they had their own serviceman/train workers, food supplies, and transport cars. This study was in line with [6], who stated that in the Metema district, 100 % of the respondents are not engaged in frankincense production, and thus gain no income from it. On the other hand, this study contradicts the findings of [5,8], they reported that local communities benefited from B. papyrifera through the sale of frankincense and from employment opportunities.

3.2.1. Contribution of major forest and NTF products to forest-related income

The sampled households identified grasses and fodder, wild edible fruits, fencing material, medicine, shade for their livestock, farming utilities, handy crafts, charcoal, firewood, and construction wood as the main forest-based activities that contribute to their annual income.

Fodder/grass: in the study area livestock production was an integral part of their livelihood system and on average the sampled households had 9.92 tropical livestock units (Table 3). During field observation time local communities used B. papyrifera dominated woodland for grazing land. In addition, according to FGD and household survey results not only used as grazing land but also used some species such as Ziziphus mucronata, Strereospermum kunthianum, Erythroxylum fischeri, Ficus sycomorus, Lannea schimperi, Rhamnus staddo, Erythrina brucei, and Anogeissus leiocarpus from it for their animal's fodder by cut and carry system in dry and fire hazard season. The survey result shows that 100 % of sampled respondents used forest areas for grazing purposes, and the mean annual income they gained from grasses for their animal feed was 1087.53 USD/household (Table 4).

Table 3.

Types of animals and their corresponding average number/household.

Types of animals	Average no/household
Oxen and cows	8.09
Sheep and goats	5.9
Donkey	1.17
Poultries	14.67

Table 4.

Local communities’ mean annual income (USD) gained from forest and NTF products.

Forest and NTF products (Annual income gained from)	Minimum	Maximum	Mean	Std. Deviation
Firewood	0	806	200.2	140.34
Charcoal	0	1342.88	138.03	236.04
Fence	0	1342.88	178.9	203.83
construction of house	0	931.07	78.97	194.4
farming utilities	0	456.58	47.2	52.36
handy crafts	0	34.38	6.92	5.64
fodder/grazing	116.38	2685.76	1082.27	526.8
medicinal purpose	0	107.43	19.73	24.97
wild edible fruits	0	89.53	8.49	14.05
gum and resin	0	0	0.00	0.000

Source: own household survey, 2023.

Fencing purpose: in the study area almost all respondent households are farmers and to control the effects of disturbance they construct fences on their crop lands. The common species used for this purpose was Acacia species. According to the household survey, 75.4 % of households used B. papyrifera dominated woodland for fencing, and also the mean annual income gained for this purpose was 178.9 USD per household (Table 4).

Shade to livestock: according to FGDs and household survey results households used B. papyrifera dominated woodland in the dry season for shade to their livestock. The major species used for this purpose were Strereospermum kunthianum, Erythroxylum fischeri, Ficus sycomorus, Lannea schimperi, and Apodytes dimidiata.

Farming utilities and Handy craft: Household farming utilities are the basic elements of agricultural production. Dryland species are multipurpose species. They have ecological, socio-cultural, and economic values. Farming utilities and Handy craft are some of its sociocultural values. This study revealed that local communities cut/used their annual farming utilities and handy craft from B. papyrifera dominated woodland. The mean annual income gained from B. papyrifera woodland for farming utilities and handy crafts was 47.21 USD and 6.92 USD/household respectively (Table 4). The result shows that 85.1 % and 88.6 % of households used their farming utilities and handy crafts from B. papyrifera dominated woodland. The common species used for this purpose were Strereospermum kunthianum and Acacia species.

Firewood and Charcoal production: Firewood and charcoal are the preferred fuel for domestic use in Ethiopia. Firewood and charcoal were the main energy sources in the study area. Crop residues are also another fuel source but it was used in a small amount. According to FGDs and household surveys, some species such as Acacia and Combretum were preferable for charcoal production whereas except for B. papyrifera, all species are used for firewood purposes. This study consistence with the result of [20]. In our study from the total sampled households: 93.9 % and 43.9 % used forests for firewood and charcoal production respectively. The mean annual income gained from the forest for firewood and charcoal wood per household in this study was 200.2 USD and 138.03 USD respectively (Table 4). In this study, 93.9 % of respondents were firewood users from B. papyrifera dominated woodland. This study was consistent with the work of [2].

Medicine: the use of woodlands for medicinal purposes is conventional elsewhere. The World Health Organization estimated that 80 % of the populations of developing countries rely on traditional medicine, mostly herbal drugs, for their primary healthcare needs [21]. In Ethiopia, the importance of traditional medicine in general and medicinal plants in particular is well recognized and it is officially acknowledged by government authorities that as much as 75–80 % of the population depend on traditional medicine. According to FGDs and 77.2 % sampled households results, locals used B. papyrifera dominated woodland to treat livestock and human disease; the most common species are Erythrina brucei, Capparis tomentose, Erythroxylum fischeri, Combretum collinum, Combretum molle, and Ziziphus mucronata, and the mean annual income gained for this purpose was 19.73 USD/household (Table 4).

Wild edible fruits: Wild edible fruits are customized and adapted especially in rural areas of developing countries. The study indicates that 63.2 % of sampled respondents used wild edible fruits from this woodland. The mean annual income they obtained was 8.5 USD/household (Table 4). The most common species they used were Grewia ferruginea, Grewia villosa, Ziziphus mucronata, and Ximenia americana.

3.3. Determinants’ of the engagement of local communities in the collection of forest and non-timber forest products from the B. papyrifera-dominated woodland

The study area's woodland provided respondent households with a foundation for their way of life by offering a variety of goods and services. It was found that local communities used around nine different kinds of forest and non-timber forest products: forest grasses and feed, wild edible fruits, fence material, medicine, livestock shade, farming utilities, handicrafts, charcoal, firewood, and construction wood. Nearly all of the sampled households used and harvested at least one type of forest and non-timber forest product, according to the sampled households' results; however, there were differences in the amount of these resources collected, and various factors had an impact on the income obtained from these resources (Table 4). Multiple linear regression analyses were conducted to determine the variations in the degree of engagement variation.

The result of multiple linear regression showed that the model was significant at a 5 % probability level (Table 5). R² of 0.959 indicated the explanatory power of the model means 95.9 % of the variation was due to the explanatory variables. All explanatory variables that of Age of respondent, sex of the respondent, marital status, distance from the household's residence to the forest, numbers of tropical livestock, household's perception towards woodland cover change, wealth status, family size, total land holding size, and education level showed positive relations to annual income derived from the woodland (Table 5) with only some variables being significant. Of all variables, the regression analysis revealed that four were significant at 1 % and 10 % probability levels. These variables were age of respondent, family size, education level, and total land holding size.

Table 5.

Determinants of local communities' engagement on the benefits from the woodland: multiple linear regression analysis result.

Explanatory variables	Coefficient	t-value	P-value
Constant	18947.66	2.3	0.023*
Respondent sex	895.68	0.55	0.58ns
Distance to forest	1.4	0.001	0.999ns
Age	12925.41	8.57	0.000***
Family size	3700.18	7.59	0.000***
Marital Status	94.11	0.101	0.92ns
Total land holding size	1463.34	1.8	0.075*
Education level	5107.07	6.97	0.000***
Wealth status	373.03	0.32	0.75ns
No of the livestock's	120.32	1.68	0.96ns
Perception towards forest cover change	1462.54	1.19	0.24ns

N = 114, R² = 0.962, adjusted R² = 0.959, ns = non-significant, *, ** and *** statistically significant at p < 0.1, p < 0.05 and p < 0.01 respectively.

Age of respondent: the regression analysis revealed that age of the respondents positively and statistically (p < 0.01) affected the probability that households earned the amount of money annually from the woodland (Table 5). This is because as people become older, households make less of an effort to gather forest and non-timber forest products from woodland areas, while young people make an attempt to do so, even when it involves traveling great distances and climbing hills. Of the respondents in this study, 39.5 %, 28.9 %, 20.2 %, and 11.4 % were in the 30–45, 45–60, >60, and 15–30 age groups. Given that the majority of respondents were under the age of 50, and that younger individuals tend to have greater energy than older people, it follows that younger households harvest forests and non-timber forest products more frequently than older ones. In addition, the younger respondents employed more fences and other necessary items for their tasks, as well as seeded more land. The age of the households increase by one unit in this study the amount of variation between the household increased by 231.43 USD. This study was in line with [21].

Family size: the result revealed, family size positively and statistically (p < 0.01) affected the probability of households collecting the amount of forest and NTF product from the woodland (Table 5). This could be related to the need for additional income (in addition to main livelihood activities) in order to support a larger number of families, and on the other hand, the quantity of production may increase as more household members participate in product collection, resulting in increased income. This is because larger families have more labor available to gather resources from the forests and eat more forest and NTF products. This study was in line with earlier findings [18,22,23], and [24], who found that close to forest lands, large-household families extracted more resources from the forest. This study contradicts with the results of [19], who reported that respondents’ family size non-significantly related to NTF products collected from the forest.

Total landholding size: regression analysis revealed that area of land holding size positively and statistically (p < 0.1) affected the probability that households harvest forest and NTF products, and generate additional income from the woodland (Table 5). This investigation aligned with the [19], and [25], who discovered a positive relationship between total land holdings and income from non-timber forest products. The positive relationship between NTFP and total land holding means that households with a greater amount of total land are more likely to harvest forest products. This could be because total land holding is the sum of crop land, which produces a lot of crops, as well as other types of land. Several forest and NTF items, including fuelwood, charcoal, fencing material, handicrafts, agricultural tools, grass/animal feed, wild edible fruit, medication, and building materials for homes, were found in this discovery. Compared to small landholders, respondent households with larger landholding sizes in this research were utilized more NTFPs and forests for their land fences. Besides, compared to other landowners, they utilized a higher quantity of farming utility. The annual amount will increase by 26.2 USD as the landholding size increases. Our findings contradict [26], who discovered a negative relationship between total landholding and household income from NTFPs.

Education level: regression analysis revealed that education level of respondents positively and significantly (p < 0.01) affected the probability that households harvest forest and NTF products from the woodlands (Table 5). This means that all respondents used forest and NTF products for their livelihood, but the amount they used varied. In this study, the majority of families were only writers and readers (42.5 %) and illiterates (36.8 %) (Table 5). Thus, illiterate, writer and reader households consumed more forests and NTF products (in birr 91.44 USD per year) than educated people. Educated people have a more positive attitude toward forest protection and exploitation than uneducated people. In this study, less educated and non-educated person utilized woodlands for grazing, charcoal production, fencing, fuelwood, and other driving forces in greater quantities. This study contradicts with the results of [19], who reported that respondents education level non-significantly related to NTF products collected from the forest.

4. Conclusion and recommendations

The livelihood activities of households in the study area were crop and livestock production, forest-related activities, and non-farm/off-farm activities. Within these diversified income sources, households exploited forest and NTF products from the woodland. However, B. papyrifera had no contribution to their annual consumption/livelihoods the reason was the species is under the control of investors, and in addition to that its socio-cultural value couldn't searched by the local communities. The woodland contributes to the locals' livelihood next to crop and animal production but the local communities collect more forest products than NTFPs. Local communities use species from the study area but their engagement to get annual consumption varies from household to household. The multiple linear regression model showed that the model was significant at a 5 % probability level and 95.9 % of the variation was due to the explanatory variables such as age of respondent, family size, total land holding size, and education level showed positive and statistically significant relations to annual income derived from the woodland. The explanatory variables increase by a number the dependent variable also increases and if this explanatory variable decreases by a number then the dependent variable also decreases to some extent. Therefore these factors should be considered during the design and implementation stage of activities related to forest and NTF products collection time.

Based on the result of the study, the following recommendations were drawn.

• ➢The findings revealed that all households used more forest products than NTFP so awareness of to use of NTFPs and sustainable utilization interventions related to deforestation and forest degradation reduction should be created within each village.
• ➢Currently, gum and resin production is under the control of legal investors so local communities did not gain any benefit from B. papyrifera, due to this reason they started expanding crop and livestock production within the woodland, in the future the population status of the woodland will be risky so to minimize this effect the District Agricultural Office should implement policy related to local communities benefit from investors.
• ➢According to field observation and household survey currently, different types of investors (stone production, Granait, gum and resin, and sand production) invaded the area this will affect the species' sustainability so the District Agricultural Office should give priority attention to this investor's effect on the woodlands.

CRediT authorship contribution statement

Adane Mulat Nigus: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Resources, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Abeje Eshete Wassie: Visualization, Validation, Supervision. Asmamaw Alemu Abtew: Visualization, Validation, Supervision.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Adane Mulat Nigus reports financial support was provided by Oda Bultum University and University of Gondar. Adane Mulat Nigus reports a relationship with Oda Bultum University and University of Gondar that includes: speaking and lecture fees. Adane Mulat Nigus has patent licensed to For all authors. Co-authors: Abeje Eshete Wassie, and Asmamaw Alemu Abtew If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article: