The time required to collect fuel wood from the natural forest and shrub lands may increase with deforestation, since the forest cover will be pushed up to the marginal areas. This showed that the tasks of fuel wood collection from the surrounding common areas are time consuming and it is proportionally correlated with distance of the site in which the fire wood is found. In line with the present result, Palmer (2009) reported that fuel wood scarcity has a positive effect on labor inputs to fuel wood collection from common areas.

The annual mean income of the parkland agroforestry introduced household in the year 2016/2017 was about 25915 birr (equivalent to 863 USD) and this was greater than the mean annual income of 21684.4 Birr (equivalent to 722 USD) earned in the same year by the households who was not introduced parkland  agroforestry   (Table 1) though there were some uncontrolled factors that can influence the income of each households.

Parkland agroforestry and fuel wood collection

From the parkland  agro  forestry  introduced  households (n=69), the responsibility of harvesting and transporting fuel wood for the whole family, only male and only female were 62.3, 29 and 8.7%  respectively (Figure 2). Majority of the household heads are of the opinion that the big trees require participation from all family members, initially to prune some selected branches of a tree which is commonly and traditionally performed by the male family members and transporting task also left mostly for all family members after the foliages and smaller part of branches are consumed by livestock. This indicates that the trees grown in the farm lands were important not only to provide fire wood and other products but also to minimize the work load of women and children by creating opportunities for labor division among all family members to harvest fuel wood and this in turn could have impact on the families’ socio-economic developments. FAO (2013) pointed out similar result that combines agricultural crop and fuel wood production through agroforestry to save woodland trees and frees up labor, especially for women, who traditionally collect fire wood. On the other hand, the survey result derived from the non-parkland agroforestry users showed that fuel wood collection responsibility in these households was inclined to same particular family members rather than distributing it to all of the family members. 71%  of  the  non-parkland agroforestry introduced households (n=69) affirmed that the fire wood collection responsibility in their family mostly rest on the shoulder of children, young female and the mothers. About 15.94% of the respondents also replied  that all family members had equal responsibility on fuel wood gathering activities and only 13% opined that the father and the young male were the most responsible to collect fuel wood from the surrounding forest and non-forest areas (Figure 2).

This shows that the children and women found in the households who have no parkland trees on their farmlands took the responsibility of fuel wood collection from the local forest and shrub lands. Kassie (2015) reported similar result that fuel wood collection responsibility from the nearby forest and shrub lands is performed by mothers and daughters.

Common parkland tree species on the farmlands of the study area

All the mature parkland trees found in the farm plots of the parkland agroforestry introduced households (n=69) was counted and a total of 807 mature scattered trees were recorded in the Fireweini village. Thus, households had owned different number of trees with a minimum of 2 trees (in 2 farmers) to a maximum of 27 trees (in 1 farmer) and on average, there was about 11.7 trees per household heads and 16.3 trees per ha.

It was also shown that Faidherbia albida was the most dominant parkland tree and it was the only tree species found under all of the parkland agroforestry practice households accounting to about 80.8% (Table 3). The main purpose of keeping this  tree  species  by  all  of  the households and in a dominant number was mainly for its better fodder value, complementary nature of the tree with growing annual food crops, fencing service and fuel wood production. Due to this reason, F. albida had been the most dominant parkland tree species; followed by Ampelocissus abyssinica, 14.6%; C. Africana 2.35%; E. camaldulensis1.5%; and 0.74% was covered by other tree species like Oleaeuropaea (Table 3).

Farmers’ judgment on the role of PLAF to minimize pressure on local forests

From parkland agroforestry user households (n= 69), majority (68.1%) of the household heads responded that the parkland trees found in their village had very high contribution in stabilizing the pressure on the local natural forest and non-forest areas by providing fuel wood/charcoal, fodder and other demands. These were even better than any other available technologies provided in the study area. Such practices are mainly introduced to minimize deforestation. Similarly, 20.3% of the respondents also replied  that contribution of the parkland agroforestry in reducing the pressure on the local forests was high, 7.2% also said it was medium while the rest 3% of the respondents said low (Figure 3).

However, no respondent believed that parkland agroforestry had very low/ no contribution on reducing the pressure on the surrounding common forests.  This idea was supported by Duguma (2010) who reported that agroforestry practice could be a promising option to solve environmental problems such as deforestation and to improve household food security by diversifying farm products and reducing vulnerability for seasonal food and fodder shortages.

Challenges of parkland agroforestry

The result indicated that 45.16% of the household heads (n=63), pointed out that the distance from home to farm plots, free grazing, shortage of farmland, prohibition of fire wood/charcoal selling and lack of awareness were considered as major limiting factors to have improved parkland agroforestry. About 24.2% of the respondents believed that farmland distance and the limited farmland size were identified as the major limiting factors. About 9.7% of the respondents also replied that small size of farmland were the only constraint to improve the parkland agroforestry practice, 8% said that free grazing and lack of awareness was the primary limiting factors for them and 4.8% believed that labor demanding nature of the parkland trees, lack of government support and shortage of farmland were main constraints; whereas 3.2% said that farmland distance, free grazing, general prohibition of fire wood/charcoal selling and weak local forest protection performance were major constraints (Figure 3).

Some respondents (1.6%)  also   believe   that  lack of awareness, general prohibition of fire wood/charcoal selling and uncontrolled grazing were main limiting factors. Some others (1.6%) also believe that lack of better tree species identification mechanism, absence of government support and small farmland size was primary constraints. Also some others (1.6%) are of the opinion that the negative effect of trees on growing field crops and being obstacle for farming activities were the main challenging factors to improve the parkland agroforestry (Figure 3).

The result from FGD (Focal Group Discussion) and from the key informants’ interview also indicated that, free grazing, lack of government focus on the farmland trees improvement, farmland distance and lack of farmers’ awareness were primary constraints to the PLAF improvement in the study area. The uncontrolled grazing of animals on farmlands after the field crops are harvested (dry season) was one of the most limiting factor in the study area, including destruction of the protected and new planted areas. In line with this idea, Mekonnen and Kohlin  (2008) was reported that free grazing on agricultural landscape was the major constraint for tree planting and maintenance in central Ethiopia. In general, the distance of farm plots from home and fragmented farmland size is among the main constraints to maximize multipurpose trees (MPTs) on farmlands;   to    which   majority    of    the    PLAF    user households were agreed upon. This implies that as distance of farm plots are increasing, farmers’ willingness to plant and protect trees are being decreased. This is mainly due to management problems on the farm, since trees require continuous care and close management efforts. It was due to this reason that more trees are observed on the farm plots found near the residential areas than the distant plots in the study area. Therefore, farmland distance and free grazing are strongly interrelated factors that have been major challenges to plant trees on distant farm plots. In the nearby plots, it was easier to grow and manage trees relatively since household members can prohibit animals from browsing after the field crops were harvested. In agreement with this result, Predo and Francisco (2012) have reported that the relative distance from home was negatively affect farmers interest to grow trees in Philippines.

The result from the focus group discussion also indicated that there is widespread problem of theft of tree products, especially animal fodder and fire wood when planted far away from living home (Figure 4).

Non-parkland agroforestry households

Despite most farmers in the study area was integrated selective trees with their farmlands especially on the plots found near their home, there were also some households who had no trees in their farmlands. Though these households had not integrated trees into their farmlands, majority of them believe that integrating selective trees on farmlands is advantageous. As a result, 91.3% (n=69) opines that integrating selective perennial trees in farmlands are beneficial alternative for them; while the rest 8.7% believed that though the parkland trees can contribute to farmers, the negative effect of the trees on the growing annual food crops and farming activities are out weighted than its benefits, hence they were reluctant to integrate trees on their farmlands.

According to the respondents introducing perennial trees into farm lands can damage annual food crops in different forms, including shading effect and becomes harbor for field crop attacking birds, weeds and becomes obstacle for agronomic activities. In agreement with this result, FAO (2013) identified a general problem of farmers’ perception on trees as incompatible with their farming activities and may not benefit from planting and managing trees as well as shrubs on their farm plots. This can also influence the introduction and implementation of agroforestry practice in wider areas.

There were some  relevant questions provided to the non-parkland introduced households (n= 63) to know the main reasons they remained without introducing trees in their farmlands if they were aware of the advantages of integrating perennial trees on farmlands.

Response from most household heads mainly revolved on a single factor that limits them to grow valuable trees on their farmlands. About 61.9% of the respondents pointed out that, the dry condition of the area was the most limiting factor for them to retain beneficial trees in their farmlands and 15.9% replied  that dry condition, availability of firewood in near areas until the near past years and lack of farmers awareness  were the main constraints on trees growth on farmlands. However, 14.3% believed that the dry condition and availability of firewood in nearby areas were the main limiting factors to grow trees on farmlands. From the respondents, about 3.2% responded that dry condition and absence of better tree species are main the constraints. The rest, 3.2%, of the household heads said that the dry condition, the negative effect of trees on the growing annual food crops, easily accessibility of fire wood until and lack of farmers awareness are the main reasons farmers do not introduce  MPTs on their farmlands.