Selection of Soil and Water Conservation Technologies and Native Tree Species for Rehabilitation of Degraded Arid Lands in Southeast Ethiopia

,


Introduction
Te Ethiopian government claims about 15.5% forest cover that comprises diferent types of forest ecosystems [1]. However, Ethiopia is among those countries with a relatively high rate of deforestation-around 1.0% per annum [2]. Te recent results of national forest inventory in fact indicated that the average annual rate of deforestation from 2000 to 2013 is about 92,000 ha-around 0.53% [1]. Te variation in the annual rate of deforestation indicates the variation in the dynamics and extent of forest cover change with time and space. Te diferent types of forest resources located in diferent parts of the country are subjected to diferent biophysical and socioeconomic descriptors that can help to determine the drivers and dynamics of forest cover changes [3,4].
Te lowland dry land areas of the country have low population density relative to the highland areas [5] due to harsh environments for living and poor infrastructures (road, water, electricity, and telephone). A newly formulated economic policy focused on facilitating the commercialization and export orientation of the agricultural sector. To realize the policy, a number of subsidiary policies, strategies, and programs were introduced to promote the expansion, commercialization, and export orientation of the agricultural sector. Te major strategy to achieve this goal is to attract direct foreign investment in large-scale commercial agriculture through allocation of mainly extensive forest lands and provisioning of tax incentives to private investors [6,7]. Accordingly, the government of Ethiopia has been leasing large forested areas in lowland areas to establish commercial agriculture, with the support of a favorable incentive structure [8]. Following this, Government of Ethiopia has leased about three million hectares of land to investors [9,10]. Tis large-scale investment is identifed as the main direct driver of deforestation and land degradation in lowland areas of Ethiopia [11]. Tis is simply confrmed by the higher deforestation rate of Acacia-Commiphora and Combretum-Terminalia woodlands biome than the other three biomes in recent times [1], suggesting the economic development policies and strategies are also bearing signifcant costs to the country's land and forest resources [8].
In addition to the large-scale investment, expansion of sedentary agricultural settlements and overgrazing due to a mounting population has become a major cause of land degradation [12]. Te increase in land degradation brought a growing concern for policy makers who fear that land degradation might reduce grazing capacities and impoverish already fragile living conditions of the pastoralists [13]. Such an increase in land degradations necessities urgent calls for interventions.
Very recently, Ethiopia has developed a new strategy called the Climate Resilient Green Economy (CRGE) strategy that take Ethiopia to middle income status by 2025 through low carbon, resilient, green growth actions [14]. Te strategy emphasized agriculture and forestry. Te CRGE strategy targets 7 million hectares for forest expansion through protecting natural forests, restoring degraded lands, and promoting sustainable use of terrestrial ecosystems by managing forests, combating desertifcation, and halting and reversing land degradation and halt biodiversity loss [1]. Forest restoration is now regarded as a cornerstone of national biodiversity conservation and sustainable development [15]. Te CRGE clearly targeted to rehabilitate one million ha of degraded lands. In addition to the national strategy, Ethiopia has also pledged to restore 22 million ha of degraded lands following the Bonn challenge, New York declaration, and the Africa Forest Landscape Restoration (AFR 100) initiative [15]. Ethiopia is striving to achieve restoration goal through implementing a number of programs that includes reducing emissions from deforestation and forest degradation (REDD+) and a 10-year national forest sector development program [15].
Following this, the country has been undertaking different types of restoration activities that include exclosure, aforestation, and reforestation [15]. National tree planting campaign is also being practiced every year. While these political commitments have acquired wide support, there are huge knowledge gaps to achieve the country's ambitious restoration targets so far agreed, and a set of efcient restoration approaches will be necessary to achieve these targets. Among others, selection of suitable native tree species from the national species pool and soil and water conservation structures is the most critical step for rehabilitation of degraded forests and reforestation programs to be successful.
In arid and semiarid areas, water availability is one of the limiting factors in tree growth [16], and thus, the adoption of erosion-control measures is a highly recommended technique for successful restoration in degraded dry land areas [17]. Te soil and water conservation structures in dry land areas enhance soil water infltration, soil organic carbon, and promote its availability for agricultural and forest species during the dry period [18][19][20]. In addition to the soil and water conservation structures, the choice of suitable tree species can also determine the efectiveness of the rehabilitation program. Tree species vary in their performance and adaptation in areas with low fertility soils, high temperature variation, and extreme drought conditions. Tus, selection of native tree species with high early growth potential that can quickly restore degraded lands is very important for long-term reforestation success.
Te motivation of this study stems from the lack of research on species selection and soil and water conservation structure for restoration of degraded forest lands. However, the ecological information necessary to select species and conservation structures which are appropriate and adapted to local conditions is inadequate in developing countries. Tis lack of data has limited the success of forest restoration programs [21,22]. To address this knowledge gap, research on the restoration potential of soil and water conservation structures and diverse native tree species is urgently needed. Tus, the main aim of this study was to select the best performing tree species and conservation structures on the degraded lands of arid and semiarid conditions in Ethiopia.

Description of Study Site.
Tis study was conducted in Liben District of Guji zone, Oromia Reginal State, Ethiopia ( Figure 1). Te district is located at about 630 km south of Addis Ababa and has a total area of 741,752 ha. Te dominant landscape of the district slopes gently from the north-west to the south-east and the altitude range from 1000 to 1600 m.a.s.l [23]. Te arid and semiarid agroclimatic zones are the dominant climatic zones in the district. Te mean monthly minimum and maximum temperature is 16°C and 28°C, respectively, with annual rainfall of 460 to 790 mm with an annual average of 609 mm. Te rainfall distribution is bimodal where the main rainy season is between February and June and accounts for 55% of the annual rainfall. Te short rainy season is between September and December that accounts for 41% [23]. Drought is common in the study area every fve to ten years. Both pastoralism and agropastoralism are the dominant livelihood in the district where most of the communities are predominantly dependent on livestock production [24]. Te area is characterized by moisture defcit, resulting on water shortages that reduced the livestock and crop production potential of the area and thus the livestock production and crop cultivation are subsistence type (Guji Zone Administration Report, 2016).
Acacia-Commiphora woodland is the dominant vegetation type where large areas are characterized by extensive grazing lands covered mainly by grasses with scattered acacia and some noxious species. Te vegetation is an important source of income through production and marketing of forest products such as gum and resin, charcoal, and fuel wood. Forest products from the vegetation have also a signifcant role as a coping strategy at times of drought. However, a number of research reports indicated that the vegetation which is used as a rangeland in the district is shrinking signifcantly and severely degraded for the last 30 years. Recurrent drought, expansion of crop cultivation, overgrazing, charcoal and fuel wood production, and bush encroachment are the main drivers of degradation while human population growth, settlement, and bans on use of fre are reported to be the underlying drivers of degradation [25,26].

Te Study Tree Species.
Four native tree species that have been growing in the study area were initially selected from the list of native plant species based on the preference of the local people through group discussion, forest ecological knowledge, traditional uses, and availability of seeds. Te four species were Cordia africana, Acacia tortilis, Acacia goetzei, and Combretum molle. Te local use of native species was surveyed through a traditional ethnobotanical approach. Te habitats of the selected species were also documented.

Experimental Design.
A factorial experiment with randomized complete block design and three replications were employed for the current study. Te frst factor was the soil and water conservation structure with fve levels (control (pitting), half-moon, spot hoeing, mowing, and trench).
Pitting which consisted of digging a planting hole at a depth and width of 30 × 30 cm. Tis is a common practice in the area and hence considered as a control. Spot hoeing is done in 1 m diameter circle and all root remnants of all plant species were removed from the soil. Tis practice is expected to loosen the compacted soil and facilitate the seedlings root penetration. Half-moon structure was established to facilitate moisture retention. Trench structure was also established by digging a trench (40 cm deep) with a contour ridge (40 cm high) used for tree planting. Mowing is the last treatment level which involves cutting of the grasses (weeds) around the seedlings two times per year during the rainy season. Te mowing is expected to decrease the above ground competition of the grass with the seedlings. Te second factor was tree species with four levels (four tree species) as indicated above.
Tree experimental blocks (considered as replications) of 4104 m 2 each (72 × 57 m) with 5 m spacing between blocks were set up on a gentle slope in 2017 and followed up until 2019. In each block, 20 experimental plots with 3 m spacing between plots were established that measure 12 m × 12 m. Each plot has a treatment combination of tree species and soil and water conservation structure. Nursery-raised seedlings were out planted in randomly assigned plots and 49 (7 * 7) seedlings that measure ± 30 cm height per species were planted with a spacing of 2 m in each plot (Table 1). Altogether 2,940 seedlings of four species were planted in June/July 2017, which is at the beginning of the wet season. After planting, weeding was carried out three times (May, July, and September) a year. International Journal of Forestry Research

Data Collection.
Te feld performance of the four planted species was evaluated based on data from a restoration planting trial. Twenty-fve inner trees in every plot were considered for survival and growth data collection, and the rest were not considered to avoid the border efects [27]. Seedling survival, root collar diameter (RCD), and seedling height were recorded every six months after planting over the two-year period. Te survival rate of planted seedlings on each plot was assessed by counting both live and dead individuals at the end of the annual growing season. RCD was measured with calipers while height was measured with a measuring tape at the stem base.

Data Analysis.
Te survival rate of the planted trees within the study plot over the two-year period was calculated using the following formula [28]: where X � survival rate of planted seedlings, Y � total number of planted seedlings, Z � total number of survived trees after two years. Te survival percentage data were then arcsine transformed while RCD and height data were square root transformed to meet the normality assumptions of analysis of variance (ANOVA). Two-way analysis of variance (ANOVA) was employed to test the signifcance of species, structures, and species vs. structure interaction efect on mean performance of planted species measured in terms of mean annual growth rates of RCD and height using a linear model. All tests were performed with R statistical software, version 2·15·0 (R Development Core Team, 2012).

Survival Rate of Tree Species.
Te overall survival rates for almost all species and structures tested were very low (>31%). However, the survival rates varied signifcantly among species and structures at P < 0.05 (Tables 2 and 3). As can be seen from the survival rate of each species, the highest survival rate was recorded for Acacia tortilis (58.22 ± 12.38%) and the least survival rate was recorded for Cordia africana (4.00 ± 1.91%). Acacia goetzei also presented a relatively higher survival rate (42.99 ± 8.81%) when compared to the Combretum molle (24.22 ± 7.34%) and Cordia africana. Acacia tortilis showed a signifcantly higher (P < 0.05) overall survival rate than the other two species, Combretum molle and Cordia africana, across all moistureconserving structures, but with Acacia goetzei. Tere were no signifcant (P > 0.05) diferences in the survival rate between Acacia goetzei and Combretum molle. Cordia africana has shown a signifcantly (P < 0.05) lower mean survival rate than the other three species (Table 2).
Almost all species planted in half-moon and trench moisture-conserving structures showed a very high survival rate than the other structures (Table 2). Te survival rates of all species grown in trench (64.00 ± 11.49%) and half-moon (52.44 ± 12.48%) structures were signifcantly higher (at P < 0.05) than the control (18.69 ± 7.46%), spot hoeing (12.40 ± 5.54%), and mowing (17.50 ± 8.48%), suggesting that water conservation structures are very helpful to enhance survival rate of planted seedlings in arid and semiarid agroecological zones. However, there was no signifcant diference (P > 0.05) between the mean survival rate of all species planted in the pitting (control), spot hoeing, and mowing. Te analysis of variance revealed that the interaction of species survival with moisture conservation structures was found to be insignifcant (P > 0.05).   (Tables 4 and 5). Te height and RCD growth of species varied signifcantly among species and structures at P < 0.05 (Tables 4 and 5).

Growth
Cordia africana showed a signifcant height growth than the three species; Acacia tortilis showed a signifcant height growth than Acacia goetzei and Combretum molle, while Acacia goetzei showed a signifcant height growth than Combretum molle. Te efect of moisture conservation structures on seedling growth in height and RCD was highly signifcant at P < 0.05 (Tables 4 and 5). Te average growth in height of    Te average height growth of planted species in mowing was signifcantly lower than species planted in spot hoeing and control.
With regard to seedling growth in RCD, seedling planted in trench (1.92 ± 0.27 cm) and half-moon (1.66 ± 0.31 cm) had higher growth performances than the other moisture conservation structures. Tere were a signifcantly higher average growth of seedlings planted in trench, half-moon, and spot hoeing (1.08 ± 0.17 cm) than the control (0.96 ± 0.20 cm) and mowing (0.81 ± 0.23 cm). Similarly, there were a signifcantly higher growth of seedlings planted in spot hoeing, half-moon, and control than mowing. However, no signifcant diference was observed in RCD growth between seedlings planted in trench, spot hoeing, and half-moon. Similarly, the growth in RCD of seedlings planted in spot hoeing, half-moon, and control did not difer signifcantly.

Discussion
Our results show signifcant diferences in survival rates among the tree species. Based on the survival rates, Acacia goetzei and Acacia tortilis showed a signifcantly higher overall survival rate than the other two species across all moisture-conserving structures. Tese results agreed with the results of Kasaye et al. [29] who have demonstrated better performance of Acacia species in degraded dry lowland areas, such as Abergelle District of Amhara Regional State, Ethiopia. Tis might be due to the nature of the dry land Acacia species which are classifed as drought avoiders [30] and their ability to fx nitrogen that helped them to perform better in degraded soil [31]. Te overall lower survival rates of Cordia africana and Combretum molle among the studied tree species in the same study suggest that their level of adaptation in the degraded arid and semiarid condition is very low [32]. Te lowest survival rate of Cordia africana seedlings in the non-water harvesting structures might not be due to an insufciency of rainfall but due to reduced soil moisture conditions or an insufciency of moisture available within the rooting zone [33,34]. Reduced soil moisture conditions or an insufciency of moisture available within the rooting zone are commonly viewed as a signifcant barrier to artifcial reforestation in moisturestressed areas [34]. A similar result was reported by Donoso et al. [35], who indicate that high water restriction afects the physiology of some tree species, provokes a reduction of stomatal conductance, net photosynthesis, and transpiration, and produces lower leaf and root biomass.
Te survival rates of species grown in trench and halfmoon structures were signifcantly higher than the pitting (control), spot hoeing, and mowing suggesting that water conservation structures are very helpful to enhance the survival rate of planted seedlings in arid and semiarid agroecological zones. Te higher survival rate of tree species planted in water conservation structures might be related to the higher efciency of the structures in capturing rainwater during the rainy season, enhancing water infltration, conservation, and allowing greater water availability during the dry season to ensure dry season tree seedling survival [36][37][38]. Successful tree seedling survival depends on the soil condition and stored soil moisture available to ensure tree seedling survival into the next growing season [36]. Te moisture-conserving structures are reported to act as a supplemental moisture and soil nutrients for vegetation growth and survival [39]. Half-moon and trench have been recognized by various authors as a technique that promotes water conservation, especially in arid and semiarid zones [37,40,41]. Tis is because the moisture harvesting structures improve moisture in the root zone of tree species [42].
Te height and RCD growth of species varied signifcantly among species and structures at P < 0.05. Based on the fndings of this study, individuals of Cordia africana have a very good local performance while Acacia species have a suboptimal growth performance. Te species diference in growth may be due to the diference in the growth strategy and genetic superiority of the tree species for the growth [43][44][45]. Cordia africana is a reasonably fast growing tree species compared to Combretum molle and Acacia tortilis which are categorized as slow growing tree species [46]. According to Reubens et al. [32], Cordia africana is one of the top three priority species selected as the indigenous, naturally growing species adapted to harsh conditions, often with a sociocultural importance, high protection, and reclamation value, and an excellent agroforestry potential. Since Cordia africana is selected by the local people for a number of its socioeconomic value and has performed well in growth and survival when it is combined with moisture conservation structure, the use of the species for rehabilitation of degraded lands in arid and semiarid conditions will contribute to successful rehabilitation schemes [47].
Te current study showed that the height and RCD growth of the four studied species was signifcantly higher in trench and half-moon moisture conservation structures than the other structures indicating that moisture-conserving structures contributed for higher growth increment and are the main factors controlling the performance of tree growth. Various studies in diferent dry land areas of Ethiopia have showed a similar result where the height and diameter growth of all planted tree species were best under moisture conservation structures than that of tree seedlings planted in normal pits [48][49][50][51][52]. A similar study in the interior dry land of the Maule Region, Chile, by Ovalle et al. [53] reported a signifcant efect of moisture-conserving structures on plant height, crown diameter, and trunk diameter. Moisture-conserving structures are also reported to facilitate the regeneration of woody plants in the degraded area [54].
Te commonly observed drought stress in arid and semiarid areas decreases the growth and development of planted tree species by reducing the photosynthetic rate [55]. 6 International Journal of Forestry Research Such stress can be overcome by constructing moistureconserving structures that enhances the growth increment in semiarid climates [56,57]. Te main role of soil moisture conservation structures are reducing runof and improve the infltration of rainwater during the rainy season. Te infltrated rainwater will be available for plants during the dry period, which vigorously enhances plant growth and increases plant biomass compared with that of trees planted without moisture harvesting structures, that is, normal pits [58,59].

Conclusion and Recommendation
Survival and growth rates are important factors in selecting the best tree species and moisture conservation structures for restoration activities in degraded arid and semiarid areas. Te performance of tree species varied in height, root collar diameter, and survival rate. Better survival rate was observed for the two Acacia species, A. goetzei and A. tortilis. A signifcant height and RCD performance was observed for Cordia african. An optimal growth performance was observed for both Acacia species. Combretum molle showed the least in growth performance while modest in survival rate.
Half-moon and trench moisture-conserving structures showed a great potential in providing suitable conditions for successful tree establishment and tree growth by promoting the highest water content in the soil profle. Almost all of the four tested tree species exhibited a signifcantly higher survival rate and growth performance when established with the two moisture conservation structures. Terefore, to restore degraded lands and to support the ongoing land rehabilitation programmes in arid and semiarid areas, we recommend the two Acacia species and Cordia africana in combination with the moisture conservation structures than the conventional method of tree planting. Authors would also like to recommend additional studies regarding the efect of moisture-conserving structures on water and soil nutrient dynamics. It would be wise to conduct a further study using a range of species with socioeconomic and ecological importance to get more alternative suitable species for successful rehabilitation of degraded lands.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.