An evaluation of farmers’ experiences planting native trees in rural Panama: implications for reforestation with native species in agricultural landscapes

In the Republic of Panama, reforestation with native species is of great interest, but many landholders often do not participate in tree planting projects and little information exists about landholder interest in, or experiences with, native trees. This study evaluates the experiences of farmers participating in a native species reforestation initiative in rural Panama to identify lessons learned that can guide on-going or future tree planting efforts. Based on the results of a questionnaire administered to program participants and non-participants (n = 68), we found that trees are important to farmers for multiple reasons, primary a variety of environmental and economic benefits. No relationship between the size of landholdings or land tenure status and the desire to plant trees was found. All participants in the program considered their experience to be positive, few had problems with their plantations, and most were interested in planting more native trees. The program’s frequent and ongoing technical support was an important factor for farmers. These results indicate widespread interest in, and success with, planting native species and underscore the need to systematically examine farmers’ interests and perceptions when planning, implementing, and evaluating reforestation initiatives.     

Tree plantations on farms: Evaluating growth and potential for success

Interest in native species is growing across the tropics as reforestation of degraded lands becomes more widespread. In this study four tree species native to Panama - Cedrela odorata, Pachira quinata, Samanea saman, and Tabebuia rosea - were grown on rural farms at two dry tropical sites in Panama for up to five years. Survivorship and growth data at these “on-farm” trials are compared to data recorded from nearby experimental or “species selection” trial sites and evaluated in terms of soil fertility and management. Participant farmers were also asked about their interest in planting trees in general as well as their interest in 61 species grown in the species selection trial.Although, on-farm survivorship was variable and generally lower than that found on the species selection trial, one species (S. saman) experienced high and consistent survivorship. High survivorship combined with growth data from farms at both sites for this species suggests it would be a good candidate for extension projects working with rural farmers. Survivorship of other species appears more sensitive to farmer management and/or local site conditions. Generally lower growth on the Los Santos farms as compared to the species-selection trial is attributed to the lower soil fertility (plant available P) at the on-farm sites compared to the species selection trial. In contrast, only one species - P. quinata - had a growth variable found to be significantly lower between the on-farm and species selection trial sites in Rio Hato. C. odorata, P. quinata T. rosea can all be used in on-farm conditions with consideration to specific site and management conditions.By 2009, approximately 80% of the farmers planting trees still wished to participate in tree planting activities. All of the farmers no longer wishing to continue with the project expressed slow growth rates of trees as a principal reason. All but one of these farmers had growth rates for his/her trees markedly below those of the species selection trial nearby. Some farmers wishing to continue had very high mortality rates (>70% for all species), suggesting non-tangible benefits for participating in project activities. Other species that were not tested on-farm but grew well in the species-selection trials and were of interest to local farmers are discussed.As long as specific site and management conditions are carefully considered, data from species selection trials can be useful in informing tree planting projects with rural farmers; however, care should be taken to manage expectations.     

Restoration of dry tropical forests in Central America: A review of pattern and process

Much information on restoration and management exists for wet tropical forests of Central America but comparatively little work has been done in the dry forests of this region. Such information is critical for reforestation efforts that are now occurring throughout Central America. This paper describes processes of degradation due to land use and provides a conceptual framework for the restoration of dry tropical forest. Most of this forest type was initially harvested for timber and then cleared for cattle in the last century (1930-1970). Only 1.7% remains largely restricted to infertile soils and remote areas on the Pacific coastal side of Panama, Costa Rica, Nicaragua and Mexico. These cleared areas are again in a state of transition due to a combination of decreasing land productivity, and land speculation for tourism development. Some farms have been sold to new landowners who are interested in reforesting to increase biodiversity and forest cover. Attempts have therefore been made to reforest by protecting the land from fire and cattle, by supplementing natural regrowth with enrichment planting, or through use of tree plantations. Experimental studies have demonstrated the ability of these lands to grow back to forests because of native species ability to sprout after cutting, and the capacity of remnant trees in field and riparian zones to provide seeds and to moderate edge environment for seed germination and seedling establishment. However, research also shows that on sites with long histories of land clearance, species diversity will remain low with functional groups missing unless some active management occurs. Under-planting with late-successional native tree species can add structure and diversity; enrichment planting with large-fruited shade-intolerant species can initiate new islands of more diverse regeneration beneath their canopies; and plantings of fast-growing, nitrogen-fixing trees that provide light canopy shade can moderate the environment below, promoting regeneration establishment of late-successional species. Plantations are the only option for lands that have lost almost all remnants of native forest, and where soils and vegetation have changed to new states of structure and function. Conversion of pastures to tree plantations that can facilitate natural regeneration beneath them is appropriate when pastures are prone to fire and/or lack immediate seed sources nearby. After the grasses have been shaded out, natural recruitment can slowly occur over a 10-15 years period. Under-planting of shade-tolerant late-successional species can supplement species composition and structure.     

Growth,survival, carbon rates for some dry tropical forest trees used in enrichment planting in the Cerro Blanco protected forest on the Ecuadorian coast

Deforestation principally to establish cattle pastures has created large expanses of land dominated by exotic grass prior to the establishment of the Cerro Blanco Protected Forest in Ecuador. A dry forest restoration program was carried out from 2008 to 2017. Tree planting sites were cleared in parallel lines through abandoned pasture and secondary vegetation using manual labor. Native tree species were produced in a nursery on site and used in yearly plantings during the rainy season. A total of 637 hectares were planted at an average stocking density of 1,000 trees per hectare. Growth data was collected from trees planted in nine sites from 2008 to 2017. Mortality rates were determined from a sample of 400 trees of twenty-two species. Survival of all species was greater than 50%. Ten species had survival rates greater than 90%. Vitex gigantea, with high survival rates, also had significantly greater growth rates than the other species (p < 0.0001) and showed the greatest yearly accumulation of total carbon (2.07 Kg C yr ^?1 stem^?1). Other species with high growth rates were Cordia alliodora and Centrolobium ochroxylum. This restoration program demonstrates that the dry forest restoration with timely maintenance and protection from fire and grazing is possible with intensive tree planting of native species.     

Tree planting by small producers in the tropics: A comparative study of Brazil and Panama

Forest regrowth is a notable phenomenon across the tropical forest latitudes. Such reforestation takes place in the wake of land abandonment, occurs cyclically in a rotational agricultural system, and may result from the deliberate planting of trees by farmers. Although less extensive than successional forest regeneration, tree planting by small farmers can have potentially important environmental impacts at both the site and global scale. This paper examines tree-planting efforts by small farmers in the tropical frontier regions of Panama and Brazil in order to gauge the magnitude of reforestation activities, and to identify factors that influence these efforts. This paper discusses the environmental regulations, forestry law, and tenure institutions in both countries, and performs a comparative analysis of reforestation efforts with information derived from household surveys (n = 356) and in-depth tree planting interviews (n = 35). Results from logistic regression are also presented. Our results show that tree planting occurs more frequently in Panama, which we attribute to greater external support in the provisioning of extension and materials and strong market incentives. We suggest that the proximity of the study sites to Panama City facilitates external support and market drivers. Finally, our results suggest that land tenure security is an influential determinant of reforestation activities in both countries.This revised version was published online in November 2005 with corrections to the Cover Date.     

Tree planting in Tigray, Ethiopia: the importance of human disease and water microdams

We examine the decision to plant trees and level of tree planting for two sites, public microdam areas and household agricultural land, and two species groups in Tigray, Ethiopia. Both sites are not perfect substitutes, as they vary with respect to distance from the household and tenure security. The role of permanent pooled water irrigation microdams to tree planting is important but unknown, because water borne diseases, which may influence household income and productivity, are thought to be enhanced by the dams. We find both disease and microdams to be important predictors to tree planting. Disease seems more important in determining whether households plant at all, and less important in the level of planting for those that do plant. For example, disease increases the probability of planting both eucalyptus and other species groups on household-own land, but households suffering from malaria plant higher-cost eucalyptus trees with lower probability at both sites, while planting of other lower-cost species increases at dam sites where other villagers can monitor the trees. We also establish a connection between planting and agricultural residues, finding a strong substitution effect on own-land. Microdam access and age are also important. Households living nearer to dam sites will plant both species groups there with higher probabilities, but the decision to plant on agricultural own-land is not affected. For older dams with more developed irrigation, households are more likely to grow crops rather than plant trees on their own land, but they plant more trees at the dam sites. This revised version was published online in June 2006 with corrections to the Cover Date.     

Intensive tree planting facilitates tropical forest biodiversity and biomass accumulation in Kibale National Park, Uganda

The extensive area of degraded tropical land and the calls to conserve forest biodiversity and sequester carbon to offset climate change demonstrate the need to restore forest in the tropics. Deforested land is sometimes replanted with fast-growing trees; however, the consequences of intensive replanting on biomass accumulation or plant and animal diversity are poorly understood. The purpose of this study was to determine how intensive replanting affected tropical forest regeneration and biomass accumulation over ten years. We studied reforested sites in Kibale National Park, Uganda, that were degraded in the 1970s and replanted with five native tree species in 1995. We identified and measured the size of planted versus naturally regenerating trees, and felled and weighed matched trees outside the park to calculate region-specific allometric equations for above-ground tree biomass. The role of shrubs and grasses in facilitating or hindering the establishment of trees was evaluated by correlating observed estimates of percent cover to tree biomass. We found 39 tree species naturally regenerating in the restored area in addition to the five originally planted species. Biomass was much higher for planted (15,675 kg/ha) than naturally regenerated trees (4560 kg/ha), but naturally regenerating tree regrowth was an important element of the landscape. The establishment of tree seedlings initially appeared to be facilitated by shrubs, primarily Acanthus pubescens and the invasive Lantana camara; however, both are expected to hinder tree recruitment in the long-term. Large and small-seeded tree species were found in the replanted area, indicating that bird and mammal dispersers contributed to natural forest restoration. These results demonstrate that intensive replanting can accelerate the natural accumulation of biomass and biodiversity and facilitate the restoration of tropical forest communities. However, the long-term financial costs and ecological benefits of planting and maintaining reforested areas need to be weighed against other potential restoration strategies.     

Improving adoptability of farm forestry in the Philippine uplands: a linear programming model

In the Philippines, smallholder farmers have become major timber producers. But the systems of timber production practiced have several limitations. In intercropping systems, the practice of severe branch and/or root pruning reduces tree-crop competition and increases annual crop yields, but is detrimental to tree growth and incompatible with commercial timber production. In even-aged woodlots, lack of regular income and poor tree growth, resulting from farmers’ reluctance to thin their plantations, are major constraints to adoption and profitable tree farming. In the municipality of Claveria, Misamis Oriental, the recent practice of planting trees on widely spaced (6–8 m) contour grass strips established for soil conservation suggests ways to improve the adoptability (i.e., profitability, feasibility and acceptability) of timber-based agroforestry systems. Assuming that financial benefits are the main objective of timber tree farmers, we develop a simple linear programming (LP) model for the optimal allocation of land to monocropping and tree intercropping that maximizes the net present value of an infinite number of rotations and satisfies farmers’ resource constraints and regular income requirements. The application of the LP model to an average farmer in Claveria showed that cumulative additions of widely spaced tree hedgerows provides higher returns to land, and reduce the risk of agroforestry adoption by spreading over the years labour and capital investment costs and the economic benefits accruing to farmers from trees. Therefore, incremental planting of widely spaced tree hedgerows can make farm forestry more adoptable and thus benefit a larger number of resource-constrained farmers in their evolution towards more diverse and productive agroforestry systems.     

Establishment of native tropical timber trees in monoculture and mixed-species plantations: Small-scale effects on tree performance and insect herbivory

Plantations of native timber species established on former pasture are a potential strategy to reduce the logging pressure on remnant natural forests in the tropics. Such plantations may help to mitigate or reverse the negative impacts of land degradation, and they may contribute to the long-term livelihood of livestock farmers. Planting native trees is, however, perceived as a risky activity due to limited knowledge of their performance and due to marked losses of newly established seedlings attributed to insect pests. Our study focuses on the small-scale effects of environmental heterogeneity, tree diversity and insecticide treatment on the performance of three native Central American timber species two years after establishment, and on damage inflicted by insect herbivores. Growth, survival and herbivore leaf damage were quantified for Anacardium excelsum (Anacardiaceae), Cedrela odorata (Meliaceae) and Tabebuia rosea (Bignoniaceae) planted in an experiment in Panama as (1) monocultures, (2) mixed stands, and (3) mixed stands protected by insecticides. Our study revealed that small-scale effects can have a substantial impact on the success of timber trees planted on former pasture. Growth performance and survival of the three species was strongly affected by small-scale environmental heterogeneity, which was expressed as significant differences in growth and survival among different plantation plots at the same study site. Establishment of trees in mixed stands did not have significant effects on tree survival and growth compared to pure stands, although it reduced herbivore pressure in one of the studied tree species. All tree species grew best and suffered lowest leaf damage when protected by insecticides, indicating a general influence of insect herbivory on growth of newly established trees. In contrast to growth performance, survival was not significantly affected by different management practices or herbivory. The large variability among plots in tree survival and growth, and also in the effects of management practices such as planting design and insecticide treatment, emphasizes the importance of small-scale environmental heterogeneity on tree survival and growth.     

广州城市园林绿地食源树种应用及其生态景观营造

【目的】城市鸟类及蜂蝶类群落结构多样性是城市生物多样性的重要组成部分和生态环境质量的重要评价标准之一。了解城市园林绿地食源树种的应用,探讨食源树种的生态特性及其景观特征。【方法】调查了广州市12处城市绿地食源树种的应用,观察鸟类及蜂蝶类的取食情况,记录绿地中食源树种的种类、应用频次、配置方式及生长状况,归纳分析食源树种群落构成、观赏特性、生态性等。【结果】广州园林绿地常见食源树种有23科33属45种,其中鸟类食源树种25种,蜂蝶类食源树种20种,以桃金娘科、桑科、大戟科为主,并以乡土种为主、外来驯化种为辅;食源树种的观赏与被取食时间主要集中于春夏季(31种),秋冬季的相对较少(16种)。【结论】广州市绿地的食源树种资源丰富,约300种,但常见鸟类及蜂蝶类食源树种仅占18%,结合良好适应性、抗逆性、取食性、观赏性,归纳出20种优势乡土食源树种;同时对食源树种群落组成结构进行分析,在兼顾食源树种的观赏性与生态性的基础上,设计出6组可在广州城市园林绿地推广应用的生态配置群落,采用树种混交与立体复合栽植的形式,乔、灌、草结合,形成丰富的水平与垂直植物群落结构,提高绿地生境异质性,增加食源树种的数量及丰富度,提供丰富稳定的食物来源,为鸟类及蜂蝶类提供多样的稳定栖息地。研究结果将对城市园林绿地食源树种的合理选择与科学配置应用,以及营造鸟类及蜂蝶类昆虫生态景观具有现实指导意义。     

兰州市行道树树种现状调查

通过对兰州市区道路树种的栽植概况调查,结果表明：目前兰州市区共有行道树树种18种,其分属10科12属,乡土树种占61%,其中刺槐、臭椿、旱柳、白蜡等树种后期病虫害严重,作为行道树将在兰州市区逐渐淘汰。并通过对行道树种的选择,筛选出适宜市区可推广应用的行道树种类有21种,分属13科17属,乡土树种占62%。     

Remnant trees and the conservation of biodiversity in Costa Rican pastures

Remnant trees may play an important role in conserving biodiversity within agricultural systems because they provide habitats and resources that are otherwise absent from agricultural landscapes. In order to determine the potential importance of remnant trees for conservation, we surveyed the density and species composition of remnant trees occurring in pastures of 24 dairy farms near Monteverde, Costa Rica. In addition, we conducted interviews with farmers to determine why they leave trees in pastures and how they manage them. In our survey of 237 ha of pastures, we counted 5583 trees of 190 species (mean density of 25 trees/ha). Primary forest trees accounted for 57% of all of the species and 33% of tree individuals. Over 90% of the species are known to provide food for forest birds and other animals. In addition, many of the species are important locally for humans as sources of timber (37%), firewood (36%) or fence posts (20%). Farmers mentioned 19 reasons for leaving trees in pastures. Of these, shade for cattle, timber, fruits for birds and fence posts were most commonly cited. Most farmers were well aware of both the economic and ecological benefits of pasture trees, and were interested in the possibility of increasing tree cover within their pastures. Although the current densities and richness of pasture trees in Monteverde are high, the size distribution indicates that diversity will decrease substantially in future years, both because farmers are harvesting trees and because saplings of primary forest trees are scarce within the pastures. This revised version was published online in June 2006 with corrections to the Cover Date.     

Conservation value of dispersed tree cover threatened by pasture management

Trees dispersed in pastures are a prominent feature of many Central American landscapes, particularly in cattle producing regions where farmers retain trees to serve as shade, fodder, timber and firewood. The presence of dispersed trees in pastures is often considered as important for the conservation of biodiversity by providing habitat and enhancing landscape connectivity. However, despite their critical productive and environmental roles, little is known about tree distribution within pastures or how farmers’ management decisions influence the trees themselves and their impact on farm productivity and biodiversity conservation. Here, we present a synthesis of (a) the abundance, composition, and size of dispersed trees in four important cattle producing regions of Costa Rica (Caňas and Río Frío) and Nicaragua (Rivas and Matiguás), based on inventory of 18,669 trees on 1492 ha of pasture, (b) the local knowledge, management and use of trees by cattle farmers, and (c) opportunities for ensuring sustainable management of dispersed trees in pasture-dominated landscapes. Dispersed trees were common in all four landscapes, with mean frequency ranging from 8.0 trees ha⁻¹ in Caňas to 33.4 trees ha⁻¹ in Matiguás. A total of 255 tree species were found in pastures across the four landscapes. The total number of tree species per landscape varied from 72 in Rivas to 101 in Caňas and Rio Frio, with mean species richness per farm ranging from 22.9 in Rio Frio to 45.9 in Matiguás. In all four landscapes, a handful of tree species dominated the pastures, with the ten most abundant species in each landscape accounting for >70% of all trees recorded. Most of these common tree species provide fruits or foliage eaten by cattle, or are important timber or firewood species, and are deliberately retained by farmers for these uses. In all four landscapes, farmers had a detailed knowledge of tree attributes affecting pasture and animal productivity, and influenced tree cover through pasture management activities and occasional tree cutting. Current farm management practices are gradually decreasing the diversity of trees in pastures, and in some cases also tree density, reducing their contribution to farm productivity and biodiversity conservation. To reverse this trend, incentives are required to encourage cattle farmers to retain and enhance tree cover in pastures, through the adoption of pasture management practices that favor the regeneration and persistence of a diverse range of tree species.     

Financial evaluation of smallholder timber-based agroforestry systems in Claveria, Northern Mindanao, the Philippines

In the Philippines, timber production on small farms has become profitable as a result of reduced supplies due to extensive deforestation and increasing demand. In the early 1990s, when the price of timber was high, farmers were promised huge returns from tree farming. However, widespread planting of few species has led to oversupply and a sharp decline in the price of farm-grown timber. Moreover, low intercrop yields as a result of competition from fast-growing trees and low timber yields due to poor tree management, further reduce net economic returns. In spite of this, interest in tree farming remains high. This paper examines the private profitability of two tree-maize systems, namely trees in blocks and trees in hedgerows, compared with the alternative of maize monocropping. The analysis reveals that maize monocropping provides higher returns to land at the current timber price, but considerably lower returns to labour, than the maize-tree systems tested. This suggests that tree farming is a more attractive option for labour and capital-constrained households or those with off-farm opportunities that compete for their labour. These farmers may raise productivity and income by planting trees on the excess land that cannot be devoted to annual crops. The analysis also indicates that wide-spaced tree hedgerows are superior to tree blocks, due to lower establishment and management costs, longer periods of viable intercropping and more rapid tree growth.     

Foresters' beliefs about farmers: a priority for social science research in social forestry

This study suggests that one of the most productive (and most neglected) uses of social science research in forestry development projects is to examine foresters' beliefs regarding rural peoples. This suggestion is illustrated with data from the Forestry Planning and Development Project, Pakistan's first nation-wide social forestry project.The operational component of this project, intended to assist small farmers to cultivate trees on their farmlands, ran into immediate difficulties. Many of the foresters involved insisted that small farmers were simply not interested in tree cultivation. A comprehensive base-line study subsequently was carried out to examine the validity of this belief. The results of this study (confirmed by the subsequent experience with the project in the field) varied markedly from the foresters' beliefs.While many of the foresters believed small farmers were opposed to having trees on their farms and would not agree to plant trees under the project, most farmers already had trees on their farms and expressed interest in planting more; while many foresters believed farmers would only be interested in planting large blocks of market-oriented exotics, most farmers requested small plantings of multi-purpose native trees; while many foresters believed farmers would plant trees only for market sale, most farmers requested trees to meet household needs for fuel and timber; and while many foresters did not think that increasing supplies of fuelwood could reduce the burning of dung, all of the evidence provided by the farmers suggested that it would.The disparity between farmer reality and forester belief is attributed to failures on the part of both foresters and social scientists — failure by foresters to distinguish their non-empirical beliefs about farmers from their empirically-based knowledge of trees, and failure by social scientists to recognize the belief systems of foresters as a legitimate and important object of study. Their study comprises three parts: finding out what the foresters think the farmers want, finding out from the farmers what they actually want, and then analyzing and explaining the differences.     

Socio-economic factors affecting smallholder tree planting and management intentions in Leyte Province,Philippines

Survey data have been collected from four rural communities in Leyte Province, the Philippines, on household tree planting and management intentions, as well as the socio-economic characteristics, attitudes to tree planting and management, farming practices and the number of trees planted. In relation to intended tree planting and management activities, respondents were asked a series of structured questions as to what tree species they intend to plant, how many individual trees of each of those species they plan to plant, and for what purpose they propose to plant each of the tree species. This paper reports the results of analyses of the stated tree management intentions of households in the four communities and identifies the socioeconomic factors that influence householders’ tree management intentions. Respondents were generally enthusiastic about the possibility for further tree planting on their land, 75% indicating they would undertake planting. About 60% indicated an interest in commercial tree farming, with no significant differences in this level of interest between communities. It was found that a number of socio-economic variables indicating higher levels of land ownership and previous experience in the forestry industry are related to the intention to plant trees to produce timber for sale. These include the use of materials from public lands, participation in community organisations and community forestry programs, and the present management of trees to produce timber for sale.     

Understanding forest transition in the Philippines: main farm-level factors influencing smallholder’s capacity and intention to plant native timber trees

The ‘when, where and how’ of decisions by smallholder upland farmers to plant trees as part of their use of natural, human and capital resources needs to be understood if policy support is to result in actual recovery of tree cover as part of a ‘forest transition’ trajectory. In large parts of the Philippines the turning point may be close. Data on resource access and tree planting decisions were gathered from a household survey, with a total of 148 respondents in four rural communities in Leyte Province in Central Philippines. Data were analysed using logistic regression analysis. Household-level results reveal that the outcomes of the decision-making processes primarily depend on the availability of land and access to remaining forest resources rather than socio-cultural or economic factors. The total area of land and number of parcels managed by the household plus security of land tenure through ownership was found to have a statistically significant effect on farmers’ decision to plant native timber trees. Access to surrounding natural forest is negatively associated with farmer tree planting.     

Influence of herbaceous ground cover on forest restoration of eastern US coal surface mines

Competitive effects of dense herbaceous vegetation (ground cover) can inhibit forest restoration on mine sites. Here we review the evidence of ground cover interactions with planted tree seedlings on coal surface mines of the eastern US, discuss recent research into these interactions, and draw conclusions concerning ground cover management when restoring forests on reclaimed coal mines. Reclaimed mine sites have a high potential to support productive forests, however forest establishment is inhibited by reclamation practices that included soil compaction, and the seeding of competitive ground covers. In the first few years after tree planting, a dense ground cover of grass and legume species commonly seeded on mine sites often affect growth and survival negatively. Herbaceous vegetation providing less extensive and competitive ground coverage may either facilitate or inhibit tree establishment, depending on site conditions. The use of quality planting stock promotes the competitive ability of seedlings by improving nutrient status and the ability to capture available resources. Herbaceous species have contrasting functional characteristics, and thus compete differently with trees for available resources. Negative interactions with trees are more frequently reported for non-native cool-season grasses than for native warm-season grasses, while the effects of legumes change over time. Further research is needed to fully understand the mechanisms of tree/ground cover interactions. The development of seeding mixes that can control erosion, facilitate survival and growth of planted trees, and allow establishment by unplanted native species would aid forest restoration on eastern US, coal mines.

     

Tree planting on deforested farmlands, Sewu Hills, Java, Indonesia: Impact of economic and institutional changes

This paper analyses how farmers in the Sewu Hills, a limestone hill complex in Java, were led to grow trees on their holdings. In the past, the area had been notorious for its deforestation, soil erosion and poverty. A historical analysis of change in land use in the area bears out a number of factors that were conducive to these changes: increased crop productivity, depletion of forest resources, better marketing opportunities for tree products, greater security of the trees and increased non-agricultural employment opportunities. Tree planting campaigns launched by the government provided the population with important incentives, establishing a critical mass of farmers adopting tree growing. Trees and their products have diversified, buffered and expanded household economies in the Sewu Hills. Yet it remains to be seen whether the system can withstand the economic and social pressures the area is facing today. A historical, multi-level analysis of land use adds considerably to cross-sectional studies in understanding what causes farmers to grow trees on their fields. This revised version was published online in June 2006 with corrections to the Cover Date.     

Tree Planting by Smallholder Farmers in the Upper Catchment of Lake Tana Watershed,Northwest Ethiopia

Tree growing by smallholder farmers is an emerging livelihood strategy in Lake Tana catchment. The objectives of this study were to (1) identify the most important tree species grown, (2) investigate the drivers of the existing pattern, and (3) identify determinants of the number and diversity of tree species and their spatial patterns. Survey data were collected from 200 households. Multiple linear regression was employed to identify the determinants of tree growing behaviour of households and spatial variables affecting the abundance of tree species. Eucalyptus globulus, Acacia decurrens and E. camaldulensis dominate woodlots. Only a fraction of the forest production is used by the households, the rest being sold as poles or charcoal. Location in relation to market centres, number of livestock owned, landholding size and age of household head were found to positively affected the number of tree species and trees grown. Gender affected the species and spatial pattern of trees. Woodlots, farm boundaries and homesteads were found to be important tree growing niches. These results substantiate the proposition that farmers assign their parcels of land to uses that increase the rent value of the land, and this value is affected by access to roads. Woodlots are on the increase at the cost of productive agricultural land. Provision of a tree planting extension service may increase participation of farmers in tree planting, and a management-oriented tree planting extension service may give desirable results.