Does outmigration lead to land degradation? Labour shortage and land management in a western Nepal watershed

Applied Geography 62 (2015) 157e170

Contents lists available at ScienceDirect

Applied Geography journal homepage: www.elsevier.com/locate/apgeog

Does outmigration lead to land degradation? Labour shortage and land management in a western Nepal watershed S. Jaquet a, *, G. Schwilch a, F. Hartung-Hofmann a, A. Adhikari b, K. Sudmeier-Rieux c, G. Shrestha d, H.P. Liniger a, T. Kohler a a

Centre for Development and Environment, University of Bern, Hallerstrasse 10, 3012, Bern, Switzerland International Union for Conservation of Nature, Nepal Country Office, Kupondole, Lalitpur, P.O. Box 3923, Kathmandu, Nepal Institute of Earth Science, University of Lausanne, Geopolis, 1015, Lausanne, Switzerland d Nepal Centre for Contemporary Research, Ekantakuna, Jawalakhel, Lalitpur Sub-metropolitan City, Nepal b c

a r t i c l e i n f o

a b s t r a c t

Article history: Available online

In Nepal, changing demographic patterns are leading to changes in land use. The high level of outmigration of men in the hills of Kaski District, Western Development Region of Nepal, is affecting the household structure but also land management. Land is often abandoned, as the burden on those left behind is too high. How do these developments affect the state of the land in terms of land degradation? To find out, we studied land degradation, land abandonment caused by outmigration, and existing sustainable land management practices in a subwatershed in Kaski District. Mapping was done using the methodology of the World Overview of Conservation Approaches and Technologies (WOCAT). While previous studies expected land abandonment to exacerbate slope erosion, we demonstrate in this paper that it is in fact leading to an increase in vegetation cover due to favourable conditions for ecosystem recovery. However, negative impacts are several, including the increase of invasive species harmful to livestock and a decline in soil fertility. Traditional land management practices such as terraces and forest management exist. To date, however, these measures fail to take account of the changing population dynamics in the region, making the question of how migration and land degradation are linked worth revisiting. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Land degradation Land abandonment Forest Outmigration Labour shortage Nepal

Introduction The relationship between land degradation and population trends has always been of great concern in the Himalayan region: an increasing population pushed people to the forest frontiers, and deforestation was one of the main causes of land degradation, for example through landslides. Land degradation in itself has been studied for decades in the Himalayas due to the region's dynamic landscape and monsoonal climate. The first theory about

* Corresponding author. Centre for Development and Environment, University of Bern, Hallerstrasse 10, 3012 Bern, Switzerland. Tel.: þ41 31 631 88 22, þ41 79 722 43 55 (mobile); fax: þ41 31 631 85 44. E-mail addresses: [email protected] (S. Jaquet), gudrun.schwilch@ cde.unibe.ch (G. Schwilch), [email protected] (F. Hartung-Hofmann), [email protected] (A. Adhikari), karen.sudmeier@ gmail.com (K. Sudmeier-Rieux), [email protected] (G. Shrestha), hanspeter. [email protected] (H.P. Liniger), [email protected] (T. Kohler). URL: http://www.cde.unibe.ch http://dx.doi.org/10.1016/j.apgeog.2015.04.013 0143-6228/© 2015 Elsevier Ltd. All rights reserved.

Himalayan land degradation was discussed in 1975 by E. Eckholm, whose paper subsequently led to the theory of Himalayan Environmental Degradation e a theory many scientists adopted over the years (Ives & Messerli, 1989). This scenario forecast that the hills of Nepal would be barren by the year 2000 (The World Bank, 1979). Although widespread land degradation did indeed occur, the causeeeffect relationship was not as simple as suggested, and the theory was widely criticized (W. M. Fleming, 1985; Gardner & Gerrard, 2003; Gerrard & Gardner, 2002; Ives, 2004; Ives & Messerli, 1989; G. Thapa & Weber, 1995). Similar predictions were made for the Phewa watershed (about 100 km2, in Kaski District, Western Development Region of Nepal), where it was said the hills would be barren by 2002 (W. M. Fleming, 1985). This predicted degradation was mainly blamed on open grazing, which was very common in the forest. The estimated number of livestock within the watershed was one per inhabitant (W. M. Fleming, 1985). The Department of Soil Conservation and Watershed Management intervened (1975e1995), and the Food and Agriculture Organization of the United Nations prepared a

158

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

study for a new basin management programme (1977e1978) (Awasthi, Sitaula, Singh, & Bajacharaya, 2002; B. Fleming & Fleming, 2009; W. M. Fleming, 1985; G. S. Paudel & Thapa, 2001): these measures were rooted in the belief that a bigger population was putting more pressure on resources. The objective was to use forests and cultivated lands sustainably, by limiting the loss of soil and nutrients and by including the government and international agencies in land use management. Several researchers have studied the implementation and impacts of the Phewa Lake Watershed Area Management Project and compared the area to other nonproject areas (Awasthi et al., 2002; G. S. Paudel & Thapa, 2001; G. Thapa & Weber, 1995). Conclusions were positive: technical help, awareness raising, and information were leading to more sustainable land use (e.g. implementation of bunds and waterways to improve water flow), even though some problems of degradation still remained (G. S. Paudel & Thapa, 2001). However, the recent past has seen a change in population dynamics due to increased outmigration. While people have migrated for decades from the hills to the low plains of Terai or to India when land became scarce (Graner & Gurung, 2003; Kollmair, Manandhar, Subedi, & Thieme, 2006; Seddon, Gurung, & Adhikari, 1998), the current exodus to foreign countries or to urban areas has led to marked changes in the demographics of the hills (Blaikie, Cameron, & Seddon, 2002; Ghimire & Upreti, 2012; Seddon et al., 1998). According to 2014 statistics, 1,3491 people leave the country every day (not counting those who leave through the open border with India or illegally). The task of managing the land falls to those left behind, mainly women and the elderly. Some villages even turn into ghost villages when entire families decide to move down to the valley or city (Gurung, personal communication, 2013). Studies investigating how demographic change affects land management have so far tended to focus mostly on whether population growth increases or decreases land degradation (G. Thapa & Weber, 1995; Warren, 2002), whereas the consequences of population decline have received little attention. Outmigration can have very different impacts on land, including land degradation and a decrease in production; land abandonment and recovery of ecosystems; and changes in land use/land cover (Aide & Grau, 2004; Gisbert, Painter, & Quiton, 1994; Radel & Schmook, 2008). The lack of knowledge on the effects of population decline makes the question of land degradation in the Phewa watershed worth revisiting. Migration has a long history in this area, with the Indian and British Armies recruiting Gurkha soldiers since the 19th century; but it has changed and increased significantly in the past ten years (Sharma, Pandey, Pathak, & Sijapati-Basnett, 2014). Today, it is not population growth which affects the landscape and the environment, but depopulation. The aim of our study in the Harpan subwatershed (a subset of Phewa watershed) was threefold: to investigate land degradation, land abandonment caused by outmigration, and existing sustainable land management practices. Study area The study area is located in Kaski District in the Western Development Region of Nepal (83 470 5600 to 83 520 5900 E and 28 1103900 to 28 150 2800 N), west of Phewa Lake and Pokhara Sub Metropolitan City. The area lies between 825 m.a.s.l. (Thulakhet village) in the east and 2517 m.a.s.l (top of Panchase) in the west (Fig. 1). The Harpan subwatershed comprises the upper part of the Harpan River, which is the main river flowing west to east and the

1 http://www.onlinenepalnews.net/2014/05/labor-migration-has-becomefashion-in.html, retrieved on 28.05.2014.

major source of water to Phewa Lake, further east. The study area comprises 36 km2 of mainly steep slopes (slope gradients of 16  e30  ) The climate of this area is subtropical humid to temperate humid, with temperatures ranging from 34  C during summer to as low as 3  C in the winter. Monsoon (June to September) sees around 80% of the rainfall, with an annual average of over 3500 mm (Dahal & Hasegawa, 2008). The area is mainly covered by forest (70%) (Fig. 1) and cultivated land (26%), the latter consisting of irrigated (khet) and non-irrigated (bari) plots. Forest covers mainly the western and southern part of the watershed. The villages located in the upper part (Bhanjyang, Upper/Lower Shidane, Mankanpur, Philinghari, Chisopani, Kuiredanda, etc.) have relatively small cultivated areas compared to the villages further downstream (Thulakhet, Ghatichina, Borang, etc.), where forest covers smaller areas. The Harpan River is joined by the Andheri River coming from the north near the village of Thulakhet. The area comprises 21 communities. Their size ranges from 6 to 150 households belonging to various ethnic/caste groups. Migration has a long history here and is significant: many men have migrated temporarily for employment and many families have moved from the hills down to the major cities and more accessible areas of Nepal. A similar phenomenon occurs at a more local scale within the study area: the population is decreasing in the more remote uphill villages on the slopes and increasing in the more easily accessible villages at the bottom of the subwatershed (Ghatichina, Thulakhet). Materials and methods Materials To record information on the households, an existing questionnaire (Sudmeier-Rieux, Jaquet, Derron, Jaboyedoff, & Devkota, 2012) was modified to fit the context of this research with a special focus on outmigration. Two land use/land cover maps and a satellite image were used for fieldwork and to present results (Table 1). The Global Positioning System (GPS) was used to locate the main features of the area and to record the position of the households surveyed. Demographic data and household survey We conducted an in-depth demographic and household survey in six selected villages, obtaining a detailed understanding of the residents and the absentee population. Villages were selected according to several criteria intended to cover a wide range of issues. The main criteria were: significance of land abandonment (high/ low), outmigration (high/low), land degradation (very much/little); location within the watershed (uphill/middle/downstream); and caste/ethnicity groups (indigenous/advantaged/marginalized). The in-depth demographic survey only covered gender, age, and the destination and work of migrant household members. The household survey covered the 5 livelihood assets (DFID., 1999) and focused specifically on outmigration. We used a systematic sampling method to select and survey 10e30 percent of the households in each village (91 households in total) during two periods, March 2013 and March 2014. This paper survey was conducted by the researchers in collaboration with field assistants. Additionally, in each village we conducted a group discussion with women and one with men. When it was not possible to have separate discussions, we conducted a mixed focus group discussion. Topics covered the village features; outmigration, its causes, consequences, and disadvantages; land management, including land degradation issues and how land is managed following outmigration; and the main concerns of the population. Since this article focuses on the

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

159

Fig. 1. Land use/land cover map of Harpan subwatershed, Nepal. The map was provided by the Machhapuchhare Development Organization, a local NGO, and is based on a Rapid Eye satellite image.

mapping outcomes of our study, we present here the survey results that relate to outmigration, land use/land cover, land degradation, and land conservation. Other results will be used in future papers with a different focus. Land use/land cover mapping During desk study, a satellite image recorded in 2012 by the GeoEye-1 satellite, consisting of a panchromatic band with a resolution of 0.5 m, and a second satellite image recorded in 2012 by the IKONOS satellite, consisting of 4 spectral bands (red, blue, green, near infrared) with a resolution of 2 m, were orthorectified, subsequently merged, and pansharpened using the PCI Geomatics remote sensing software (Table 1). The topographic coordinates of the resulting ortho-image were validated during a 3-month field survey from January to March 2014 in the Harpan subwatershed using the Global Positioning System (GPS) and a

Garmin GPS receiver. Evenly distributed and fixed waypoints such as temples and other buildings were located by mans of the GPS for comparison with the topographic information in the final ortho-image. In the initial stage of the field survey, we visited different villages to obtain information about the present land use/land cover. Based on this survey, we defined eight land use/land cover categories (Table 2). Subsequently, we manually mapped the present land use/land cover (as observed in March 2014) based on visual assessment and information gathered from the local population, using the eight categories. Mapping was done directly in the orthoimage (scale: 1:4000). The map was then digitalized using the ArcGIS Geographic Information System software. Land was defined as abandoned if it had no longer been used for cultivation (for a minimum of two years) and had been converted into a different land use/land cover type (shrubland, grassland, or grazing land). Identification of abandoned plots required

Table 1 Details of land use/land cover maps and satellite image. Type of map and year

Provider

Purpose

Land use/land cover map compiled from a Rapid Eye satellite image, 2012

Machhapuchhre Development Organization (MDO), a local NGO

Satellite image, Geoeye1 and Ikonos, 2012

Purchased by the University of Lausanne, Switzerland Survey Department, Government of Nepal

Base map for land use/land cover mapping using the World Overview of Conservation Approaches and Technologies (WOCAT) methodology Base map for (manual) mapping of present land use/land cover (time of fieldwork, March 2014) This map was used for comparison over time; the date of 1996 was chosen because this was the most recent available map predating the Maoist insurgency (which led to an increase in outmigration)

Land use/land cover map (1:25,000) compiled from 1:50,000-scale aerial photography, 1996 (land use/land cover categories: cultivation, forest, grass, bush, sand, river)

160

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

Table 2 Land use/land cover categories defined for the mapping. Land use/land cover Category

Description

Forest Cultivated land Shrubland Grassland Grazing land

Dense tree cover Mostly terraced, identifiable by its levelled structure; can be khet (irrigated land) or bari (rainfed land) Covered by shrubs and bushes Covered by high grasses and to a minor part with bushes Covered by short grass with the presence of dung from pasturing cattle; often zoned by stonewalls or fences to protect cultivated or other land use/land cover plots against damage from cattle Cliffs and uncovered surfaces along steep slopes; a barren surface can also occur in connection with erosion (mostly within forest, e.g. after a landslide) Land prepared for the construction of new buildings (existing settlements were not included) Channels of streams and rivers as well as the uncultivated area along them

Barren land Built-up Riverbed

Table 3 Selected types of land degradation according to WOCAT (Liniger et al., 2008). Type of degradation

Specific types

W: Soil erosion by water, including

Wg: Gully erosion/gullying Wm: Mass movements/landslides Wo: Offsite degradation effects Wr: Riverbank erosion Wt: Loss of topsoil/surface erosion

E: Soil erosion by wind C: Chemical soil deterioration, including H: Water degradation, including

B: Biological degradation, including

information from the local population, as fieldwork took place during the dry season, when most of the terraces (especially bari) are not cultivated. We found abandoned fields to be characterized by certain visual criteria, namely a visible terrace structure with the following traits (classified through fieldwork): - Development of dense grass cover and compact topsoil (by contrast, the topsoil in seasonally cultivated terraces shows patterns of leftover crop stalks or a loosened surface after ploughing) - Conversion into shrubland (characterized by high shrubs and small trees) - Conversion into grassland (characterized by high grasses and fodder production) - Conversion into grazing land (characterized by trampling and dung of cattle) We mapped abandoned plots according to their land use/land cover, marking whether each mapped polygon occurred on abandoned land or not. We combined all polygons covering abandoned land to a single layer using ArcGIS software and calculated the total extent of the abandoned area and percentages of each land use/ land cover type.

Cn: Fertility decline and reduced organic matter content Ha: Aridification Hg: Change in groundwater/aquifer level Hs: Change in quantity of surface water Bs: Quality and species composition/diversity decline Bc: Reduction of vegetation cover Bp: Increase of pests/diseases

Mapping land degradation and sustainable land management practices using WOCAT tools The mapping methodology of the World Overview of Conservation Approaches and Technologies (WOCAT) (Liniger, Van Lynden, Nachtergaele, & Schwilch, 2008; Schwilch et al., 2011) is a qualitative tool which relies on expert assessments and is used to map multiple types of land degradation (as opposed to only soil erosion by water; see Table 3), as well as sustainable land management practices and land use/land cover trends. The methodology requires a base map with fixed mapping units. Most commonly, two maps e representing the land use/land cover system and administrative boundaries, respectively e are combined to get a workable number of easily identifiable units. The methodology is then implemented with the help of a group of land experts, a questionnaire, and an online database. The experts can have different backgrounds but must have sound land use knowledge. They can be community forest chairmen, farmers, agricultural extension workers, members of organizations working in the area (e.g. local NGOs), or chairwomen of mothers' groups. Expert groups are generally small (1e3 persons) due to the amount of time required for the mapping. In the lower area of the subwatershed, for example, the expert group comprised of the president of the

Table 4 Degrees of degradation according to WOCAT (Liniger et al., 2008). Degree of degradation

Definition

1: Light

There are some indications of degradation, but the process is still in an initial phase. It can be easily stopped and damage repaired with minor efforts. Degradation is apparent, but its control and full rehabilitation of the land is still possible with considerable efforts. Evident signs of degradation. Changes in land properties are significant and very difficult to restore within reasonable time limits. Degradation beyond restoration.

2: Moderate 3: Strong 4: Extreme

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

161

Table 5 Definition of the effectiveness of management measures according to WOCAT (Liniger et al., 2008). Effectiveness of management measures

Definition

1: Low

The measures need local adaptation and improvement in order to reduce land degradation to acceptable limits. Much additional effort is needed to reach a “high” standard. The measures are acceptable for the given situations. However, loss of soil, nutrients, and water retention capacity exceeds the natural or optimal (as with “high”) situation. Besides maintenance, additional inputs are required to reach a “high” standard. Regarding water and vegetation degradation, the measures only slow down the degradation process, but are not sufficient. The measures control the land degradation problems appropriately. For example, soil loss does not greatly exceed the natural rate of soil formation, while infiltration rate and water retention capacity of the soil, as well as soil fertility, are sustained; only maintenance of the measures is needed. Concerning water and vegetation degradation, the measures are able to stop further deterioration, but improvements are slow. The measures not only control the land degradation problems appropriately, but even improve the situation compared to the situation before degradation occurred. For example, soil loss is less than the natural rate of soil formation, while infiltration rate and/or water retention capacity of the soil, as well as soil fertility, are increased; only maintenance of the measures is needed. Either the measures have strongly improved water availability and quality (addressing water degradation), or vegetation cover and habitats have been highly improved (addressing biological degradation).

2: Moderate

3: High

4: Very high

local NGO (running a programme on ecosystem adaptation to climate change), a local representative of the International Union for Conservation of Nature (IUCN), and a farmer. As for the base map, the Machhapuchhre Development Organization, a local NGO, provided a land use/land cover map which had been prepared from a Rapid Eye satellite image (2012) (Table 1). For the administrative boundaries map, we used the local boundary map available from the Survey Department in Kathmandu and representing the Village Development Committee (VDC) boundaries with their subdivisions (wards). In some cases, where the wards covered an area spanning large altitude differences, these were again subdivided following natural boundaries such as river courses. The two maps were then merged using GIS software to produce a map with units designating every type of land use/land cover per administrative unit (e.g. forest within ward x). This map was uploaded into the online WOCAT mapping database (qm.wocat.net) as a shapefile (.shp). To assess land degradation, sustainable land management, and land use/land cover trends, local stakeholders, land users, and land experts first gathered for a workshop where they were introduced to the methodology; fieldwork was then conducted in smaller subteams over the following days. The WOCAT mapping questionnaire guided the assessment of each mapping unit. For each unit, the direct and indirect causes, extent, rate (from 3 ¼ rapidly increasing degradation to 3 ¼ rapidly decreasing degradation), degree (Table 4), and impacts of land degradation were determined. A similar procedure was carried out for sustainable land management practices (extent, effectiveness, purpose, time, impacts) and land use/land cover trends (intensity, surface). The WOCAT definition of effectiveness is presented in Table 5 for illustration. The results were then entered into the online database, enabling an immediate visual overview by means of the resulting maps and allowing the team to present the results to local experts and stakeholders for verification directly after fieldwork. Results

as shown by the age pyramid below (Fig. 2), which excludes the absentee population.2 Kaski District is part of this region and the case study area presents the same characteristics, even though the sample villages are very small. The collected demographic data enabled us to construct age pyramids for the villages of the study area. In all cases, the male population aged between 20 and 44 is mostly absent. Thulakhet (Fig. 2) with a total of 59 households is located downstream, closer and with better access to the major city of Pokhara, but the phenomenon is similar, with a clear absence of men in the same age groups. Chisopani and Philinghari are two villages with a total of 25 households; the age pyramid presents a partial absence of men between 20 and 44 and even a total absence of men between 25 and 35 years old. These villages are located uphill, are difficult to access and lack infrastructure such as a road and or a health post. Fig. 3 maps the selected villages and represents the percentage of households with migrant members and households without migrant members. Of the villages studied, the majority has more than 67% of households with at least one migrant member. The villages located uphill (Shidane, Makwanpur, Chisopani, and Philinghari) have a higher percentage of households with migrant members than the villages downstream in the valley floors (Ghatichina and Thulakhet). Land use/land cover change and land abandonment Fig. 4 shows superposed forest cover from 1996 (topographic map) and 2014 (land use/land cover map from fieldwork), revealing a 12% forest increase between 1996 and 2014. The current forest cover can be visualized by adding the increase of forest area 1996e2014 to the stable forest area. The increase is particularly clear in the northern, western, and southern (higher) parts of the watershed, and less so in the eastern (lower) part of the watershed. Abandoned areas currently represent 22% of cultivated land: the map shows that areas with forest increase and land abandonment are often adjacent. The land abandoned in 2014 has been converted into three types of land cover: shrubland, grazing land, and grassland (Table 6). Two-thirds of abandoned land (67.3%) is covered by

Demography of the area At the regional level (Western Development Region), migration of the male population between 20 and 44 years old is very visible

2 Absentee: Persons away or absent from their birthplace or usual place of employment, study, or business are considered part of the “absent population” and are not counted as the present population (GoN, 2011).

162

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

Fig. 2. : Age pyramids of the Western Development Region in Nepal, based on data from the official population census (GoN, 2011), as well as the village of Thulakhet and the villages of Chisopani and Philinghari, based on data from our demographic survey in 2014.

shrubland, with banmara (Ageratina adenophora) being one of the first species to grow (Fig. 5). Terraces are invaded by grass or bushes and slowly disappear, even though their shape may still be visible. A total of 19.2% of the abandoned land is now covered by grassland. We assume this land was abandoned recently, as local land users said that grass can only be used (for cut-and-carry stall feeding) for about two years after abandonment. The map also shows land that is cultivated or has another land use/land cover e that is, land which has not been abandoned or has remained non-forested. This land might be left fallow and, eventually, converted into shrubland or forest in the future.

Our household survey results confirm those from the mapping. Within the study area, 42.9% of the households interviewed have abandoned land. Of these, 69% said the main reason was lack of labour: Women and elderly people in particular stay behind and take care of the land in addition to their usual tasks of looking after their children, house, and livestock. They are overburdened and often decide to leave part of their land unused. Some respondents mentioned other factors such as the distance between their land and home, lack of water, low production, or loss of land due to natural disasters. Some households with migrant members decided not to abandon land, either because they own only very little land

Fig. 3. Size of selected villages and proportion of households with migrant members and households without migrant members, Harpan subwatershed, Nepal.

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

163

Fig. 4. Forest change (1996e2014) and current land abandonment, Harpan subwatershed, Nepal.

(some less than 0.15 ha) or have enough male members at home to work in the field. Finally, some households possess no land at all. Land degradation and land management practices mapping The main type of degradation found in the area is biological; specifically, changes in the quality and composition of species (especially invasive species) and decline in vegetation diversity (68% of the cultivated area and 30% of forest is affected). Further degradation types are: chemical deterioration (soil fertility decline and reduced organic matter content, affecting 32% of the cultivated area) and reduction of vegetation cover (affecting nearly 9% of the forest area) (Table 7). Only few areas are affected by mass movements/landslides (0.5%), gullying (0.3%), riverbank erosion (1%), and surface erosion (1%) within the watershed, even though the slope is generally steep. However, these mass movements affect several land use/land cover types: cultivated land (5%), forest (1%), built-up area (5% affected by off-site degradation), and areas along the river (riverbed area, 60%). Topsoil erosion is often seen as a normal issue of terrace management and is not considered degradation by local land users. It is worth noting that riverbank erosion, which affects only a very small portion of the area (1%), is nonetheless located in the most populated area with access to the main road. Several problems related to water, such as aridification and change in surface water quantity have also been found in the area. It is not clear if they are linked to a change in rainfall or to a lack of irrigation management. Of the entire study area, half of the land (51%) is not affected by any type of land degradation (62% of forest and 24% of cultivated land are free from degradation) (Table 7).

To prevent or mitigate degradation, land management practices such as terracing or forest management have been applied in the area. Forest management measures are several, covering 48% of the forest area; terraces as a traditional management measure cover 52% of the cultivated area (Table 8). Additional measures include protection against natural disasters along the river such as gabion walls with vegetative measures (1%), grazing land management (0.6%), and conservation of natural biodiversity or nutrient management (0.7%). On 47% of the area no sustainable land management practices are applied. Below, we present land degradation and land management practices in more detail. We focus on forest and cultivated areas, as they represent 96.7% of the area; but we also present the riverbank and surroundings because they host a large population and the main road access to the watershed. Forest areas Degradation. Invasive species mainly affect cultivated land but are also seen in the forest, albeit to a lesser degree (Fig. 6 and Table 7). Two invasive species mainly affect the area: banmara, present at all altitudes; and nilo gandhe (Ageratum houstonianum), below 1700 m.a.s.l. While nilo gandhe is rarely seen in the forest, banmara grows more easily on the edge of the forest, near the river, along roadsides, and on abandoned land. Another type of degradation affecting 6% of the watershed is the reduction of vegetation cover (Table 7). Although the forest area is increasing (Fig. 4), some zones suffer from decreasing vegetation cover due to grazing, mismanagement, and population pressure. Open grazing inside the forest still exists but has been highly

164

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

Table 6 Current conversion of abandoned land. Conversion of abandoned land to …

Area (hectares)

%

Shrubland Grazing land Grassland Total abandoned land

137.3 27.7 39.1 204.1

67.3 13.6 19.2 100

reduced by strict Community Forest (CF)3 grazing management rules. One specific forest area in the centre of the watershed, west of Ghatichina, suffers from reduction of vegetation cover amid a conflict between two communities in the implementation of a new CF. In the CF Shanthi Salgari, (Fig. 6), population growth is to blame. In the village of Thulakhet, the population is increasing due to better access to roads and services such as health care, education, job opportunities, and proximity to Pokhara. Though managed under a CF, the share of forest per household is diminishing every year and pressure on resources is increasing. Forest management measures. Two types of forest management exist in the area: CF and Panchase Protected Forest (PPF). CF is the most popular and proved an effective way to manage forest in Nepal. The rules regarding grazing, cut-and-carry for fodder, and grass and timber cutting, may vary from one Community Forest User Group (CFUG) to another. PPF4 is aimed at protecting the rich biodiversity of this area, especially the orchids. As it was put in place only recently (2011), it is difficult to estimate its level of effectiveness. Fig. 7 shows the effectiveness of those measures independent of their extent. Nearly half (48%) of the forest area within the watershed is managed either by communities or the PPF programme. CFs have been in place for two decades in some locations in this subwatershed, with visible positive effects. Fewer people are moving to Tamagi CF, which has improved the forest quality and led to better forest management. The Harpan CF is also interesting: after a large landslide in 1934, forest was planted to stabilize the soil, and then turned into a CF. The area in the south used to be seasonal meadows, but the decrease in cattle coupled with better land management have shown very positive effects.

increase is slower than within the cultivated lands. In some villages (Damdame and Khorpakha) of the watershed the rate is higher. These areas are affected by a high level of migration and thus land abandonment. However, the cultivated lands close to the river in the eastern part of the watershed are less affected. This is due to demographic dynamics: the lower area is densely populated with a growing population, leading to a higher intensity of cultivated land use/land cover. In turn, this means land maintenance is better, with regular weeding, terrace wall maintenance, and irrigation. This has restricted the spread of invasive species in these areas. Reasons for the increase in invasive species are multiple but still unclear, according to most of the local land users and land experts. Two causes are most frequently cited: for one, land abandonment and insufficient or no land management (e.g. terrace maintenance with infrequent weeding of the risers); for another, use of agrochemicals (less livestock means less manure). Bush encroachment is also a severe issue in the area due to the spread of banmara. To a lesser extent, the expansion of settlement and change in seasonal rainfall have also been cited as causes of increased spread of invasive species. Negative consequences of the spread of invasive species are numerous. They lead to a scarcity of grass in some areas because the invasive species need less fertilizer and soil moisture. They hinder the growth of some trees and bushes, leading to loss of biodiversity. Loss of soil fertility has also been noticed where invasive species are

Cultivated land Land degradation. As for the forest areas, the main degradation occurring on cultivated land is related to the quality and composition of species (Table 7). Local experts say nilo gandhe appeared on the lower fields five years ago and banmara more than ten years ago. Both species are spreading gradually to upper slopes due to increasing temperatures and land abandonment. They can only be removed by frequent weeding of terraces, picking, and burning. Even though banmara can be used as manure, preparing it is very work intensive and thus hindered by the labour shortage in this area. Fig. 8 represents the trend of biological degradation over a recent period of time (about ten years): in most of the area, the rate is slowly to moderately increasing. Within the forest area, the

3 Under the Community Forest (CF) system, the government hands over the particular area of forest to its users (usually to the traditional users who have been using it for generations e often the issue of demarcation becomes an issue of conflict among two neighbouring communities). A CF User Group is formed, makes rules according to the community forestry guidelines, and manages the forest accordingly. 4 The rules of the Panchase Protected Forest (PPF) are stricter than for CF. In particular, it has stricter rules regarding the use of forest for timber and fuelwood. The protected forest is managed by the PPF Council under the Department of Forest. The CFs are managed by the communities that have been given the right of use.

Fig. 5. A) Landscape with a succession of cultivated (rice) terraces and uncultivated terraces (red arrows). B) Abandoned terraces colonized by banmara (Ageratina adenophora). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

165

Table 7 Extent of degradation per land use/land cover and as percentage of total area. Land use/land cover type

Land use/land cover area coverage within the watershed

Bs

Bc

Bp

Cn

Ha

Hg

Hs

Wg

Wm

Wo

Wr

Wt

O

Na

Forest Cultivated land Shrubland Grassland Built-up Riverbed Total

70.4 26.3 0.9 0.8 0.2 1.4 100.0

30.1 67.9 13.0 e e 5.2 39.2

8.7 e e e e e 6.1

e 3.9 e e e e 1.0

e 32.1 e e e e 8.4

e 9.0 e e e e 2.4

e 0.3 e e e e 0.1

e 0.7 e e e e 0.2

e 0.9 e e e e 0.3

0.2 1.4 e e e e 0.5

e 0.3 e e 5.0 e 0.1

e 0.6 e e e 59.6 1.0

0.8 1.9 2.9 e 0.0 0.0 1.0

61.9 23.7 13.8 e 95.0 34.5 50.6

1.5 2.6 70.3 100.0 0.0 0.8 3.2

Note: Bs: Quality and species composition/diversity decline, Bc: Reduction of vegetation cover, Bp: Increase of pests/diseases, Cn: Fertility decline and reduced organic matter content, Ha: Aridification, Hg: Change in groundwater/aquifer level, Hs: Change in quantity of surface water, O: No degradation, Wg: Gully erosion/gullying, Wm: Mass movements/landslides, Wo: Offsite degradation effects, Wr: Riverbank erosion, Wt: Loss of topsoil/surface erosion, Na: missing data.

Table 8 Percentage of land management practices per land use/land cover type. Land use/land cover type

Land use/land cover area coverage within the watershed

Forest protection

Grazing land management

Others

Protection against natural hazards

Terraces

No conservation measures

Missing data

Forest Cultivated land Shrubland Grassland Built-up Riverbed Total

70.4 26.3 0.9 0.8 0.1 1.4 100.0

47.5 0.9 e e e e 33.7

0.1 1.4 14.3 e e e 0.6

e 2.5 e e e e 0.7

0.2 2.6 e e e 17.7 1

e 52.4 e e e e 13.8

50.7 37.6 15.4 0 100.0 81.5 47

1.5 2.6 70.3 100.0 e 0.8 3.2

Fig. 6. Degree of biological degradation weighted by extent per land use/land cover unit, Harpan subwatershed, Nepal.

166

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

Fig. 7. Effectiveness of management measures weighted by extent per land use/land cover unit, Harpan subwatershed, Nepal.

abundant. Another very important issue related to nilo gandhe specifically is its toxicity to animals. The decline in soil fertility is apparent on 32% of the cultivated areas (Table 7, Fig. 9), but follows no specific locational pattern. The upstream areas in the centre and the northern part of the watershed have a moderate degree of fertility decline. These areas are affected by land abandonment (and thus invasive species and livestock decrease). Livestock is an important source of manure on fields in an area where very little chemical fertilizer is used. This is why fertility decline is much lighter in the eastern, downstream area next to the river, where sufficient manure and fertilizer is applied. Nevertheless, there is still some fertility decline, as a population increase here is responsible for an intensive cropping pattern (two to three crops per year), leading to excessive removal of biomass. Management measures in the cultivated areas. Terraces are a traditional way of managing land in the Nepali hills. Stones are sometimes used to reinforce the terrace risers, but the effectiveness of terrace management is not the same everywhere. Terracing is moderately to highly effective in the lower part of the watershed, but only lightly effective in the western, mostly steep upper part e mainly because the labour shortage means less maintenance. Mice and porcupines affect terraces everywhere, making holes and eating the harvest. By contrast, livestock trampling on uncultivated terraces appears not to be a big issue. Though some terraces show degradation due to cattle trampling, it is not the case everywhere and is hardly mentioned by local land users.

River and surroundings Land degradation. Riverbank erosion is mainly located along the Harpan River, and the degree of degradation is especially high in the eastern part of the watershed at the confluence of the Harpan River and the Andheri River. (Table 7 and Fig. 1). The degree of degradation is lighter upstream and moderate on both sides of the river. There appears to be no riverbank erosion along other streams of the watershed. The causes mentioned by local experts are mainly linked to rainfall patterns with perceived changes in seasonal rainfall confirmed by Duncan, Biggs, Dash, and Atkinson (2013). Seasonal flooding is also mentioned as a cause of riverbank erosion, as is the extraction of stones and gravel for construction of buildings. The population increase has led to an increase in cultivation and grazing at the valley bottom (e.g. Ghatichina village), making it more susceptible to economic losses and infrastructure damage in case of natural events. For instance, in 2007 heavy rainfall triggered floods and landslides, destroying land in and around Ghatichina. Protection against natural disasters. Natural disaster protection measures mainly include gabion walls, sometimes combined with vegetative measures such as planting bamboo (Dandrocalamus spp.) or broom grass (Thysanolaena maxima). Gabion walls are widely used to protect fields from erosion, and roads from landslides; they were erected along the Harpan River in the 1970s and 1990s to protect the riverbank from erosion. After the 1990s, the District Soil and Water Conservation Office planted bamboo and other vegetation to help stabilize these walls, and which could also be used for livestock fodder and domestic purposes. However, their effectiveness is considered to be light.

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

167

Fig. 8. Rate of biological degradation over the last 10 years weighted by extent per land use/land cover unit, Harpan subwatershed, Nepal.

In other areas, gabion walls are few and quite new, and it is difficult for local experts to assess their effectiveness. In addition, they may not have reached their full effectiveness if combined with still-growing vegetation such as bamboo or broom grass. But our degradation assessment shows only a low presence of landslides/ gully degradation in the watershed (Table 7). Discussion A recent report based on the Nepal Living Standard Survey showed that most of those absent are men (73%) between 15 and 44 years old (Sharma et al., 2014), which confirms our household survey's results. In areas with high levels of subsistence agriculture, the absence of males has a large impact on every aspect of the households' livelihoods. According to recent studies, outmigration in Nepal has a negative impact on agriculture, as it leads to a labour shortage and increases the work burden on women (Gartaula, Visser, & Niehof, 2012; K. P. Paudel, Dahal, & Shah, 2012). Our case study supports this finding, demonstrating in addition that this labour shortage leads to a change in agricultural practice and land use, often resulting in insufficient land management and land abandonment. In a context of labour shortage, those left behind have to make choices on how to manage the household with fewer people and with off-farm income, mainly remittances (Maharjan, Bauer, & Knerr, 2012). The decision to abandon land may occur as a result of the labour shortage, distance to fields, production level,

water scarcity, and, sometimes, the consequences of natural disasters. Surprisingly, abandoned land does not mean that land is available to use or buy. Instead of selling their land, locals prefer to leave it fallow or to practice a rental system called adhiya (the owner rents the land and receives 50% of the production). In many cases and especially within one of the ethnic groups living in the area, migrants expect to return back because they are hopeful the area will be developed in the future or that their land will increase in value due to its proximity to Pokhara. Some also cited symbolic reasons (e.g. ancestral property) for not selling. Significantly, our study did not confirm the findings of previous studies which have linked population dynamics such as a population decline to radical effects of land use/land cover changes (Davis & Lopez-Carr, 2014; Gartaula et al., 2012; Li & Tonts, 2014; McCarthy, Carletto, Davis, & Maltsoglou, 2006). According to these studies, such changes increase the risk and occurrence of forest fires, intense erosion, gullying, and landslides (Harden, 1996; Koulouri & Giourga, 2007; Munroe, van Berkel, Verburg, & Olson, 2013). In areas with steep slopes (like those found in the Harpan subwatershed), erosion such as landslides and gullying is expected (Petley et al., 2007). In addition, the labour shortage and ensuing lower maintenance are expected to render terraces less effective in preventing slope erosion, leading to potentially hazardous events in some areas (Gardner & Gerrard, 2003; Khanal & Watanabe, 2006; Tarolli, Preti, & Romano, 2014). Our study found that large

168

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

Fig. 9. Degree of soil fertility decline weighted by extent per land use/land cover unit, Harpan subwatershed, Nepal.

areas of terraced land had been abandoned and become forest. But in contrast to the above risk expectations, we found that lower population pressure and better forest management measures (CF and PPF) fostered forest growth, as less fodder and fuelwood was collected. Extensive vegetation cover within the watershed stabilized the slopes and decreased natural erosion amid favourable ecosystem recovery conditions (favourable climate, effective forest management measures). However, monitoring is important: in other areas of Nepal with a longer history of land abandonment, collapse of terraces has led to significant damage due to erosion and gullying (Khanal & Watanabe, 2006). Whether the positive effects observed in our study area will decrease risks of flooding in downstream areas requires further investigation. The downstream villages of our study watershed have shown significant riverbank erosion: there is growing housing and infrastructure construction, and agricultural fields are being created in former riverbeds. While the latter increases vulnerability, the area may benefit from the stabilized slopes upstream. Soil fertility decline is also often cited in the literature as a general problem for agriculture in the hills (Desbiez, Matthews, Tripathi, & Ellis-Jones, 2004; Gartaula et al., 2012; Gautam, Webb, Shivakoti, & Zoebisch, 2003; G. S. Paudel & Thapa, 2001; Regmi & Zoebisch, 2004; G. B. Thapa, 1996). We (more precisely, the expert groups) found a decline in soil fertility both upstream and downstream e but for different reasons, both related to population dynamics. Our results very clearly showed more land abandonment upstream and more intensive land use downstream, with consequences not only for soil fertility but also for forest use and growth of invasive species. While the villages located uphill are becoming

smaller due to high levels of migration, villages located in the valley bottom such as Ghatichina and Thulakhet have an increasing population. The soils of the less populated villages on the slopes or uphill are less fertile, as there is less livestock and thus less manure. The valley bottom villages of Ghatichina and Thulakhet are also experiencing fertility decline, but this is due to excessive removal of biomass e a consequence of having increased cropping cycles from two to three per year. Experts and land users in our study area consider invasive species to be the biggest challenge. Invasive species have taken hold in the watershed over the past decade, hindering growth of natural vegetation. Though the problem is recognized and studied (IUCN., 2013), effective solutions are not yet clear, hindered by the apparent lack of knowledge about invasive species and lack of support from agricultural extension services. Climate change is pushing some species higher to areas used for grassland and grazing land; nilo ghande poses a particular problem as it is toxic for animals. The issue of invasive species hampers well-working forest management measures such as community forestry, known for its efficiency (Adhikari, Di Falco, & Lovett, 2004). Nevertheless, the negative impact of invasive species is considerably smaller than in unmanaged forest, proving that effective management is possible and advisable. Conclusion In the 1970s, the discourse on degradation in the Phewa watershed focused on resource overuse and the impact this had on land. Today, the uphill areas are underused. The Harpan

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

subwatershed is very dynamic in terms of population and land use/land cover changes: people, mostly men, migrate for labour work abroad or to Nepal's urban areas, leaving behind women and the elderly; in addition, families migrate downstream to be closer to the main facilities. This subwatershed shows an increasingly diverging development path between uphill areas and valleys, as is typical for many mountain regions worldwide. Uphill, a marked decrease in population has consequences such as land abandonment, spread of invasive species, decline in soil fertility, and lack of terrace management. Downstream, population increase has led to more intensive land use, reduction of vegetation cover, soil fertility decline, and encroaching on land close to the riverbeds (riverbank erosion). The degradation identified in the mapping can sometimes appear very similar but has different causes. Differentiating these causes is possible using the WOCAT methodology, which allows us not only to represent the degradation but also to understand its causes and impacts. The environment of this watershed and its agricultural practices are very well known. However, effective management measures are still lacking. In addition, what measures there are insufficiently reflect the dynamics of the region, as they fail to take into account the changes in population (i.e. absence of men) and increase in forest cover. Degradation is less and different than it used to be, and the overall watershed seems to be in better shape than others in the area. However, better land management is still urgently needed, especially as new problems such as growth of invasive species are emerging. This is important in order to support those left behind in the outmigration context, and to leave the door open for future return migration. Significantly, migrants, with their links to their places of origin, could have huge potential to substantially contribute to sustainable development of the area, with the help of the knowledge, financial means, and skills gained in the migration process. Acknowledgements This 2-year study (2012e2014) was financed by the Swiss Network for International Studies, based in Geneva. We would like to send sincere and deep thanks to the communities and their people who helped us during the fieldwork and who always welcomed us with a smile, especially to the communities of Ghatichina, Thulakhet, Lower Shidane, Makwanpur, and ChisopaniPhilinghari. Many thanks also to our partners IUCN Nepal and NCCR and to our field assistants Pratima Baral and Gyanendra Subedi. A special thanks to the Institute of Forestry in Pokhara and its director, Prof. Dr. Chiranjibi Upadhyaya, who welcomed the final workshop for the WOCAT mapping and for providing help and advice, and to Sanjaya Devkota his valuable assistance in the logistical organization. Thank you to Tina Hirschbuehl and Marlene Thibault for language editing this article. References Adhikari, B., Di Falco, S., & Lovett, J. C. (2004). Household characteristics and forest dependency: evidence from common property forest management in Nepal. Ecological Economics, 48(2), 245e257. http://dx.doi.org/10.1016/j.ecolecon.2003. 08.008. Aide, T. M., & Grau, H. R. (2004). Globalization, migration, and Latin American ecosystems. Science, 305(5692), 1915e1916. http://dx.doi.org/10.1126/ science.1103179. Awasthi, K. D., Sitaula, B. K., Singh, B. R., & Bajacharaya, R. M. (2002). Land-use change in two Nepalese watersheds: GIS and geomorphometric analysis. Land Degradation & Development, 13(6), 495e513. http://dx.doi.org/10.1002/ ldr.538. Blaikie, P., Cameron, J., & Seddon, D. (2002). Understanding 20 years of change in West-Central Nepal: continuity and change in Lives and ideas. World Development, 30(7), 15.

169

Dahal, R. K., & Hasegawa, S. (2008). Representative rainfall thresholds for landslides in the Nepal Himalaya. Geomorphology, 100(3e4), 429e443. http://dx.doi.org/ 10.1016/j.geomorph.2008.01.014. Davis, J., & Lopez-Carr, D. (2014). Migration, remittances and smallholder decisionmaking: Implications for land use and livelihood change in Central America. Land Use Policy, 36(0), 319e329. http://dx.doi.org/10.1016/j.landusepol.2013.09. 001. Desbiez, A., Matthews, R., Tripathi, B., & Ellis-Jones, J. (2004). Perceptions and assessment of soil fertility by farmers in the mid-hills of Nepal. Agriculture, Ecosystems & Environment, 103(1), 191e206. http://dx.doi.org/10.1016/j.agee. 2003.10.003. DFID.. (1999). Sustainable livelihoods guidance sheet. London: UK Department for International Development. Duncan, J. M. A., Biggs, E. M., Dash, J., & Atkinson, P. M. (2013). Spatio-temporal trends in precipitation and their implications for water resources management in climate-sensitive Nepal. Applied Geography, 43(0), 138e146. http://dx.doi.org/ 10.1016/j.apgeog.2013.06.011. Eckholm, E. P. (1975). The deterioration of mountain environments. Science, 189. Fleming, W. M. (1985). Basin management in Nepal: Translating technical information into resource policy alternatives. Paper presented at the Jerusalem Symposium. Fleming, B., & Fleming, J. P. (2009). A watershed conservation success story in Nepal: land use changes over 30 years. Waterlines, 28(1), 29e46. http:// dx.doi.org/10.3362/1756-3488.2009.004. Gardner, R. A. M., & Gerrard, A. J. (2003). Runoff and soil erosion on cultivated rainfed terraces in the Middle Hills of Nepal. Applied Geography, 23(1), 23e45. http://dx.doi.org/10.1016/S0143-6228(02)00069-3. Gartaula, H. N., Visser, L., & Niehof, A. (2012). Shifting perceptions of food security and land in the context of labour out-migration in rural Nepal. Food Security, 4(2), 181e194. http://dx.doi.org/10.1007/s12571-012-0190-3. Gautam, A. P., Webb, E. L., Shivakoti, G. P., & Zoebisch, M. A. (2003). Land use dynamics and landscape change pattern in a mountain watershed in Nepal. Agriculture, Ecosystems & Environment, 99(1e3), 83e96. http://dx.doi.org/10. 1016/S0167-8809(03)00148-8. Gerrard, J., & Gardner, R. (2002). Relationships between landsliding and land use in the Likhu Khola Drainage Basin, middle Hills, Nepal. Mountain Research and Development, 22(1), 48e55. http://dx.doi.org/10.1659/0276-4741(2002)022 [0048:rblalu]2.0.co;2. Ghimire, A., & Upreti, B. R. (2012). Migration, mobility and Development: Enhancing our understanding in Nepal, SAs RCO discussion paper 6. Kathmandu: South Asia Regional Coordination Office of National Centre of Competence in Research North-South and Nepal Centre for Contemporary Research. Gisbert, M. E., Painter, M., & Quiton, M. (1994). Gender issues Associated with labor migration and dependence on off-farm income in rural Bolivia. Human Organization, 53(2), 110e122. GoN. (2011). Population and housing census. Kathmandu. Graner, E., & Gurung, G. (2003). Arab ko lahure: looking at Nepali labour migrants to Arabian countries. Contributions to Nepalese Studies, 30(2), 30. Harden, C. P. (1996). Interrelationships between land abandonment and land degradation: a case from the Ecuadorian Andes. Mountain Research and Development, 16(3), 274e280. http://dx.doi.org/10.2307/3673950. IUCN.. (2013). Impact assessment of invasive plant species in selected ecosystems of Bhadaure tamagi VDC, Kaski. Kathmandu. Ives, J. D. (2004). Himalayan Perceptions: Environmental change and the well-being of mountain peoples. from http://www.myilibrary.com?id¼5088. Ives, J. D., & Messerli, B. (1989). The Himalayan Dilemma Reconciling development and conservation. London and New York: ROUTLEDGE London and New York. Khanal, R., & Watanabe, T. (2006). Abandonment of agricultural land and its consequences. Mountain Research and Development, 26(1), 32e40. Kollmair, M., Manandhar, S., Subedi, B., & Thieme, S. (2006). New figures for old stories: migration and remittances in Nepal. Migration Letters, 3(2), 9. Koulouri, M., & Giourga, C. (2007). Land abandonment and slope gradient as key factors of soil erosion in Mediterranean terraced lands. CATENA, 69(3), 274e281. http://dx.doi.org/10.1016/j.catena.2006.07.001. Liniger, H. P., Van Lynden, G., Nachtergaele, F., & Schwilch, G. (2008). Questionnaire for mapping land degradation and sustainable land management (QM). WOCAT, LADA, DESIRE (pp. 45). CDE Bern, FAO Rome, ISRIC Wageningen. Li, L., & Tonts, M. (2014). The impacts of temporary labour migration on farming systems of the Loess Plateau, Gansu Province, China. Population, Space and Place, 20(4), 316e332. http://dx.doi.org/10.1002/psp.1832. Maharjan, A., Bauer, S., & Knerr, B. (2012). International migration, remittances and subsistence farming: evidence from Nepal. International Migration, 51, 249e263. McCarthy, N., Carletto, G., Davis, B., & Maltsoglou, I. (2006). Assessing the impact of massive out-migration on agriculture (December 2006 ed.). FAO. Munroe, D. K., van Berkel, D. B., Verburg, P. H., & Olson, J. L. (2013). Alternative trajectories of land abandonment: causes, consequences and research challenges. Current Opinion in Environmental Sustainability, 5(5), 471e476. http://dx. doi.org/10.1016/j.cosust.2013.06.010. Paudel, K. P., Dahal, D., & Shah, R. (2012). Study report on abandoned agriculture land in mid-hills of Nepal, Status, causes and consequences. Nepal: Australian Centre for International Agriculture Research, World Agroforestry Centre, Forest Action, IUCN. Paudel, G. S., & Thapa, G. B. (2001). Changing farmers' land management practices in the Hills of Nepal. Environmental Management, 28(6), 789e803. http:// dx.doi.org/10.1007/s002670010262.

170

S. Jaquet et al. / Applied Geography 62 (2015) 157e170

Petley, D. N., Hearn, G. J., Hart, A., Rosser, N. J., Dunning, S. A., Oven, K., et al. (2007). Trends in landslide occurrence in Nepal. Natural Hazards, 43, 23e44. Radel, C., & Schmook, B. (2008). Male transnational migration and its Linkages to land-use change in a Southern Campeche Ejido. Journal of Latin American Geography, 7(2), 59e84. http://dx.doi.org/10.2307/25765219. Regmi, B. D., & Zoebisch, M. A. (2004). Soil fertility status of Bari and Khet land in a small watershed of Middle Hill Region of Nepal. Nepal Agriculture Research, 5, 7. Schwilch, G., Bestelmeyer, B., Bunning, S., Critchley, W., Herrick, J., Kellner, K., et al. (2011). Experiences in monitoring and assessment of sustainable land management. Land Degradation & Development, 22(2), 214e225. Seddon, D., Gurung, G., & Adhikari, J. (1998). Foreign labour migration and the remittance economy of Nepal. Himalayan Research Bulletin, 18(2). Sharma, S., Pandey, S., Pathak, D., & Sijapati-Basnett, B. (2014). State of migration in Nepal. Kathmandu: Centre for the Study of Labour and Mobility. Sudmeier-Rieux, K., Jaquet, S., Derron, M.-H., Jaboyedoff, M., & Devkota, S. (2012). A case study of coping strategies and landslides in two villages of Central-

Eastern Nepal. Applied Geography, 32(2), 680e690. http://dx.doi.org/10.1016/j. apgeog.2011.07.005. Tarolli, P., Preti, F., & Romano, N. (2014). Terraced landscapes: from an old best practice to a potential hazard for soil degradation due to land abandonment. Anthropocene, (0). http://dx.doi.org/10.1016/j.ancene.2014.03.002. Thapa, G. B. (1996). Land use, land management and environment in a subsistence mountain economy in Nepal. Agriculture, Ecosystems & Environment, 57(1), 57e71. http://dx.doi.org/10.1016/0167-8809(96)88021-2. Thapa, G., & Weber, K. (1995). Status and management of watersheds in the Upper Pokhara Valley, Nepal. Environmental Management, 19(4), 497e513. http:// dx.doi.org/10.1007/bf02471963. The World Bank. (1979). Nepal: Development performances and prospects (pp. 152). Washington D.C.: The World Bank. Warren, A. (2002). Land degradation is contextual. Land Degradation & Development, 13(6), 449e459.