The Government of Nepal has identified opportunities in agricultural commercialization, responding to a growing internal demand and expansion of export markets to reduce the immense trade deficit. Several cash crops, including coffee and bananas, have been identified in the recently approved Agriculture Development Strategy. Both of these crops have encouraged smallholder farmers to convert their subsistence farming practices to more commercial cultivation. Identification of suitable agro-ecological zones and understanding climate-related issues are important for improved production and livelihoods of smallholder farmers. Here, the suitability of coffee and banana crops is analyzed for different agro-ecological zones represented by Global Environmental Stratification (GEnS). Future shifts in these suitability zones are also predicted. Plantation sites in Nepal were geo-referenced and used as input in species distribution modelling. The multi-model ensemble model suggests that climate change will reduce the suitable growing area for coffee by about 72% across the selected emission scenarios from now to 2050. Impacts are low for banana growing, with a reduction in suitability by about 16% by 2050. Bananas show a lot of potential for playing an important role in Nepal as a sustainable crop in the context of climate change, as this study indicates that the amount of area suited to banana growing will grow by 40% by 2050. Based on our analysis we recommend possible new locations for coffee plantations and one method for mitigating climate change-related problems on existing plantations. These findings are expected to support planning and policy dialogue for mitigation and support better informed and scientifically based decision-making relating to these two crops.

Worldwide, water availability will be a key issue in the 21st century. Per capita water availability is projected to fall from 6600 to 4800 m 3 between 2000 and 2025 because of uneven distribution of water resources. However, most of the world's population will have below 1700 m 3 per capita (Cosgrove and Rijsberman 2000). Mountains, in particular, are of great importance because more than half of humanity relies on freshwater that originates from mountains (Liniger et al 1998; Viviroli 2001). The rising demand for water from growing populations, climate change affecting water supplies , and increased uncertainty in relation to natural hazards (eg, increased occurrence of landslides, glacial lake outburst floods) are some of the reasons for future concern (Kundzewicz et al 2001). In the context of the Hindu Kush-Himalayan (HKH) region with its pronounced monsoonal climate, the following dual concern can be demonstrated by the current status of water-related issues:

No abstract

The global population has tripled in the last 100 years, whereas water use by humans has increased 6-fold. Major water shortages seem certain if these trends continue. With the aim of avoiding the risk of water deficiency, the World Water Council's World Water Vision report proposed a strategy that includes efforts to increase water availability and improve storage, reform water resources management institutions, appraise ecosystem functions, and increase transboundary cooperation. To address these issues and the concerns of the residents of rural watersheds in the Middle Mountains of Nepal (see the paper by Merz et al in the Research section of this issue), new ways of managing available water resources must be adopted. A study conducted in 2 watersheds in Nepal-the Jhikhu Khola (JKW) and the Yarsha Khola (YKW)-by the People and Resource Dynamics in Mountain Watersheds of the Hindu Kush-Himalayas Project (PARDYP), operated by the International Centre for Integrated Mountain Development (ICIMOD), revealed that: 14

The semi-distributed, conceptual hydrological model HBV was applied to Tamor Nadi in order to estimate runoff at Tapethok, Taplejung, in Eastern Nepal. As there was no discharge data available for this particular location, the model was first calibrated and validated for the bigger, gauged basins at Mulghat and Majithar. However due to its structure HBV shows difficulties in modelling low and high flows correctly at the same time. Therefore two parameter sets were produced: one with focus on the model performance during low flows and the second one, on high flows. Those parameters were then applied to the basin at Tapethok. Generally HBV was able to correctly simulate low flows except for some sharp peaks due to isolated precipitation events. However, pre-monsoon discharge was overestimated while the runoff of the monsoon season were most of the time underestimated. The main reasons for this situation are: (1) HBV generates runoff from one single groundwater reservoir for the entire catchment, leading to sharp peaks with a rapid recession and therefore exaggerated reactions on precipitation during dry season; (2) during pre-monsoon snow and ice melt gain in importance and add to the mentioned problem; (3) due to the simplified representation of storages in the model structure the catchment area drains too quickly.