Discussion papers

1. Land use impacts on water resources: a literature review - Benjamin Kiersch

Benjamin Kiersch, Land and Water Development Division
Food and Agriculture Organization
Rome, Italy

Introduction

Land use practices are assumed to have important impacts on both the availability and quality of water resources. These impacts can be both positive and negative. It is intuitively appealing that the benefits of improved land management, or the costs associated with negative impacts of inadequate land use on the water resources, might not only be felt by water users who cause them, but also by others who live downstream or - in the case of groundwater - make use of the affected groundwater resources. In order to assess these costs and benefits, it is important to get a clear picture, from a landscape perspective, of the extent that different land use practices affect hydrologic regime and water quality and at which watershed scale the impacts are of importance.

The present paper proposes a typology of land use impacts on water resources, and attempts to evaluate the importance of each impact type in relation to spatial scale on the basis of a literature review.

Impacts of land use on water resources

In order to establish linkages between upstream land and downstream water users, it is important to have a clear picture of the possible impacts of land uses on both hydrologic regime (water availability) and water quality, and the scales at which these impacts are relevant. In the following sections, an attempt is made to categorise land use impacts on water resources, to analyse the main determining factors behind the impact, and to provide some examples from the literature.

The review focuses on impacts from agricultural land use, as well as from grazing, forestry and fisheries, as these all fall under FAO’s mandate. Other land uses, like mining and quarrying, urbanization and industrialization, which also have important impacts on the hydrologic regime, are not included in this review. Furthermore, the review focuses on the physical impacts on water resources. Impacts on living aquatic resources, e.g. on fish and other aquatic organisms, aquatic ecosystems and wetlands, are not discussed explicitly. It is an open question, however, whether and how these should be included in this typology.

It is difficult to arrive at universally valid statements about land use impacts on water resources for several reasons. The impacts of land use on water resources depend on a host of natural and socio-economic factors. Natural factors include climate, topography and soil structure. Socio-economic factors include economic ability and awareness of the farmers, management practices, and the development of infrastructure, e.g. roads. Furthermore, the impacts of agricultural land use may be difficult to distinguish from natural or other human impacts, such as the impact of agricultural runoff versus rural sewage systems on degradation of surface water and groundwater.

Land use impacts on hydrologic regime

With regard to the hydrologic regime, impacts on surface water resources and groundwater resources can be distinguished. Impacts of land use practices on surface water can be divided into (i) impacts on the overall water availability or the mean annual runoff, and (ii) impacts on the seasonal distribution of water availability. With regard to the latter, impacts on peak flows and impacts on dry season flows are of importance. With regard to groundwater, the effect of the land use on groundwater recharge has to be examined.

Mean surface runoff

The impact of land use on the mean runoff is a function of many variables, the most important being the water regime of the plant cover in terms of evapotranspiration (ET), the ability of the soil to hold water (infiltration capacity), and the ability of the plant cover to intercept moisture.

A change of land cover from lower to higher ET will lead to a decrease in annual stream flow. From a review of 94 catchment experiments, Bosch and Hewlett (1982) concluded that the establishment of forest cover on sparsely vegetated land decreases water yield. Coniferous forest, deciduous hardwood, brush and grass cover have (in that order) a decreasing influence on water yield of the source areas in which the covers are manipulated.

Conversely, a change from higher-ET plants to lower-ET plants will increase the mean surface runoff: reduction in forest cover increases water yield (Bosch and Hewlett, 1982; Calder, 1992). The impact, however, depends very much of the management practices and the alternative land uses. Careful, selective harvesting of timber has no or little effect on stream flow. Stream flow after maturation of the new plant cover may be higher, the same or lower than original value, depending on vegetation (Bruijnzeel, 1990).

An exception to this rule are cloud forests, which can intercept more moisture (fog drip) than consumption by ET (Bosch and Hewlett, 1982), and very old forests, which, depending on the species, may consume less water than the vegetation that establishes itself after clear-cutting (Calder, 1998).

Stream flow gains decline over time with establishment of new plant cover, but time scales can vary greatly. In humid warm areas, the effect of clear-cutting is shorter lived than in less humid areas, due to rapid regrowth of vegetation (Falkenmark and Chapman, 1989).

Increasing water yield from changing plant cover does not necessarily increase water availability downstream. Stream flow might decrease because of other factors, e.g. water consumption by riparian vegetation or through transmission losses (channel infiltration) (Brooks et al., 1991).

Peak flow/floods

Peak flows can increase as a result of a change in land use if the infiltration capacity of the soil is reduced, for example through soil compaction or erosion, or if drainage capacity is increased. Peak flow may increase after trees are cut down (Bruijnzeel, 1990). Relative increases in storm flow after tree removal are smallest for large events and largest for small events. As the amount of precipitation increases, influence on storm flow of soil and plant cover diminishes (Bruijnzeel, 1990; Brooks et al., 1989).

An increase of peak flows may also result from the building of roads and infrastructure. Studies in the north-western USA have shown that the construction of forest roads can intensify peak runoff from forested areas significantly (La Marche and Lettenmaier, 1998; Bowling and Lettenmaier, 1997). Consolidation of smaller plots to large fields can lead to higher runoff rates, due to drainage systems and asphalt access roads (Falkenmark and Chapin, 1989). Conversely, peak flows may decrease as a result of an increased soil infiltration capacity.

In larger basins, effects of land use practices on peak flow are offset due to time lag between different tributaries, different land use and variations in rainfall (Bruijnzeel, 1990) In larger watersheds, this de-synchronisation effect can lead to a reduction in peak discharge, although overall storm flow increases due to land use changes in individual subwatersheds (Brooks et al., 1991).

Base flow/dry season flow

The effect of land use change on dry-season flow depends on competing processes, most notably changes in ET and infiltration capacity. The net impact is likely to be highly site specific (Calder, 1998).

In tropical areas, afforestation can lead to decreased dry-season flows due to increased evapotranspiration. In the Mae Thang watershed (Thailand), afforestation programmes led to water shortages downstream, which resulted in a seasonal closure of a water treatment plant and lower availability for irrigation (Chomitz and Kumari, 1996). Similarly, in the Fiji Islands, large-scale pine afforestation (60 000 ha) in watersheds previously covered by grassland led to reductions in dry-season flow of 50-60 percent, putting the operation of a hydro-electric plant and drinking water supply at risk (FAO, 1987).

Most experimental evidence in rainfall-dominated regimes suggests that forest removal (or change from high-water-use plants to low-water-use plants) increases dry season flows (Brooks et al., 1991). In contrast, dry-season flows from deforested land may decrease if the soil infiltration capacity is reduced, e.g. through use of heavy machinery (Bruijnzeel, 1990). Low flow resulting from extended dry periods or droughts may not be substantially altered by changes in vegetative cover (Brooks et al., 1991).

Groundwater recharge

The groundwater recharge may be increased or decreased as a result of changing land use practices. The major driving forces are the ET of the vegetative cover and the infiltration capacity of the soil. Groundwater recharge is often linked with dry-season flows, as groundwater contributes much of the river discharge during the dry season.

The water table may rise as a result of decreased evapotranspiration, e.g. following logging or conversion of forest to grassland for grazing. Recharge may also increase due to an increased infiltration rate, e.g. through afforestation of degraded areas (Tejwani, 1993).

In contrast, the water table may fall as a result of decreased soil infiltration, e.g. through non-conservation farming techniques and compaction (Tejwani, 1993). Also, heavy grazing may lead to reduced infiltration and groundwater recharge (Chomitz and Kumari, 1996). If the infiltration capacity is substantially reduced, this can lead to water shortages in dry seasons, even in regions where water is usually abundant, like in the case of shifting cultivation in Cherapunji province, India (FAO, 1999). Likewise, groundwater recharge can be reduced as a result of planting of deep rooting tree species, e.g. eucalyptus (Calder, 1998).

Impacts of land use on water quality

Land use practices can have important impacts on water quality, which in turn may have negative or, in some cases, positive effects on downstream uses of water. Impacts include changes in sediment load and concentrations of nutrients, salts, metals and agrochemicals, the influx of pathogens, and a change in the temperature regime.

Erosion and sediment load

Forests are checkers of soil erosion. Protection is largely because of understorey vegetation and litter, and the stabilising effect of the root network. On steep slopes, the net stabilising effect of trees is usually positive. Vegetation cover can prevent the occurrence of shallow landslides (Bruijnzeel, 1990). However, large landslides on steep terrain are not influenced appreciably by vegetation cover. These large slides may contribute the bulk of the sediment, as for example in the middle hills of the Himalayas (Bruijnzeel and Bremmer, 1989).

Afforestation does not necessarily decrease soil erosion. Splash erosion may increase substantially when litter is cleared from the forest floor (Bruijnzeel, 1990). The spectrum for the size of the drops that are formed by the canopy varies widely among different species, resulting in large differences in the potential of splash erosion (Calder, 1998).

Deforestation may increase erosion. In Malaysia, streams from logged areas carry 8-17 times more sediment load than before logging (Falkenmark and Chapman, 1989). The actual soil loss, however, depends largely on the use to which the land is put after the trees have been cleared. Surface erosion from well-kept grassland, moderately grazed forests and soil-conserving agriculture are low to moderate (Bruijnzeel, 1990).

Road construction may be a major cause for erosion during timber harvesting operations. In the USA, forest roads are estimated to account for 90 percent of the erosion caused by logging activities (Brooks et al., 1991; Bruijnzeel, 1990).

Effects of erosion control measures on sediment yield will be most readily felt on-site. There is an inverse relation between basin size and sediment delivery ratio. In basins of several hundred km², improvements may only be noticeable after a considerable time lag (decades), due to storage effects (Bruijnzeel, 1990).

Downstream sediment yields cannot always be ascribed to the changing of upstream land use practices. Human impacts on sediment yield may be substantial in regions with stable geological conditions and low natural erosion rates. In regions with high rainfall rates, steep terrain, and high natural erosion rates, however, the impact of land use may be negligible. In the Phewa Tal watershed in Nepal, for example, only six percent of the total sediment yield has been calculated to stem from surface erosion (Bruijnzeel, 1990).

Sediment can act both as a physical and a chemical pollutant. Physical pollution characteristics of sediment include turbidity (limited penetration of sunlight) and sedimentation (loss of downstream reservoir capacity, destruction of coral reefs, loss of spawning grounds for certain fish). Chemical pollution of sediment includes adsorbed metals and phosphorous, as well as hydrophobic organic chemicals (FAO, 1996).

Nutrients and organic matter

A change in land use can alter the nutrient content of surface and groundwater, most notably nitrogen (N) and phosphorus (P) levels. Deforestation can lead to high nitrate (NO₃) concentrations in water due to decomposition of plant material and a reduced nutrient uptake by the vegetation. Nitrate concentration in runoff in deforested catchments can be 50 times higher than in a forested control catchment over several years (Falkenmark and Chapman, 1989; Brooks et al., 1991).

Agricultural activities can lead to an increased influx of nitrogen into waterbodies as a result of many factors, including fertiliser application, manure from livestock production, sludge from municipal sewage treatment plants, and aeration of the soil. In Europe, agriculture accounts for substantial nitrogen emissions into surface and groundwater. With regard to inorganic N, agriculture accounts for 50 percent in Denmark and 71 percent in the Netherlands (FAO, 1996). High nutrient leaching losses can occur when fertiliser is applied to short-term crops on permeable soils. In Sri Lanka, NO₃ -N concentration in groundwater under intensive chilli and onion cultivation reaches 20-50 mg/L (BGS et al., 1996). Continuous soil cover reduces N leaching; fallow periods and soil disturbance increases leaching (BGS et al., 1996). Ploughing can increase NO₃ concentrations in surface and groundwater, as oxygenation of the soil causes nitrification (Falkenmark and Chapman, 1989). In rice paddies, leaching losses are likely to be small, due to denitrification in the soil and volatile losses (BGS et al., 1996). Application of manure from livestock production and direct runoff can lead to acidification of soils due to the volatilisation of ammonia, which in turn may increase the solubility of metals in the soil (FAO, 1996).

Phosphate (PO₄) leaching into water is inhibited by sorption processes to clay particles (BGS et al., 1996). Livestock production, however, can be a major source of P in waters. Direct runoff from intensive livestock farms can lead to serious degradation of surface and groundwater. In the EU, livestock wastes account for 30 percent of P load in surface waters, other agricultural uses account for 16 percent (FAO, 1996).

Phosphate-laden sediment can form a nutrient pool on the bottom of eutrophic lakes, which can be released into the water under anoxic conditions. This makes it difficult to control eutrophication in the short term through limitation of P inflow. Eutrophication can be mitigated by dredging sediment or oxidising the hypolimnion, but these options are quite costly (FAO, 1996).

The precise role of agriculture in the contamination of ground and surface water is difficult to quantify. In most countries, monitoring is not sufficient to establish the extent of nutrient pollution from agricultural land use. In rural areas, it may be difficult to distinguish between agricultural pollution and pollution by untreated sewage (BGS et al., 1996).

Freshwater aquaculture can add substantial nutrient loading to surface water through waste feed that is not consumed by the fish, and the fish’s faecal production (FAO, 1996).

Pathogens

Land use activities may affect the bacteriological quality of water, which can create health concerns for downstream water users. The concentration of pathogenic bacteria in surface waters may increase as a consequence of riparian grazing activities or waste influx from livestock production.

A reduction of stream flow, for example, as a consequence of upstream diversion for irrigation, may lead to ponding in riverbeds, which in turn may provide breeding grounds for vectors of waterborne diseases, such as malaria. Where low flow leads to saltwater intrusion in estuaries, vectors breeding in brackish water may spread (FAO, 1995).

Pesticides and other persistent organic pollutants

Generally, the application of pesticides poses a danger to surface and groundwater resources, since pesticide compounds are designed to be toxic and persistent. Pesticide leaching into groundwater depends on the chemical’s persistence and mobility, as well as the soil structure. Pesticide metabolites might be as toxic and as mobile as the parent compound (BGS et al., 1996). In humans and animals, pesticides can have both acute and chronic toxic effects. Lipophilic compounds can accumulate in fatty tissue (bio-concentration) and in the food chain (bio-magnification) (FAO, 1996).

Pesticide residues can find their way into water resources through their use in agriculture, forestry and aquaculture. Furthermore, unsafe stockpiling and dumping of old and obsolete pesticides can cause severe ground and surface water contamination (FAO, 1996). Aquaculture can lead to the introduction of biocides, disinfectants and medicines into surface water (FAO, 1996).

The actual impact of pesticide contamination of downstream water resources is often difficult to quantify. Pesticide monitoring is difficult because concentrations are very low, large samples and careful sampling, as well as sophisticated analytical instruments, are required (BGS et al., 1996). Since many pesticides are transported in association with suspended matter, water analyses may yield incomplete results. For some pesticides, the analytical capability may not be accurate enough to determine presence or absence for the protection of human health. Newer pesticides which are soluble and degrade more quickly can only be detected shortly after application; therefore, typical monitoring programmes operated on a monthly or quarterly basis are unlikely to be able to quantify the presence and determine the significance of pesticides in surface waters (FAO, 1996).

Salinity

An increase in salinity of surface and groundwater can have detrimental effects on downstream water uses, for example for irrigation or domestic water supply. The impact of land uses on salinity depends on climatic as well as geological factors.

Irrigation and drainage activities may lead to increased salinity of surface and groundwater as a consequence of evaporation and the leaching of salts from soils. This is of special concern in arid areas, where subsurface drainage water always has higher salt concentrations, an increased hardness and a higher sodium absorption ratio than the supply water (FAO, 1997a). Drainage from irrigated agriculture may also lead to an increased concentration of selenium in ground and surface water (Postel, 1997).

A high application rate of potassium chloride fertiliser can lead to an increased leaching of chloride into groundwater. In Sri Lanka, for example, it has been estimated that in some areas of intensive agriculture, groundwater chloride levels may rise to 400 mg/L by 2010 at current rates of fertiliser application, which by far exceeds the acceptable concentration for drinking water as determined by WHO (250 mg/L) (BGS et al., 1996).

In coastal areas, water abstraction for land use activities may indirectly contribute to the salinization of water resources. Groundwater extraction for irrigation, domestic and industrial purposes can result in the intrusion of seawater into the aquifer, and consequently a salinization of the groundwater resources (FAO, 1997). A decrease in river flow due to upstream abstraction or the building of reservoirs can lead to an inland intrusion of brackish water in the estuarine zone (FAO, 1995).

Heavy metals

Land use practices may directly and indirectly contribute to an increased concentration of heavy metals in water resources. A direct pathway is the application of livestock manure and sludge from sewage treatment plants, which may contain high concentrations of heavy metals. For example, pig manure often contains high concentrations of copper (FAO, 1996).

Indirectly, land use may affect heavy metal concentration in surface and groundwater by increasing the mobility of metals in the soil from anthropic or geological origin. Heavy metals in the soil may be transferred into waterbodies by erosive processes. The acidification of soil, caused by ammonia volatilization from manure application or in animal feedlots, may increase the solubility of heavy metals stored in the soil, and thus the influx into surface and groundwater. High abstraction rates of groundwater for irrigation can alter the chemical environment in the soil, leading to an increased mobility of heavy metals of geological origin. This may be the reason for increased arsenic concentration in Bangladesh (Ahmed and Amin, n.d.).

Changes in thermal regime

The thermal regime of surface water can be affected by land use practices. In small streams, removal of riparian vegetation can cause temperature increase in the water (thermal pollution) (Brooks et al., 1991). Also, tail water discharge from irrigated areas may cause a rise in temperature of the receiving stream (FAO, 1997a). A temperature rise leads to reduced oxygen solubility, which can negatively affect the biological activity in the water as well as the self-cleaning capacity of the river.

Scale consideration

The above review of land use impacts on water resources does not take into account spatial and temporal distribution aspects. Scale considerations, however, are of fundamental importance when assessing these impacts as they indicate whether a land use upstream may affect a water use downstream.

Spatial scale

With regard to the spatial scale, i.e. the size of river basin, the land use impact can become less important because of offset effects, such as de-synchronisation (e.g. in the case of floods), storage capacity of the river bed (sedimentation) or the self-cleaning capacity of the river (organic pollution). At the same time, the impact can become more important with increasing scale due to accumulative effects, e.g. in the case of salinity.

Land use induced changes of the hydrologic regime and sediment load decrease with the size of the river basin. The effects will be most readily felt in smaller watersheds of up to several hundred km². One well-documented case is the Ganges-Brahmaputra-Meghna basin. Studies show that in small-scale catchments (<50 km²) in the basin, erosion and stream flow may be strongly influenced by changing land use patterns (Ives and Messerli, 1989). However, the lowland flooding in Bangladesh is not related to the increased peak flow and erosion resulting from deforestation in the Himalayan uplands in Nepal. The main driving forces behind the flood events in the plains are naturally occurring rainfall events in the lowlands, which may be augmented by human interventions in the floodplains, such as road and river embankments (Hofer, 1998a; Ives and Messerli, 1989). Similarly, the bulk of the sediment load in the Ganges-Brahmaputra river system does not stem from human-induced erosion, but rather from large landslides not influenced by human activity (Bruijnzeel and Bremmer, 1989).

With regard to water quality impacts, the picture is much less clear. Observations show that some land use impacts on water quality, like salinity or pesticide load, can also have downstream effects in medium to large watersheds, like the Murray-Darling basin (Australia) and the Colorado basin (USA/Mexico). Other downstream impacts, like organic matter and pathogens, are relevant only at smaller scales.

The spatial dimensions of land use effects can be summarized as follows:

Legend: × = Observable impact; - = no observable impact

Temporal scale

Temporal scale is another important aspect of land use impacts, as it determines the perception of the impact as well as the economic cost associated with it. Two aspects are important with regard to temporal scale of land use impacts. First, the time it takes for a land use to have an impact on downstream uses, and, second, in the case of negative impacts, the time it takes for remedial measures to take effect, if the impact is reversible.

The temporal scales of land use impacts vary widely. Depending on the impact, they may range from less than one year, as in the case of bacterial contamination, to hundreds of years, as in the case of salinity. Similarly, time scales of recovery from adverse impacts are very diverse, depending on the impact. However, in most cases, the time it takes to restore an aquatic system after an adverse impact is much longer than the time it takes for an impact to appear (Peters and Meybeck, 2000).

Conclusion

With regard to land use impacts on hydrologic regimes and sediment transport, there is an inverse relationship between the spatial scale in which the impacts can be observed and the scale in which the redistribution of benefits might be important. These impacts can be most readily felt in small spatial scales. At the same time, the number of water users who might benefit or suffer from this land use change, increases with the size of the watershed. Due to the decreasing magnitude of impact, the respective costs and benefits will be small. Impacts of land use practices on water quality, like salinity, pesticide pollution and eutrophication due to nutrient influx, however, may be relevant in medium- to large-scale river basins as well. These impacts may affect many downstream uses, including providers of drinking water, industries, fisheries and other agricultural uses.

Bibliography

Ahmed, I. & Amin, A. n.d. Arsenic on tap. http://www.dainichi-consul.co.jp/english/arsenic/as18.htm (23/11/98)

BGS/ODA/UNEP/WHO. 1996. Characterisation and assessment of groundwater quality concerns in Asia-Pacific Region. Doc. UNEP/DEIA/AR.96-1. Nairobi: UNEP.

Bosch, J.M., & Hewlett, J. D. 1982. A review of catchment experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology, 55: 3-23.

Bowling, L.C., & Lettenmaier, D.P. 1997. Evaluation of the effects of forest roads on streamflow in Hard and Ware creeks, Washington. University of Washington, Seattle, WA, USA Water Resources Series Technical Report, No.155.

Brooks, K.N., Ffolliott, P.F., Gregersen, H.M., & Thames, J.L. 1991. Hydrology and the management of watersheds. Ames, Iowa: Iowa State University Press.

Bruijnzeel, L.A. 1990. Hydrology of moist tropical forests and effects of conversion: A state-of-knowledge review. Paris: UNESCO International Hydrological Programme.

Bruijnzeel, L.A., & Bremmer, C.N. 1989. Highland-lowland interactions in the Ganges-Brahmaputra river basin: A review of published literature. ICIMOD Occasional Paper, No. 11.

Calder, I.R. 1998. Water-resource and land use issues. SWIM Paper 3. Colombo: IIMI.

Childon, P.J., Jégat, H.J., & Stuart, M.E. 1995. Groundwater and agriculture: The interrelationship. Proceedings of an international seminar, Merida, Venezuela 14-17 October 1992. British Geological Survey Technical Report, No.WD/95/26.

Chimatiro, S.K., & Vitsitsi, E.G. 1997. Impact of different land-use activities in the catchment on small-scale fish farming in the southern region of Malawi: Coordination with other sectors. p. 231-241, in: K. Remane (ed) African inland fisheries, aquaculture, and the environment. Rome: FAO.

Chomitz, K.M., & Kumari, K. 1996. The domestic benefits of tropical forests. A critical review emphasizing hydrologic functions. World Bank Policy Research Working Paper, No. 1601.

Falkenmark, M., Andersson, L., Castensson, R., & Sundblad, K. 1999. Water: A reflection of land use. Options for counteracting land and water mismanagement. Stockholm: Swedish Natural Research Council.

Falkenmark, M., & Chapman, T. (eds). 1989. Comparative hydrology. An ecological approach to land and water resources, Paris: UNESCO.

FAO. 1987. Guidelines for economic appraisal of watershed management projects. Written by H.M. Gregersen, K.N. Brooks, J.A. Dixon and L.S. Hamilton. FAO Conservation Guide No. 16.

FAO. 1993. Prevention of water pollution by agriculture and related activities. Proceedings of the FAO Expert Consultation. FAO Water Reports, No. 1.

FAO. 1995. Environmental impact assessment of irrigation and drainage projects. Written by Dougherty, T.C., & Hall, A.W. FAO Irrigation and Drainage Paper, No. 53.

FAO. 1996. Control of water pollution from agriculture. Written by Ongley, E.D. FAO Irrigation and Drainage Paper, No. 55.

FAO. 1997. Seawater intrusion in coastal aquifers. Guidelines for study, monitoring and control. FAO Water Reports, No. 11.

FAO/Netherlands. 1999. Water. Background Paper 6 of the Conference on the Multifunctional Character of Agriculture and Land. Maastricht.

Forsyth, T. 1999. Questioning the impacts of shifting cultivation. Watershed, 5(1): 23-29.

Graaff, J.de. n.d. The price of soil erosion. An economic evaluation of soil conservation and watershed development. Mansholt Study 3. Wageningen, Netherlands: Agricultural University.

Harden, C.P. 1993. Land use, soil erosion, and reservoir sedimentation in an Andean drainage basin in Ecuador. Mountain Research and Development, 13(2): 177-184.

Hofer, T. 1998a. Floods in Bangladesh. A highland-lowland interaction? Geographica Bernensia G 48.

Hofer, T. 1998b. Do land use changes in the Himalayas affect downstream flooding? Traditional understanding and new evidences. Memoir Geological Society of India, 19: 119-141.

Ives, J.D., & Messerli, B. 1989. The Himalayan dilemma. Reconciling development and conservation. London: United Nations University Press.

La Marche, J., & Lettenmaier, D.P. 1998. Forest road effects on flood flows in the Deschutes river basin, Washington. University of Washington, Seattle. Water Resources Series Technical Report, No.158.

Liniger, H., & Weingartner, R. 1992. Water and Soil resource conservation and utilization on the northwest side of Mount Kenya. Mountain Research Development, 12(4): 363-373.

Liniger, H., & Weingartner, R. 1997. Mountains and freshwater supply. Unasylva, 49(195): 39-46.

NCSAI [National Council of the Paper Industry for Air and Stream Improvement]. 1999. Scale considerations and the detectability of sedimentary cumulative watershed effects. NCSAI Technical Bulletin, No. 776. Research Triangle Park, NC, USA.

Peters, N.E. & Meybeck, M. 2000. Water quality degradation effects on freshwater availability: Impacts of human activities. Water International, 25(2): 185-193.

Postel, S. 1997. Last oasis: Facing water scarcity. 2nd edition. New York: W.W. Norton.

Tejwani, K.G. 1993. Water management issues: Population, Agriculture and Forests - a focus on watershed management. In: Bonell, M., Hufschmidt, M.M. and Gladwell, J.S. Hydrology and water management in the humid tropics. Paris: UNESCO, pp 496-525.

Volker, A., & Henry, J.C. (eds). 1988. Side effects of water resources management. Wallingford, UK: International Association of Hydrological Sciences.

2. Instruments and mechanisms for upstream-downstream linkages: a literature review - Benjamin Kiersch

Benjamin Kiersch, Land and Water Development Division
Food and Agriculture Organization
Rome, Italy

Introduction

This paper presents the results of a survey of mechanisms and instruments which may be applied to develop social, economic and institutional linkages between upstream land users and downstream water users, based on a desk study on land-water linkages in rural watersheds.

The survey focuses on mechanisms and instruments that are explicitly designed to link land users and water users in a watershed context. It proposes a typology of mechanisms and instruments and identifies examples in which these instruments have been applied.

Some of the instruments discussed below are not designed to provide linkages between specific upstream and downstream stakeholders, but rather between farmers or land users and the society at large, as in the case of some subsidy schemes. They have been included because they might also be applied in more specific upstream-downstream contexts. Furthermore, such instruments may substantively influence the magnitude of land use impacts on water resources. In some developing countries, for instance, fertilizer subsidies are as high as 50 to 60 percent, leading to low application efficiency and thus resulting in environmental pollution, with negative consequences for the water quality (Lankoski, 1996).

In practice, these measures usually do not stand alone, but a mix of instruments is employed. For example, economic incentives may be linked to awareness-building measures.

An essential prerequisite for the implementation of any mechanism or instrument linking upstream land users and downstream water users is the assessment of the downstream costs and benefits that arise from upstream land use. In practice, this assessment is a demanding task (see also discussion paper 1). There are many reasons for this, among others:

• Land use impacts may be difficult to distinguish from natural or other anthropogenic effects, e.g. from mining and urbanisation. In comparison with other anthropogenic or natural effects, the land use impacts may be small.

• Some land use impacts, e.g. sedimentation, may occur over long time scales, making the present net value of the impact very small.

• With regard to some land use impacts, i.e. on the hydrologic regime and sediment transport, there seems to be an inverse relationship between the scale at which the impacts can be observed and the scale at which the redistribution of benefits might be important. These impacts can be most readily be felt in small spatial scales, while the number of water users increases with the size of the watershed. Due to the decreasing magnitude of impact at larger scales, the respective costs and benefits to the individual water user may be small.

Instruments and mechanisms for upstream-downstream linkages

Instruments can be structured into: regulatory instruments; economic instruments; educational and awareness-building measures; mechanisms to increase market access; the building of organisational structures; and participatory approaches.

Regulatory instruments

Regulatory instruments (command and control measures) are widely used in developed countries to protect water resources from agricultural land use practices, including pollution. In Germany, the Federal Water Act provides for the establishment of water protection zones to protect public water supply sources, groundwater and watercourses from sediment, fertilizer and pesticide runoff. In these zones, practices that have a negative impact on water resources may be restricted or forbidden, such as the use of certain pesticides. Farmers suffering economic loss because of these restrictions are eligible for compensation by the state (Kraemer and Buck, 1997).

In Thailand, there have been attempts to protect watersheds by imposing land use restrictions according to the vulnerability of the area. The total land area has been divided in five watershed classes, from forested upland watersheds with steep slopes (class 1a) to gentle slopes and flat areas with intensive agriculture (class 5). The most vulnerable watersheds, comprising 16 percent of the total land area in Thailand, have been designated for protection from any use except forest and ecological rehabilitation, including an evacuation of any residents of these areas. This approach, however, has not proven feasible. There is strong political pressure from rural populations demanding compensation for relocation and the restrictions imposed on their lands, as well as limited enforcement capacities on the part of the authorities. As a consequence, the government has granted an amnesty to illegal forest squatters (Krairapond and Atkinson, 1998).

Economic instruments

Economic instruments to distribute benefits and costs resulting from land use impacts on water resources between upstream and downstream stakeholders include subsidies, taxes, and transferable property or use rights for land, water and emissions.

Subsidies

Subsidies include direct and indirect payments, such as tax exemptions, price regulations and protective measures, by the state to achieve certain objectives. With regard to land use impacts on water resources, there are direct and indirect subsidy schemes in place aimed at compensating farmers for the costs arising from water protection.

The water tax scheme in the province of Baden-Württemberg, Germany, for example, illustrates the use of both direct and indirect subsidies. A tax is levied on all surface and groundwater abstractions. The proceeds are used to finance compensation payments to farmers for restrictions on fertilizer use in water protection zones (direct payment). Farmers can get a rebate of up to 90 percent on the tax payments for agricultural water use (indirect payment). However, this subsidy is conditional on taking all available measures to save water, and on using surface water instead of groundwater (Kraemer and Buck, 1997). The latter condition is imposed because groundwater is the primary source of drinking water in Germany.

Under the New York City Watershed Agreement, owners of forests are eligible for a rebate of up to 80 percent on their property taxes if they prepare a forest management plan and commit themselves to implementing it over a period of 10 years. The management plan includes measures for the maintenance of water quantity and quality (New York Watershed Agricultural Council homepage; Tobias, 1999).

In the UK, there is a subsidy scheme in place which compensates farmers for the adoption of farming practices that reduce nutrient leaching, such as reduced fertilizer use or the conversion of arable land to grassland (Kraemer and Buck, 1997).

Taxes

Taxes are another instrument employed to curb negative impacts on water resources of land use practices. The economic incentive for the farmer is the same as in the case of the subsidy. There is an important difference, however: in the case of subsidies, the government pays the farmer for avoiding pollution, while with taxes, the farmer has to pay for activities which are increasing pollution, or the pollution itself. From a property rights viewpoint, the subsidy approach implicitly gives the environmental property rights to the farmer, while in the case of taxes, the rights are allocated to society at large, and the farmer has to pay to use them.

Possible approaches include taxes on agricultural inputs (fertilizer, pesticides), taxes based on the nutrient balance, and taxes based on effluent concentration. Taxes on inputs are easy to implement, but their environmental impact may be low for several reasons (see discussion in Lankoski, 1996). Thus, their principal effect may be the reduction of farmer’s income. Alternative approaches like a tax based on nutrient balance or effluent concentration may be more efficient in controlling pollution. Their implementation, however, poses immense difficulties, because of problems related to the assessment of non-point source pollution.

Flexible property or use rights

One way of protecting water resources from land use impacts is through the acquisition of land or land use titles. The New York City Watershed Agreement provides an example of this approach. Under the agreement, the City of New York can purchase land in sensitive areas (e.g. near watercourses, wetlands and reservoirs) in upland watersheds to protect its water supply. These lands are set aside from use or can only be used for certain recreational activities, like hiking or fishing with a special permit. Alternatively, the city can purchase the right to develop a property through a so-called conservation easement. Under this arrangement, the land remains the property of the original owner. However, the owner forfeits the right to develop the land, e.g. to construct buildings or roads. The conservation easement is granted for an indefinite time (Tobias, 1999).

Another possibility to account for land use impact is to introduce a system of permits for river pollution. An example of such an instrument is the salinity mitigation programme within the Murray-Darling river basin in Australia. In view of the increasing cost of salinity for downstream users (cities, industry and agriculture) the three riparian states jointly financed a programme to divert saline groundwater seepage in the lower part of the basin, which decreased downstream salinity. In return, the upstream states receive entitlements to dispose of saline drainage water from irrigation within defined limits. A state can increase its "salinity credit" by contributing to the costs of further downstream groundwater diversion projects. The limited number of available salt disposal entitlements has led to major improvements in irrigation practices and water use efficiency in upstream states. The costs for improvements are shared by the community and the state government. The state government has an incentive to contribute to the improvements to avoid having to invest in projects to increase the salt disposal entitlements (Murray-Darling Commission home page).

Educational and awareness building

Educational programs are used to encourage farmers to switch to less polluting farming practices. In the New York City Watershed Agricultural Program, farmers can participate in environmental audits of their business, which include the identification of potential pollution sources, pollution barriers and hydrologically sensitive areas. (New York Watershed Agricultural Council home page; Walter and Walter, 1999) Similarly, in the UK, a programme of the Ministry of Agriculture offers free farm visits to prepare pollution-risk assessments and waste management plans for farmers (Kraemer and Buck, 1997).

These programmes are usually coupled with an incentive programme to reduce pollution risks and to improve the economic performance of the farm. In the New York case, for example, farmers receive financial assistance for pollution-preventing structures, such as cement pipes.

Market support

The improved access of upstream farmers to downstream markets is another mechanism to improve cooperation between stakeholders: one that may increase farmers’ income and, in the framework of a watershed agreement, can be used as an incentive to conserve the resources. (Preston, 1997) For example, under the New York City Watershed Agricultural Program, restaurants, markets and purveyors have committed themselves to purchasing produce from participating farmers (New York Watershed Agricultural Council homepage).

Organizations

Organizational development is a prerequisite for the successful implementation of instruments establishing linkages between upstream land users and downstream water users.

Organizations have two important functions:

First, they provide a forum of exchange between upstream and downstream stakeholders. The institutional framework in the Murray-Darling river basin is a good example of such an entity. It consists of three bodies: the Murray-Darling Ministerial Council and the Murray-Darling Basin Commission, comprising ministers of the riparian states and the Australian Federal Government, and the Community Advisory Committee, which is made up of representatives of the watersheds that make up the basin, together with special interest groups. Within this framework, decisions are made regarding the sharing of water resources and management costs, as well as long-term management planning for the basin. Decisions in the Ministerial Council and the Basin Commission are taken unanimously, i.e. with the consent of all riparian states (Murray-Daring Basin Commission home page).

Second, organizations are vital as a forum to consolidate the interests and opinions of scattered groups of users, e.g. upstream farmers. An example is the Watershed Agricultural Council in New York. This entity was formed by farmers and agri-business leaders upstream of New York City in order to negotiate the Watershed Agreement with the City of New York. The Council now administers the Watershed Agricultural Program aimed at securing the city’s drinking water supply (Walter and Walter, 1999).

Participatory approaches

Participatory approaches to curb negative impacts of land use practices on water resources are frequently applied to improve the management of natural resources, for example through soil and water conservation, and to increase sustainability by including the local population in the planning and implementation process. In addition to the environmental benefits, the aim of the participatory approaches includes economic benefits such as improved farmer income and better livelihood security, as well as social benefits such as the establishment of organizations and decreased out-migration.

Usually, the participatory watershed planning and management projects focus on the community level and encompass only very small land units. Reviews suggests that participatory watershed management projects on this level have been very successful and yield better results than soil and water conservation projects focusing on individual farm plots (Hinchcliffe et al., 1995; Farrington and Lobo, 1997).

There are some problems with employing participatory approaches to address problems between upstream and downstream communities.

First, due to the small scale of this approach, the benefits accrue mostly to the participating farmers themselves. Second, sometimes the hydrological watershed is not socially meaningful as a planning unit for local people, for example, if a community extends over several watersheds. In order for the participatory approach to work, the planning area might have to be adjusted, which might make it more difficult to establish upstream-downstream links (FAO, 1996; FAO, 1998). Third, the scaling-up of participatory approaches to larger watershed units is difficult as it involves the cooperation of government agencies and the establishment of organizations at the watershed level. In particular, this applies to project implementation. In a case study of tank irrigation systems in Sri Lanka, it was found that while in theory water availability could be greatly enhanced through participatory planning at the level of sub-watersheds, implementation of these plans proved to be impossible because of a lack of organizations at the watershed level and reluctance of the local government to support the plans (Jinapala et al., 1996).

Conclusion: Criteria for successful implementation of instruments for upstream-downstream linkages

As a first result of the survey, the following criteria of success for the implementation instruments can be formulated:

• The impact of upstream land use on downstream water use is well understood.

  A clear understanding of the land use impact is an essential prerequisite for the implementation of upstream-downstream links. As the impacts of land use vary greatly because of site-specific conditions, they must be carefully evaluated in each case (see discussion paper 1).

• The impact of land use on water resources clearly dominates over natural impacts or other anthropogenic impacts.

  In many instances, natural or anthropogenic impacts are present that overlap with land use impacts. In these cases, upstream-downstream links focusing on land uses will only be successful if it can be established that the upstream land use is the dominating factor determining water availability and quality downstream. This may become more difficult as the size of the river basin increases (see discussion paper 1). For Andean watersheds, for example, it has been estimated that landslides occurring during high rainfall years, stream bank erosion, roads and urban development cause more erosion than agricultural activities (Estrada and Posner, 1999).

• The groups of upstream and downstream stakeholders are few and well organized.

  The fewer the groups of stakeholders living in a watershed, the easier it will be to implement upstream-downstream links. Similarly, if upstream land users and downstream water users are well organized (e.g. in production associations, water user groups, or through a municipal water utility), the establishment of linkages will be more effective than if there is no connection between individual stakeholders within a group e.g. upstream farmers.

• The economic impact on up- and downstream stakeholders can be quantified.

  In the case of New York, for example, the cost of installing filtration equipment to provide safe drinking water was US$6 000 to 8 000 million (Hofmann, 1999), while upper watershed protection measures cost US$600 million, providing a powerful incentive to invest in watershed protection. (Watershed Protection Update).

• The difference in costs between alternatives must be high enough to provide an incentive for the participation in the linkage programme.

  If costs for downstream users are small or occur in the distant future, downstream users are not likely to invest in upper watershed protection. For example, in Columbia, managers of a hydro-electric dam in the Rio San Antonio watershed may be reluctant to finance the conversion of erosion-intensive coffee plantations upstream of the dam because, at current erosion rates, the hydro-electric capacity will only be affected after 80 years (Estrada and Posner, 1999).

  The absence of a sufficient economic value for downstream users of upstream land use impacts, however, does not imply that upstream watershed protection is not necessary. In some cases, downstream users might be dependent on the resources, but engage in activities that do not generate enough revenue (e.g. subsistence farming) to fund upstream conservation measures (Estrada and Posner, 1999). In other cases, there might be impacts which affect environmental goods which do not have an economic value (e.g. aquatic ecosystems which rely on a minimum instream flow).

• There is political commitment to establish upstream-downstream linkages.

  If linkages between upstream and downstream users are developed, there has to be a commitment from the government having jurisdiction over the watershed. National and local governments can act as important facilitators and provide institutional support for upstream-downstream linkage programmes. Conversely, if upstream-downstream linkages are not supported by governments, it may be difficult to implement programmes that go beyond the community scale.

• There is a strong institutional and legal framework that facilitates the implementation of linkage instruments.

  Many linkage instruments require a stable institutional and legal framework for implementation. For example, for the transfer of property rights over land and water, it is imperative that these rights are defined and allocated to land and water users in the first place. Similarly, if regulatory instruments are implemented, institutions are necessary to enforce the regulations.

Bibliography

Abernethy, C.L. 1998. Institutions and processes for sharing water among multiple types of use. Agriculture and Rural Development, 2: 10-13.

Bakker, M., Barker, R., Meinzen-Dick, R., & Konradsen, F. (eds) 1999. Multiple uses of water in irrigated areas: A case study from Sri Lanka. SWIM Paper 8. Colombo, Sri Lanka: IWMI.

Barrow, C.J. 1998. River basin development planning and management: A critical review. World Development, 26 (1): 171-186.

Dourojeanni, A., & Jouravlev, A. 1999. El Código de Aguas en Chile: Entre la ideología y la realidad. Document LC/L.1263-P. Santiago: CEPAL.

Estrada, R.D., & Posner, J. 1999. Heresy from the hilltop: How useful is the watershed as an organizing principle for research and development? Draft Paper. Lima: CONDESAN. (http://www.inrm.cgiar.org/documents/centers/heresy2.pdf)

FAO. 1994. Water policies and agriculture. In: The State of Food and Agriculture: 228-297.

FAO. 1994a. Preparing national regulations for water resources management. Principles and practice. Written by Burchi, S. FAO Legislative Study, No. 52.

FAO. 1995a. Water sector policy review and strategy formulation. A general framework. FAO Land and Water Bulletin. No. 3.

FAO. 1995b. Reforming water resources policy. A guide to methods, processes and practices. FAO Irrigation and Drainage Paper, No. 52.

FAO. 1995c. Methodology for water policy review and reform. FAO Water Report, No. 6.

FAO. 1996. Steps towards a participatory and integrated approach to watershed management. Report of the Inter-regional Project for Participatory Upland Conservation and Development, GCP/INT/542/ITA, FAO/Italy Cooperative Programme. Tunis. Written by Fé d’Ostiani, L., & Warren, P. (eds).

FAO. 1998. Developing participatory and integrated watershed management. A case study of the FAO/Italy Inter-regional project for Participatory Upland Conservation and Development (PUCD). Written by Warren, P. FAO Community Forestry Case Study Series, No. 13.

Farrington, J., & Lobo, C. 1997. Scaling up participatory watershed development in India: Lessons from the Indo-German watershed development programme. ODI, London, Natural Resource Perspectives. No. 17.

Hearne, R.R., & Trava, J.L. 1997. Water markets in Mexico: Opportunities and constraints. IIED, London, Discussion Paper, No. 97-01.

Hinchcliffe, F., Guijt, I., Pretty, J.N., & Parmesh, S. 1995. New horizons: The economic, social and environmental impacts of participatory watershed management, IIED, London, Gatekeeper Series, No. 50.

Hoffman, R. 1999. The New York Ciy Watershed Agreement. Water Resources Impact 1(5): 2-4.

Jinapala, K., Brewer, J.D., & Sakthivanidivel, R. 1996. Multi-level participatory planning for water resourcs development in Sri Lanka. IIED, London, Gatekeeper Series, No. 62.

Kraemer, A., & Buck, M. 1997. Water subsidies and the environment. Report OECD/GD(97)220. Paris: OECD.

Krairapanond, N., & Atkinson, A. 1998. Watershed management in Thailand: Concepts, problems and implementation. Regulated Rivers: Research and Management, 14: 485-498.

Lankoski, J. 1996. Controlling agricultural nonpoint source pollution: The case of mineral balances. UNCTAD, Geneva, Discussion Paper No. 116.

Le Moigne, G., Easter, K.W., Ochs, W.J., & Giltner, S. (eds) 1994. Water policy and water markets. World Bank Technical Paper, No. 249.

Lyon, R.M. 1989. Transferable discharge permit systems and environmental management in developing countries. World Development, 17 (8): 1299-1312.

Murray-Darling Basin Commission homepage: http://www.mdbc.gov.au (17/11/99)

New York Watershed Agricultural Council homepage: http://www.nycwatershed.org (18/11/99).

Perry, C.J., Rock, M., & Seckler, D. 1997. Water as an economic good: A solution, or a problem? IIMI, Colombo, Research Report No. 14.

Pratt, D.J., & Preston, L. 1998. The economics of mountain resource flows. Unasylva, 49(195): 31-38.

Preston, L. (ed) 1997. Investing in mountains: Innovative mechanisms and promising examples for financing conservation and sustainable development. Franklin, WV, USA: The Mountain Institute.

Rhoades, R.E. n.d. Participatory watershed research and management: Where the shadow falls. IIED, London, Gatekeeper Series, No. 81.

Sanders, D.W., Huszar, P.C., Sombatpanit, S., & Enters, T. (eds) 1999. Incentives in soil conservation: From theory to practice. New Delhi: Oxford & IBH.

Tobias, D. 1999. Protection of New York City’s water supply through land acquisition and stewardship. Water Resources Impact, 1(5): 9-15.

Walter, M.T., & Walter, M.F. 1999. The New York City Watershed Agricultural Program: A model for comprehensive planning for water quality and agricultural economic viability. Water Resources Impact, 1(5): 5-9.

Watershed Protection Update. http://www.ci.nyc.ny.us/html/news/wsprot.html. (12/11/1999)