7 Land use, hydrological function and economic valuation

B. Aylward
Deschutes Resources Conservancy, Bend, USA

I N T RO D U C T I O N impacts on the economy can be summarised according to whether

they feed back into the economic system through a reduction in
Land use change affects economic activity both directly and indi- on-site production (soils) or through a more distant, downstream
rectly. In the process of land colonisation that accompanies eco- impact on off-site production or consumption (streamflow quality
nomic development and population growth, naturally occurring and quantity).
vegetation is typically affected in one of three ways: (1) avail- To economists the theoretical implications of the on-site impacts
able biomass and species are harvested and then left to regenerate of land use change are fairly straightforward. In a farming context,
before harvesting again, (2) the vegetation is simplified (in terms McConnell demonstrates that as long as farmers’ objectives are
of its biological diversity) in order to increase production from consistent with society’s objectives and social and private discount
selected species or (3) the existing vegetation is largely removed rates are identical, on-site losses of productivity due to soil erosion
to make way for the production of domesticated species, the instal- can be expected to follow an optimal path (McConnell, 1983).
lation of infrastructure or urbanisation. The direct, and desired, That is, soil would be ‘used’ over time so as to maximise the net
impact of land use change under these circumstances is to raise present value of its contribution to production. The question of
the economic productivity of the land unit. Of course, many indi- course is whether the assumptions of McConnell’s model hold in
rect (and perhaps unintentional) environmental impacts result as the real world. As a result, considerable effort has been devoted
well. These impacts reflect the economic values attributed to nat- to investigating policy, institutional and social imperfections that
ural vegetation and biogeophysical processes. Conversely, efforts may lead to excessive rates of soil degradation (loss of soil depth or
to recuperate degraded lands or to protect natural ecosystems may soil quality). Nevertheless, in the absence of serious imperfections,
forsake direct productive benefits in favour of fostering these indi- neoclassical economists are fairly sanguine about the ability of the
rect environmental values. market to provide a relatively efficient level of incentive for soil
The loss of biodiversity and alteration of ecological processes conservation (Crosson and Miranowski, 1982; Southgate, 1992;
accompanying the logging and conversion of forestland have cap- Lutz, Pagiola, and Reiche, 1994).
tured the public imagination in the 1990s, with corresponding In addition to the on-site impacts of soil degradation, a series
growth in research aimed at illustrating these indirect ecolog- of downstream hydrological impacts also accompany the distur-
ical and economic impacts (Perrings, Folke, and Maler, 1992; bance of natural vegetation. Regardless of the perceived seri-
Barbier, Burgess, and Folke, 1994). This chapter concerns itself ousness of the ‘soil erosion problem,’ economists and natural
with another type of environmental value: the impact of land use scientists have traditionally agreed that the downstream effects
change on the hydrological cycle. Vegetation is an important vari- of land use change are potentially very serious (Crosson, 1984;
able in the hydrological cycle as it is the medium through which Clark, 1985b; Pimentel et al., 1995). This belief is based on
rainfall must pass to reach the soil and begin the journey back the general perception that the hydrological impacts of land
to the sea. Further, land use change invariably involves not just use change have unambiguously negative impacts on produc-
modification of land cover but alteration of soil surface and sub- tion and consumption and the suspicion that these impacts are
surface conditions. The hydrological impacts that result from these often large in magnitude. As the effects are external to the land
changes are often grouped in terms of their impact on soils and use decision-making process of landholders, the failure of the
changes in streamflow quality and quantity. The nature of these market to internalise these effects (externalities) is unquestioned.

Forests, Water and People in the Humid Tropics, ed. M. Bonell and L. A. Bruijnzeel. Published by Cambridge University Press.


C UNESCO 2004.

99
100 B . AY LWA R D

Consequently, this chapter uses the term ‘hydrological externali- L A N D U S E A N D H Y D RO L O G Y

ties’ to refer to these downstream hydrological impacts of land use
change. As a means of introducing the hydrological issues and concepts
This chapter examines the existing knowledge base with regard employed, a brief overview of the hydrological impacts of land
to the application of the tools of economic analysis to the valuation use change is provided, particularly as it relates to the case of the
of these hydrological externalities of land use change, with an humid tropics.
emphasis on the humid tropics. The objectives are to:
r specify the general theoretical linkages that govern the rela- Hydrological impacts of land use change
tionships between land use, hydrological function and down-
stream economic welfare; Disturbance of tropical forests can take many different forms,
r assess the existing empirical evidence in the economics liter- from light extraction of non-timber forest products through to
ature regarding the significance of these hydrological exter- wholesale conversion. Each type of initial intervention will have
nalities; and its own particular impacts on the pre-existing hydrological cycle
r assess what a priori claims can be made regarding the (Hamilton and King, 1983). These hydrological impacts may be
direction and magnitude of the economic consequences of loosely grouped according to whether they are related primarily
land use change and resulting downstream hydrological to water quality or water quantity. Under this typology erosion,
impacts. sedimentation and nutrient outflow are grouped together under
the heading of water quality impacts; and changes in water yield,
Interest in the environmental benefits provided by forests and seasonal flow, stormflow response, groundwater recharge and pre-
catchment management has never been greater (Johnson, White, cipitation are considered as water quantity issues. Beginning with
and Perrot-Maitre, 2001). Investments in forest conservation and water quality and moving on to water quantity, the hydrologi-
catchment management and the derivation of new regulations and cal impacts of changes in land use and conversion of tropical
market incentives in this regard are of increasing importance in forests can be summarised by compiling the general nature of these
both temperate and tropical zones. Thus, a systematic understand- impacts as extracted from a number of authoritative reviews on
ing of the relationships between upstream land use, hydrology and the subject, including those in this volume (Hamilton and Pearce,
downstream economic activity, as well as practical methods for 1986; Bruijnzeel, 1990; Calder, 1992; Bruijnzeel and Proctor,
the quantitative evaluation of these linkages is required to guide 1995; Bruijnzeel, 1997, 1998, 2002; Pielke et al.,1999; and Bonell,
project investments and policy-making. Callaghan and Connor; Bonell et al.; Bruijnzeel; Chappell et al.;
Given the emphasis in other chapters of this book on the latest Grip et al.; Heil Costa; Scott et al.; this volume).
scientific findings in forest hydrology, the chapter begins with
just a short and stylised summary of the biophysical impacts (1) Erosion increases with forest disturbance, at times dramati-
of land use change on hydrological function (sedimentation, cally, depending on the type and duration of the intervention.
water yield, seasonal flows, peakflows, etc.). This knowledge (2) Increases in sedimentation rates are likely as a result of
is used as a point of departure for a simple theoretical presen- changes in vegetative cover and land use and will be deter-
tation of the linkages between land use, hydrology and indi- mined by the kind of processes supplying and removing
vidual utility. Hydrological services may enter into an individ- sediment prior to disturbance.
ual’s utility function directly through consumption, indirectly (3) Nutrient and chemical outflows following conversion gen-
through the household production function or as factor inputs in erally increase as leaching of nutrients and chemicals is
production. increased.
The types of economic impacts that can be expected to result (4) Water yield is related inversely to forest cover, with the excep-
from changes in hydrological services that are, in turn, related tion of upper montane cloud forests where horizontal precip-
to changes in land use are then reviewed. The range of impacts itation may compensate for losses due to evapotranspiration.
that are caused by land use and subsequent hydrological change is (5) Seasonal flows, in particular dry season baseflow, may
amply demonstrated in the literature and the magnitude of these increase or decrease depending on the net effect of changes
impacts is discussed. The ensuing section then discusses the gen- in evapotranspiration and infiltration.
eral nature of these linkages between land use and hydrological (6) Peakflow may increase if hill-slope hydrological conditions
externalities, drawing upon the empirical and theoretical ideas pre- lead to a shift from sub-surface to overland flows, although
sented in the two previous sections. A final section summarises the the effect is of decreasing importance as the distance from
findings of the chapter and presents recommendations for future the site and the number of contributing tributaries in a river
research in this area. basin increase.
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 101

(7) Groundwater recharge is generally affected in a similar fash- However, given the existence of at least the possibility of one
ion to seasonal flows. relationship that is decreasing (baseflow) no generalisation can be
(8) Local precipitation is probably not significantly affected by made about the net hydrological impact of a given change in land
changes in forest cover (at least up to a scale of 10 Km). use in terms of first order effects. In any case, such a generalisation
Exceptions are cloud forests (elevated cloud base following would have little meaning in practical terms as the direction of
large-scale downslope forest clearance) and large continental change of the hydrological function does not predetermine the
basins (such as the Amazon which is partially enclosed). direction of the accompanying change in economic welfare.
Three possibilities present themselves as to how the vector of
Finally, the authors cited above generally agree that in assess- hydrological outputs relates to utility (the economist’s measure of
ing the hydrological impact of land use changes it is important well-being):
to consider not just the impacts of the initial intervention but
also the impacts of the subsequent form of land use, as well as (1) H may enter directly into individual utility, for example if
the type of management regime undertaken (Bosch and Hewlett, the degree of suspended sediment in surface waters affects
1982; Hamilton and King, 1983; Bruijnzeel, 1990; Calder, 1999; the aesthetic pleasure derived by a recreationalist from sight-
Bruijnzeel, 2002). seeing or hiking.
(2) H may be an input into the household production of utility-
yielding goods and services, for example if poor quality of
L A N D U S E C H A N G E , H Y D RO L O G Y A N D water drawn from a stream affects the health of people in the
E C O N O M I C W E L FA R E household.
(3) H may serve as a factor input in the production of a marketed
A change in hydrological function as provoked by alteration of good that in turn enters into the production of other marketed
land use or land management practices will lead to changes in goods, household production or individual utility: for exam-
the downstream hydrological outputs associated with a given land ple if streamflow is used for hydroelectric power generation
unit. These outputs may be summarised generally as consisting of which in turn is consumed by businesses, households and
the streamflow over a given time period and the level of sediment individuals.
and nutrient concentrations contained in this streamflow. The spa-
tial and temporal point at which these outputs are evaluated will A simple theoretical presentation of each of these cases is pre-
depend on the type and location of the affected economic activ- sented below. In the discussion, an effort is made to identify the
ity. However, in general, a hydrological production function for a general type and nature and importance of downstream effects as
given site can be defined that relates land use, L, and a vector Y of they are felt through each medium in developed and developing
other biophysical parameters to a vector of hydrological outputs, economies (Freeman, 1993).
as follows: The approach taken in this chapter tends to focus on the ways
in which land use affect hydrology and the ways that the result-
H = H(L , Y) (7.1)
ing physio-chemical changes (in water, nutrients, sediment, etc.)
The vector H then refers to the different hydrological outputs feed into the economy. This is a very linear and straightforward
(H = h1 , . . . , hi , . . . , hm ) including sediment yield, annual water approach to what is necessarily a complex and intertwined set of
yield, peakflow, dry season baseflow, etc. Somewhat arbitrarily, L factors and events.
is defined such that an increase in L represents a change away from The same changes in land use and in hydrology may also affect
undisturbed natural forest (or vegetation) towards less vegetation economic activity through knock-on effects that are transmitted
and a more ‘productive’ land use. As noted above, the removal through changes in riparian zone and aquatic ecology. Changes in
of forest cover tends to increase sediment yield, SY, as well as water quality and timing of water flow can have important ecolog-
raising nutrient and chemical levels, FL. Similarly the effect of ical impacts that affect, for example, fish populations and those
an ‘increase’ in land use is to raise annual water yield, WY, as who depend on fish for their livelihood or income. At the same
well as peakflows, PF. The effect on dry season baseflow, BF, is time changes in land use such as forest conversion or restoration
indeterminate. Thus a majority of the relationships between land can have direct impacts on these same riparian zones and aquatic
use and individual hydrological functions are increasing: ecosystems. Increases in light due to reduction in forest cover
may lead to beneficial impacts on fish – at least up to some point
∂SY ∂FL ∂WY ∂PF ∂BF
> 0, > 0, > 0, > 0, < 0, (Zalewski, Thorpe, and Naiman, 2001). Even further, downslope
∂L ∂L ∂L ∂L ∂L
∂BF riparian zones may play an important role in mitigating changes
or > 0. in water quality due to forest conversion upstream (Hubbard and
∂L
102 B . AY LWA R D

Lowrance, 1996; Snyder et al., 1998; Sheridan, Lowrance, and In developed economies, the principal manner in which change
Bosch, 1999). in hydrological function will affect utility directly, would be a
Examination of these ‘ecohydrology’ impacts remains a rel- change in water quality or quantity that directly affects aesthetic
atively young science and the integration of these impacts into values. As in the example mentioned above, muddied waters may
empirical work on economic valuation is a challenge for the future. affect the attractiveness of a recreation or urban site, which then
It is important, however, to note that the addition of an ecohydrol- directly reduces the utility associated with the aesthetic aspect of
ogy perspective to the argument presented here would not change the experience. There is also the possibility that people may hold
the outcome fundamentally – many of the studies that are emerging existence values for the natural streamflow regime. For example,
suggest that, a priori, ecosystem modification cannot be consid- individuals may derive satisfaction or pleasure directly from the
ered to be ‘negative’ and that ecosystems can indeed be managed knowledge that free-flowing rivers continue to exist in their nat-
in order to optimise the services they provide. ural state, regardless of their past or planned future use of the
river or its associated products and services. Donations to river
conservation organization are one example of how such existence
Hydrological outputs that enter directly into utility values translate into willingness-to-pay for conservation. In devel-
As it is practically impossible for an upstream land user to pre- oping economies it is more difficult to conceive of many instances
vent downstream users from enjoying or suffering (as the case where water quantity and water quality will simply be consumed
may be) the consequences of upstream land use change, hydro- directly by an individual, that is enter directly into the utility (or
logical functions may be considered as non-exclusive in nature economic welfare) of the individual (Hearne, 1996). The excep-
(Aylward and Fernández González, 1998). Absent regulation pro- tion may be the very poor where existence is litterally ‘hand to
ducers are unlikely to bear any downstream costs attributable to mouth’. In any event it is probable that hydrological outputs are
their upstream activities. Likewise, upstream ‘producers’ cannot more likely to enter directly as an input into household produc-
capture any downstream benefits of their actions (or their restraint) tion processes in rural, developing households than in developed
by selling hydrological outputs in markets. This is not to preclude countries (or urban, developing households) where the house-
the possibility that property rights exist for these outputs further hold typically purchases basic services from public or private
downstream. In many areas, for example, streamflow is appropri- utilities.
ated under a system of private property rights. Deposited sedi-
ment may also be a marketable commodity once it is deposited.
Hydrological outputs as inputs to the
For example, in Thailand sediment dredged from rivers is subse-
household production
quently resold (Enters, 1995). To the extent that these rights or
products are then tradeable, these hydrological outputs may be In the case of the household production function, utility of the
marketable. household is assumed to be derived from a vector of final services,
However, these cases involve the development of exclusiv- Z, that yield utility:
ity, whether through institutional arrangements or investment in
U = U (Z) = u(z 1 , . . . , z k , . . . , z 0 ) (7.4)
resource harvesting, only at the downstream end of the ‘produc-
tion’ change. It remains the case that an upstream change in land These final services are themselves produced by a technology that
use will alter the physical availability of the output regardless of is common to all households and employ as inputs vectors of both
any legal claim to the output, whether constituted as streamflow marketed goods and non-marketed hydrological outputs:
or sediment.1 For this reason the vector of hydrological outputs
z k = z k (X, H) (7.5)
may be assumed to enter into utility as a non-marketed good or
service alongside a vector of marketed goods, X: For example, changes in dry season baseflow or water quality (H)
may affect the quantity of bottled water or the number of water
U = U (X, H) (7.2) filters (X) that are purchased by the household in providing drink-
where U(•) is a well-behaved and increasing individual utility ing water (zk , the utility-yielding service) for household mem-
function and X is composed of private good quantities (X = bers. Again, the budget constraint can be formulated as reflecting
x1 , . . . , xj , . . . , xn ). The individual is then assumed to maximize the need to spend less on the marketed goods than is available
utility subject to the budget constraint, where M equals money
1 For an in-depth discussion of this topic and the possibility of a ‘Coasian
income and p refers to the prices of the marketed goods:
Bargain’ wherein upstream and downstream parties may develop a voluntary


n arrangement that is in the interest of both parties see (Aylward and Fernández
pjxj ≤ M (7.3) González, 1998) and for real-world examples see (N. Johnson et al., 2001;
j=1 Rojas and Aylward, Forthcoming).
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 103

in money income. Thus, the household is assumed to maximise their economic impacts will be site specific, it is not possi-
utility subject to the budget constraint, the level of H and the ble, a priori, to draw any generalisation about which effect will
constraints implicit in Eqn 7.5. predominate.
In developed countries this model is applicable to certain cases Change in hydrology will thus alter both the cost curve for x as
of recreation. For instance, streamflow may be a factor along with well as the demand for inputs of capital, k, and labour, w. Given
canoes, equipment and other inputs in producing a household factor prices, p, the cost function is:
canoeing trip. Similarly, changes in water quality may affect river-
C = C( p w , p k , x, H) (7.7)
ine, estuarine or lacustrine ecological conditions, in turn affect-
ing biomass and species composition of systems that are prized The producer is assumed to minimise cost and the impacts of
for fishing or diving. Stormflow and flooding are other exam- a change in H are felt by consumers (as prices change) or by
ples where hydrological outputs may affect developed households producers in the input markets (as demand, and hence prices, for
directly, but by and large household ‘use’ of water and other capital and labour inputs change).
hydrological outputs is more often achieved through the purchase As suggested above, the analysis of economic consequences of
of marketed outputs produced by the state or the private sector, changes in land use and hydrology for developed countries will
for example potable water for domestic use, electric power from often draw on this formulation of the problem, particularly as
hydroelectricity, food produced by irrigators and navigation from it relates to impacts on hydroelectric power production, domestic
ferry services. water treatment and supply, and industrial water supply. The same
In developing countries, the use of water for recreation is likely goes for developing countries where urban households, industrial
to be limited to that by higher income or foreign recreationalists. concerns and commercial farmers purchase water-related products
Most probably, hydrological function more directly affects the from public/private utilities and state agencies.
rural household that ‘uses’ water for domestic and agricultural use,
waterways for navigation, and waterpower as an energy source.
Thus, streamflow and water quality may serve as inputs (along D OW N S T R E A M E C O N O M I C I M PAC T S
with other marketed or non-marketed inputs of labour and capi- O F C H A N G E S I N H Y D RO L O G I C A L
tal) into the preparation of food and drink, subsistence farming, FUNCTION
transport of produce to market, the accomplishment of repetitive,
small-scale mechanical tasks, etc. In developing countries then the A number of the points typically held as conventional wisdom
bulk of rural populations will experience the hydrological impact regarding the downstream impacts of changes in hydrological
of land use change through the household production function. function needexamination . . . The empirical literature on the eco-
nomic valuation of hydrological externalities is then reviewed.
This literature is critiqued as a prelude to the next section, which
Hydrological outputs as factor inputs into production
revisits conventional wisdom on the topic in drawing some gen-
The vector of hydrological outputs can also appear directly in the eral conclusions regarding the direction and magnitude of these
production function along with other factor inputs. Production of externalities.
the marketed good, x, then depends on the production function as The conventional wisdom emerging from the literature holds
follows:2 that forest conversion (or ‘deforestation’ as it is often called
in developing countries or clear-cutting in developed countries),
x = x(k, w, . . . , H) (7.6)
leads to large costs in terms of losses in on-site productivity and
Production is assumed initially to be an increasing function of costly sedimentation of downstream hydropower, water supply
capital, k, and labour, w, so that an additional unit of each will and irrigation facilities. In addition, conventional wisdom holds
yield an increase in x. Typically, production is assumed to be an that the forest attracts rainfall and acts as a sponge, soaking
increasing function of the environmental service. As formulated up and storing excess water for use at later times, thus provid-
in the case of H, this may not be strictly true. An increase in water ing benefits in terms of increased water supply, flood reduction,
yield may be beneficial while an increase in sediment yield may improved navigation and dry season flow to agriculture and other
not improve production. For example, an increase in streamflow productive activities. Although these views seem to be shared
(as a result of forest conversion) may be assumed to have a posi-
tive impact on production in the case of hydroelectric power gen- 2 Following on the tradition of ‘bioeconomic’ modelling, such a production
eration. Meanwhile, an increase in sediment delivery may lower function could be called a ‘hydroeconomic’ production function. However,
in order to avoid confusion this function is simply referred to as an ‘eco-
production, other things being equal – e.g. holding expenditure nomic’ production function in order to distinguishes it from the ‘hydrolog-
on dredging constant. Given that the hydrological functions and ical’ production functions that model the land use-hydrology relationship.
104 B . AY LWA R D

across developed and developing regions, they are often empha- demonstrate the range of economic activities that may be affected
sised in humid areas of the tropics where ‘rainforests’ are the by change in hydrological functions. The second objective is to
dominant natural vegetation type. give the reader an idea of the degree to which these impacts
There exists another strand of conventional wisdom, which have been explored in both developed and developing countries.
concerns ecological systems that receive less rainfall, oftentimes The third objective is to summarise what this research has to say
including ecosystems where forests are not the native vegeta- about the relative magnitude and importance of these downstream
tion. Conventional wisdom emphasises the negative effects of effects, as well as noting the direction (positive or negative) of
the choice of agricultural production technology on hydrologi- the externalities identified. As will be shown, there are consider-
cal function rather than questioning the choice of land use per se. able gaps and misinterpretations in the literature. Thus, the final
In this context, the debate over the severity of the erosion problem objective, which is taken up in the next section, is to suggest
and its economic impact on productivity is complemented by the the extent to which the direction of the individual impacts can
debate over the relative magnitude of the off-site costs of erosion be generalised as increasing or decreasing with respect to land
and other surface and sub-surface water quality impacts of agri- use.
cultural land use (some of which may result indirectly from the Prior to turning to the empirical literature it is worth stating
need to fertilise eroded and degraded soils). While most of the that there are a large number of techniques available for use
evidence comes from North America, the issue clearly applies in in the valuation of non-marketed environmental goods and ser-
other regions. Although the evidence is far from conclusive, many vices. Many authors have surveyed the use of these methods in
analysts have suggested that these off-site impacts may be at least determining the user cost of soil erosion (Pierce et al., 1983;
as important as the on-site costs. Stocking, 1984; Bishop, 1992; Olson, Lal, and Norton, 1994;
Another issue receiving increased attention in the North Barbier and Bishop, 1995; Bishop, 1995; Barbier, 1998). Less fre-
American context is the growing evidence that the over- quent in the literature are surveys that include methods for use in
appropriation and abstraction of instream flows for irrigation, valuing downstream changes in hydrological function (Gregersen
urban and industrial use is having increasingly negative impacts et al., 1987; De Graaff, 1996; Aylward, 1998; Enters, 1998). For
on recreation and fish stocks. According to this view, an increase example, Gregersen et al. (1987) investigate systematically dif-
in streamflow would restore these use and existence values. The ferent aspects of hydrological function (including downstream
implicit suggestions being that altering land use and land manage- effects) and suggest appropriate valuation techniques. The tech-
ment practices so as to increase streamflow would have the same niques they consider, while perhaps still the most applicable tech-
affect as reducing water abstraction for agricultural, domestic and niques, represent only a small subset of currently available tech-
industrial uses. niques. Aylward (1998) provides a more recent survey of valuation
The earlier discussion of the hydrological impact of land use methods and identifies those applicable to the valuation of hydro-
change noted that the conventional wisdom regarding the rela- logical externalities.
tionship between forest conversion (and reforestation) and water
yield, seasonal flows, flooding and precipitation is often at
Valuation of water quality impacts
odds with the scientific understanding, particularly in the trop-
ics (Hamilton and King, 1983; Bruijnzeel, 2002). Much however The literature on water quality impacts is fairly well spread out
remains to be learned in this regard as many of the existing studies over developed and developing countries (see Table 7.1). The lack
have been undertaken at small scales (less than 10 km2 ) in head- of cited studies from European countries does not indicate that they
water basins and over relatively short durations, making accurate do not exist, rather it probably reflects the reliance in this review
extrapolation and ‘upscaling’ difficult (Bonell, pers. com.). More- on English language sources, primarily those from the United
over, the net economic effect of land use change in a given circum- States. At the same time, applied work in natural resource and
stance will depend not only on the land use and hydrological func- environmental economics has a longer history in United States
tion relationship but also the direction of the relationship between universities than in their European counterparts.
hydrological change and economic welfare. Accurate identifica- The bulk of the literature on water quality impacts in both devel-
tion, quantification and valuation of the hydrological externalities oped and developing countries surrounds the off-site effects of
associated with land use change is complicated further by the erosion, otherwise referred to as ‘sedimentation.’ This literature
need to consider both a range of potential changes in hydrological is reviewed first before assessing what material is available regard-
function and a series of potential economic impacts that may be ing the effects of nutrient and chemical outflows.
associated with a given hydrological function. Studies of externalities associated with sedimentation are found
Below, a review of the available literature on these topics is in the literature on tropical moist forests and temperate agricultural
undertaken with four objectives in mind. The first objective is to production systems. The specific economic activities examined
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 105

Table 7.1. Summary of valuation literature on water quality

Region Country Sourcea

Africa Cameroon Ruitenbeek (1990)

Morocco Brooks et al. (1982)
Latin America Chile Alvarez et al. (1996)
Costa Rica Quesada-Mateo (1979; Duisberg (1980); Rodrı́guez (1989); CCT and
CINPE (1995); Aylward (1998)
Dominican Republic Veloz et al. (1985); Santos (1992); Ledesma (1996)
Ecuador Southgate and Macke (1989)
Asia Indonesia Magrath and Arens (1989); De Graaff (1996)
Lao PDR White (1994)
Malaysia Mohd Shahwahid et al. (1997)
Panama Intercarib S.A. and Nathan Associates (1996)
Philippines Briones (1986); Cruz et al. (1988); Hodgson and Dixon (1988)
Sri Lanka Gunatilake and Gopalakrishnan (1999)
Thailand Johnson and Kolavalli (1984); Enters (1995)
North America Canada Fox and Dickson (1990)
United States of America Guntermann et al. (1975); Kim (1984); Clark (1985a); Duda (1985);
Forster and Abrahim (1985); Crowder (1987); Forster et al. (1987);
Holmes (1988); Ralston and Park (1989); Hitzhusen (1992);
Pimentel et al. (1995)

a
These studies include a number that are summary studies in the sense that they report on results obtained by other
researchers.

and type of values estimated by these studies are summarised (6) The increase in costs of water treatment associated with sed-
below:3 imentation (CCT and CINPE, 1995; Forster et al., 1987; Fox
and Dickson, 1990; Gunatilake and Gopalakrishnan, 1999;
Holmes, 1988).
(1) The loss of hydroelectric power generation due to sedimenta-
(7) The increasing dredging costs associated with harbour silta-
tion of reservoirs (Aylward, 1998; Briones, 1986; Cruz, Fran-
tion (Magrath and Arens, 1989).
cisco and Conway, 1988; De Graaff, 1996; Duisberg, 1980;
(8) The loss in production due to the effects of sedimentation
Gunatilake and Gopalakrishnan, 1999; Ledesma, 1996;
on subsistence or commercial fisheries (Hodgson and Dixon,
Magrath and Arens, 1989; Quesada-Mateo, 1979; Rodrı́guez,
1988; Gunatilake and Gopalakrishnan, 1999; Johnson, 1984;
1989; Santos, 1992; Southgate and Macke, 1989; Veloz et al.,
Ruitenbeek, 1990).
1985).
(9) The loss of tourism revenues or recreational benefits (includ-
(2) The loss of irrigation production due to sedimentation
ing fishing) following sedimentation of water systems (Fox
of reservoirs (Briones, 1986; Brooks et al., 1982; Cruz,
and Dickson, 1990; Hodgson and Dixon, 1988; Ralston and
Francisco and Conway, 1988; De Graaff, 1996; Magrath and
Park, 1989).
Arens, 1989).
(10) The loss of hydroelectric power production and increased
(3) The loss of flood control benefits due to sedimentation of
dredging costs associated with sedimentation of settling
reservoirs (De Graaff, 1996).
ponds (Mohd Shahwahid et al., 1997)
(4) The increase in operation and maintenance costs incurred
(11) The loss of navigation opportunities associated with sedi-
by sedimentation of drainage ditches and irrigation canals
mentation of water supply reservoirs used to supply water to
(Alvarez et al., 1996; Brooks et al., 1982; Forster and
canal locks (Intercarib S.A. and Nathan Associates, 1996).
Abrahim, 1985; Fox and Dickson, 1990; Gunatilake and
Gopalakrishnan,1999; Kim, 1984; Magrath and Arens, In the most comprehensive examination of the off-site costs of
1989). erosion in the United States to date, Clark (1985a) identifies the
(5) The increase in dredging and maintenance costs associated
with sedimentation of hydroelectric reservoirs (Rodrı́guez, 3 Studies that merely present the results of other studies or aggregate them
1989; Southgate and Macke, 1989). are not included in this list.
106 B . AY LWA R D

full range of economic impacts that eroding soils may cause. Of in this chapter where the concern is with nutrient and chemical
these impacts, a number are missing from the list above includ- outflows related to a change in vegetation accompanying a change
ing: impact of sediment on biological systems, lake clean-up, in land use.
damage caused by sediment in floods and damage caused to pro- Nonetheless, Clark’s estimates serve the purpose of dramatis-
ductive activities and consumption by residual sedimentation in ing the potential magnitude of the off-site damage caused by soil
end use water supplies. Thus, even a single hydrological out- erosion. Clark’s compilation also suggests that the literature on
put, sedimentation, may cause an enormous number of external the topic as reported on in this chapter is but a representative
effects. sample of a much larger literature. However, it must be acknowl-
The results of these studies confirm the intuition that in gen- edged that the quality of a majority of the studies drawn upon by
eral utility will be a decreasing function of sedimentation and, Clark and, indeed, of those gathered for this chapter, is mediocre.
consequently, that utility will be a decreasing function of land Holmes (1988) summarises this criticism by stating that the Clark
use. In other words, land use change that increasingly modifies (1985a) study ‘is based to a large degree on ad hoc interpretation
natural vegetation can be expected to produce negative hydro- of a widely divergent group of studies.’ The majority of these
logical externalities. A dissenting voice on this topic is that of studies rely on simple damage function estimates of changes in
Enters (1995) who cautions that sedimentation may also confer costs or revenues, with no consideration of optimising behaviour
benefits and not just costs on society. This claim is based on the on the part of consumers and producers as reflected in supply and
author’s observation that illegal dredging of deposited sediment in demand curves.
the Ping River, Thailand, demonstrates positive externalities asso- Interestingly, Holmes’ (1988) more sophisticated study of the
ciated with sedimentation. It has also been noted that erosion and nationwide costs of soil erosion to the water treatment industry
sediment transport lead to increased soil fertility on footslopes produces a range of US$35 million to US$661 million per year.
(van Noordwijk et al., 1998; Malmer et al. this volume). Still, This range is close to that provided by Clark (1985a) of from
these benefits are likely to simply reduce the net negative effect US$50 to US$500 million, even though Holmes’ best estimate of
of sediment rather than suggesting that sedimentation impacts are US$353 million is three times larger than Clark’s best estimate of
positive on net. US$100 million. At the same time, it must be acknowledged that
These observations are complemented by noting that in many despite the sophistication in methods, the large range obtained by
river systems (e.g. the Nile, the Senegal, the Mekong) natural Holmes indicates continued uncertainty over the true magnitude
flooding and sedimentation historically played vital roles in the of these sorts of damage estimates.
renewal of soil fertility in floodplain and recession agriculture Clearly much work remains to be done in refining such esti-
systems, as well as the renewal of geomorphological processes in mates. In particular, one difficulty of many of these studies is that
delta ecosystems. The loss of these downstream services due to they simply measure existing damage levels and do not consider to
the construction of dams or their confinement to river channels by what extent these damages could be mitigated by alternative land
levees has now led to interest in the possibility of re-establishing uses or production technologies. Nor do they subsequently assess
natural flood regimes and instream flows artificially so as to restore the trade-off between alternatives and the existing situation. This
the benefits of sedimentation. At a larger, basin-scale then, the may be an important point as even improved technologies will
issue of costs and benefits of natural and accelerated erosion and produce some erosion and sedimentation. Of course, oftentimes
sedimentation requires a careful assessment. an understanding of how damage relates to different sediment
A number of the studies demonstrate significant external effects. levels is missing from the studies as well, making it difficult to
For the United States, Clark (1985a) gathers related research on understand the form of the relationship and how it might be altered
practically every conceivable off-site impact of eroding soils and by partial reductions in sedimentation rates. The application of a
provides a nationwide estimate of the annual monetary damage damage function approach that evaluates the choice between the
caused by soil erosion of US$6.1 billion (in 1985). Even so, Clark option to undertake conservation and postpone the decision may
concludes that this figure may be severely under-estimated as the be worth investigating in this regard (Walker, 1982).
impact of erosion on biological systems and subsequently on eco- In sum, it is likely that substantial off-site damages are caused
nomic production and consumption is not included. At the same by soil erosion due to agricultural production in the United States
time it should be acknowledged that Clark includes in his analysis and similar areas around the world. Whether the claim is accurate
the effects of ‘erosion-associated’ contaminants. In other words, that these damages are as big as, if not larger than, the on-farm
the figures relate to water quality more generally, not simply the impacts is probably a moot point, given that the estimates of on-
effects of soil erosion, and include the effects of pesticides and farm losses are just as debatable as the off-site losses on method-
fertilisers that are used in agricultural production. This of course ological grounds. For example, Crosson (1995) elegantly rebuts
goes beyond the scope of the hydrological externalities envisioned the exaggerated claims made by Pimentel et al. (1995) regarding
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 107

on-site productivity losses due to soil erosion. What is probably would have essentially no effect on water supply and would lead
more important to evaluate is whether off-site damages are impor- to only a minimal disturbance of hydropower generation through
tant enough to merit action, a point that is often disregarded by sedimentation of the settling ponds. In other words, the gains
the literature. To be fair, however, it may be difficult to generalise from logging could easily compensate for the losses incurred by
due to the site-specific nature of the biophysical and economic the hydroelectricity producer due to sedimentation. Finally, in Sri
relationships involved. Lanka a comparison of measures for preventing or mitigating the
In tropical regions, many of the studies are more explicit in impact of sedimentation on the Mahaweli reservoirs suggested
targeting land use per se as the cause of hydrological externali- that the costs of the measures outweighed their potential benefit
ties, particularly the conversion of tropical forests to other uses. (Gunatilake and Gopalakrishnan, 1999).
A number of these studies even go so far as to include damage In sum, the results are mixed on the magnitude of the eco-
estimates into cost-benefit analyses in order to demonstrate the nomic impact of sedimentation as caused by the conversion and
need for changes in policies affecting land use or to justify con- modification of tropical forests. Such a conclusion is not counter
servation projects. For example, in Ruitenbeek’s valuation of the intuitive as it is logical to expect that site-specific characteristics
Korup Project in Cameroon, the benefits from erosion control such as geology and climate, drainage area and topography, type
were estimated to be almost half of the direct conservation bene- and size of reservoir or other infrastructure, and demand for end
fits of conserving the forest, benefits which outweighed the sum use goods and services, will determine the magnitude of these
of the direct and opportunity costs of conservation (Ruitenbeek, effects in particular cases. In addition, it must be said that many
1990). Santos (1992), Southgate and Macke (1989), and Veloz of these studies present only fairly crude estimates, just as in the
et al. (1985) all suggest that sedimentation will have significant case with the studies from developed countries.
effects on hydroelectric power plants in Latin America and the Turning briefly to water quality issues beyond merely the off-
Caribbean. site effect of erosion, no studies were found in the developing
Nevertheless, there is an additional series of studies demonstrat- country literature that specifically assess the downstream exter-
ing that the externalities associated with sedimentation are often nalities associated with nutrient or chemical outflows associated
not terribly large or important. In the Philippines, the effect of with land use change (though see Proctor, Connolly and Pearson,
sedimentation derived from the conversion of large areas to open this volume, for more on the biogeochemical impacts). In a devel-
grasslands in the Magat Basin on the length of life of the reservoir oped country context, there are of course many studies of the eco-
downstream was valued at 0.10 Pesos ha−1 yr−1 , or under one US nomic damage caused by poor water quality (Bouwes, 1979; Epp
cent per hectare per year (Cruz et al., 1988). Meanwhile the ben- and Al-Ani, 1979; Young, 1984; Ribaudo, Young, and Shortle,
efits of erosion control through reforestation in the Panama Canal 1986; Lant and Mullens, 1991). Typically these studies are not
Zone comes to a present value of just $9 ha−1 in terms of its affect linked to land use in specific geographical areas, nor do they eval-
on storage reservoirs and water supply for navigation (Intercarib uate damage that is directly and only related to land use change.
S.A. and Nathan Associates, 1996). Oftentimes the measure of water quality that can actually be eval-
In Arenal, Costa Rica, the present value of the cost of sedi- uated (as perceived by recreationalists, for example) is extremely
mentation from pasture (as opposed to reforestation) in terms of crude (i.e. water quality is good or bad), so that associating the
lost hydroelectric production ranged from US$35 to US$75/ha measure of damage with a particular type of non-point source pol-
(Aylward, 1998). The Arenal study is unusual in that it employed lution is impossible. These are precisely the ‘erosion-associated’
a formal model of the impact of sedimentation on both the dead contaminants surveyed by Clark. Clearly these (gross) impacts are
and live storage areas of the reservoir, enabling it to separate out important and perhaps particularly so in the case of the biological
the differential effects on these areas. Given the large dead storage impacts that Clark does not estimate. The extent to which they
relative to sediment inflow for this particular reservoir the effect are associated with land use per se and not simply the prevalence
of sedimentation on dead storage produced benefits, not costs, of pesticide and fertiliser use as part of a production technology
in the case of Arenal as the sediment effectively displaces water package is difficult to assess.
upwards into the live storage during dry periods. Arenal is an inter-
annual regulation reservoir and thus during a series of dry years in
Valuation of water quantity impacts
which the reservoir does not fill but is gradually drawn-down, the
sediment occupying the dead storage effectively makes additional The external effects of land use change on streamflow levels
water available for power generation (Aylward, 1998). will affect four types of hydrological outputs: (1) annual water
In Malaysia, a simulation of the effect of logging on down- yield, (2) seasonal flows, (3) peakflow and (4) groundwater lev-
stream run-of-stream hydroelectric power and treated water pro- els (Gregersen et al., 1987). These outputs will in turn affect
duction indicated that a programme of reduced impact logging a host of different economic activities, including most of those
108 B . AY LWA R D

Table 7.2. Summary of valuation literature on water quantity

Region Country Function valued Source

Latin America Bolivia Flood control Groundwater recharge Richards (1997)

Costa Rica Dry season flow Peak flows Quesada-Mateo (1979)
Annual water yield Aylward (1998)
Dry season flowa
Guatemala Dry season flow Brown et al. (1996)
Panama Dry season flow Intercarib S.A. and Nathan
Associates (1996)
Africa Cameroon Flood control Ruitenbeek (1990)
South Africa Annual water yield de Wit et al. (2000)
Annual water yield de Wit et al., forthcoming
Asia China Annual water yield Guo et al. (2001)
Indonesia Annual baseflow Pattanayak and Kramer (2001a, b)
Malaysia Dry season flow Kumari (1995)
Thailand Dry season flow Vincent and Kaosa-ard (1995)
Temperate countries Australia Annual water yield Creedy and Wurzbacher (2001)
United Kingdom Annual water yield Barrow et al. (1986)
United States of America Annual water yield Kim (1984)

a
Sensitivity analysis only.

affected by water quality changes. An increase in water yield of the changes is insignificant or simply represents an ignorance
or baseflow will change reservoir storage and irrigation capac- of the biophysical impacts of land use change on water yield is
ity leading to changes in water supply for hydropower, irriga- unclear. As an indication that this situation is changing, 12 of the
tion, navigation, recreation, etc. Similarly, changes in water yield 16 studies were published since 1995. Interestingly, seven of the
and baseflow may affect these activities directly in the absence studies considered water quantity issues but did not raise the issue
of hydrostorage capacity in the system. Changes in peakflows of water quality (Barrow et al., 1986; Brown et al., 1996; Guo
are felt principally through a change in localised flood frequency et al., 2001; Pattanayak and Kramer, 2001a, b; Richards, 1997;
and can damage infrastructure (bridges, culverts, roads, embank- Vincent et al., 1995).
ments) and agriculture (sedimentation of crop land with coarse An additional avenue of research, primarily in a developed
material), as well as putting homes and lives at risk. Changes in country context, concerns the valuation of increases in instream
the groundwater table in upland areas will influence directly spring flows. A number of studies have examined the recreation, fishery
discharges used for local water supply and have downstream and hydroelectric power benefits that would be gained by restoring
impacts on the productivity of local biological systems (such as instream flows in the western United States (Daubert and Young,
wetlands) that may provide recreational or preservation benefits, 1981; Narayanan, 1986; Ward, 1987; Johnson and Adams, 1988;
as well as affecting downstream agricultural and other productive Brown, Taylor, and Shelby, 1992; Duffield, Neher, and Brown,
systems. 1992). Once again, these studies are not linked directly to land
The methods that may be applied in valuing such external effects use, but could be used to indicate the economic benefits associ-
are essentially no different than those in the case of water quality. ated with land use change that subsequently alters streamflow.
Nonetheless the literature on this topic is scanty in comparison
to that on water quality effects. Just 13 studies were found in A N N UA L WAT E R Y I E L D
comparison to the 34 studies of sediment. The countries for which Of the seven studies on annual water yield reviewed here, five sug-
such studies were found are listed in Table 7.2. gest that catchment protection values are negative, i.e. that utility
Of the studies that examined the off-site costs of sedimenta- is increasing as a function of land use. In the earliest study of this
tion, only five considered the attendant issue of water quantity nature, Kim (1984) simulates the increase in annual water yield
(Aylward, 1998; Intercarib S.A. and Nathan Associates, 1996; associated with a change in land use management from no grazing
Kim, 1984; Quesada-Mateo, 1979; Ruitenbeek, 1990). Indeed, to grazing in the Lucky Hills catchment of southeastern Arizona.
such impacts were rarely, if ever, even identified and listed in qual- Based on a review of the literature Kim (1984) assumes a 30%
itative terms. Whether this is due to a suspicion that the magnitude increase in water yield under grazing over a simulated 50-year
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 109

rainfall cycle (based on climatic records). Under the additional The South African study by De Wit et al. (2000) examines
assumption that all the extra water would be used for irrigated issues related to the catchment management charge (approxi-
agriculture and employing a US$1.2 m−3 value for irrigation mately US$1 ha−1 yr−1 ) that is to be levied on forestry activities
water based on studies from the region, Kim calculates the net as Stream Flow Reduction Activities under existing legislation.
present value over the 50 years to be US$342 at a 7% discount Combining information from detailed hydrological studies of the
rate. Unfortunately, it is not clear if this is the catchment total or a effect of forestry on evapotranspiration, the authors’ calculate that
per acre figure. Assuming the former this comes out to a little over forestry consumes 7% of South Africa’s water (see also Scott et al.,
US$7 ha−1 for the 44-hectare catchment. When Kim adds in the this volume). Collation of macroeconomic data on value added in
costs of excavating the sediment settling ponds (US$1068) and the forestry suggests that the value added per cubic metre in forestry
benefits of animal weight gain (US$740), the net present value of is low (2.8 Rand or about US$0.50) but still higher than irrigation.
the returns to the land use management change are barely positive De Wit et al. (2000) use an input-output model to confirm that
at $14 or about US$0.25 ha−1 . due to the existence of higher value uses for water (than forestry)
A study of the effects of afforestation on hydroelectricity gener- such changes lead to economy-wide gains in output. In a related
ation in the Maentwrog catchment in Wales and forty-one catch- study De Wit (forthcoming) calculates the present value cost of
ments in Scotland by Barrow et al. (1986) indicates that the water consumed by black wattle (Acacia mearnsii) in South Africa
increased evaporation under reforestation (in comparison with as US$1.4 billion using information on the difference between
grazing) lead financially marginal sites (for forestry) to become streamflow and value added of black wattle as versus alternative
financially sub-marginal once hydropower losses were included land uses.
into the analysis. While there was some variation in results depend- In a study of ecological services in Victoria, Australia a counter-
ing on site conditions, the example clearly shows the negative example to the trend shown above is provided by Creedy and
impact on productivity associated with afforestation in a hydro- Wurzbacher (2001). In this case the authors are assessing the
electric catchment. effect of harvesting old-growth Eucalypt forest. These forests have
A study in Arenal, Costa Rica, confirmed the results obtained the unique property that they transpire very little water. Thus, the
by Barrow et al. (1986) by showing that water yield losses due to effect of harvesting and allowing regrowth will lead to a decline in
reforestation of pasture areas may lead to large efficiency losses annual water yield, not an increase as would be otherwise expected
in downstream hydroelectric power production (Aylward, 1998). (Vertessy et al., 1998). Creedy and Wurzbacher (2001) do not pro-
The externalities associated with water yield effects were calcu- vide explicit value estimates in per hectare terms. However, they
lated to be one order of magnitude greater than those associated do show that given the projected costs of alternatives sources of
with the sedimentation costs (as already referred to above). Best water to the public utility, incorporating the loss of water benefits
estimates for both cloud and non-cloud forest areas suggested alongside the wood benefits of logging leads to an infinite length
positive present values in the range of US$250 to US$1100 ha−1 of the optimal rotation. In other words logging is not worth the
for pasture. Sensitivity analysis showed that the values will be costs it incurs in terms of forgone water supply.
reduced to two-thirds of these figures with higher discount rates In examining the value of ecosystem services in Xingshan
and in the event that all the water yield gain under pasture were County of Hubei Province, (north-eastern) China a study by Guo
to arrive during the wet season (instead of being received pro- et al. (2001) purports to value the water conservation value of
portionately across wet and dry seasons). The values may also forests in terms of ‘hydrological flow regulation’ and ‘water reten-
rise to almost US$5000 ha−1 if dry periods lengthen or occur tion and storage’. However, all the figures employed in the study
early in the seventy-year simulation period. Further sensitivity are annual, thus it can only be concluded that this is a study of
analysis examined what would be the economic outcome if refor- annual water yield. Unfortunately, the authors’ definition of for-
estation resulted in net gains in dry season flow, in spite of the est hydrological function is confused, leaving out transpiration
expected overall losses in total annual water yield. A switching and defining canopy interception as one of the elements of rain-
value (where the value of total hydrological externalities go to water ‘conserved’ by a forest ecosystem. The author’s empirical
zero) was obtained only when all of the annual water yield gain analysis concludes that in comparison with a scenario of forest
and an amount equal to an additional 50% of this amount was conversion to shrub and grass the forest alternative ‘conserves’
redistributed to arrive during the wet season (when water is less such large amounts of water that 42% of the value of downstream
valuable for power generation). When the analysis of livestock hydroelectric production is due to the conservation of forest. This
productivity was incorporated into a cost-benefit analysis of land study only serves to illustrate how inadequate hydrological anal-
use options, strong synergies between livestock production and ysis and simplistic applications of economic valuation can lead to
hydroelectric power generation in the catchment were demon- gross exaggerations of hydrological externalities (see also Cheng,
strated (Aylward and Echeverria, 2001). 1999).
110 B . AY LWA R D

F L O O D C O N T RO L storage, as well as a change in stormflow regime. To the detriment

The remainder of the literature that was surveyed portrays utility of the analysis, the author assumes erroneously that the removal
as a decreasing function of land use. Ruitenbeek (1990) estimates of forest cover will lead to an increase in peak flow during the wet
the flood control benefits to be generated by protecting forested season and a decrease in baseflow during the dry season. While
catchments in Korup National Park in Cameroon. Ruitenbeek’s the first assumption is likely to be correct, the latter does not
calculation is based on the share of local income that would be lost necessarily follow.
in a flood event multiplied by the percentage of cleared forest area
in the Park. As reported by Bonell and Elsenbeer (this volume), the D RY S E A S O N F L OW A N D G RO U N DWAT E R
hydrological literature does not support definitively the contention S T O R AG E : H Y D RO L O G I C A L A NA LY S I S
that land use change would lead to changes in flood frequency or Eight studies were found that attempt to quantify the purported
magnitude at the scale suggested by Ruitenbeek, and thus the benefits provided by forest cover in terms of enhanced ground-
results must be regarded as suspect until proven otherwise. water storage and subsequent dry season baseflow. All but the
Richards (1997) examines the potential benefits of a flood con- Quesada-Mateo (1979) study (reviewed above) are recent in ori-
trol programme in the Taquina catchment in the Bolivian high- gin and most of the studies suffer from the same problem, namely
lands. The approach taken is more data intensive than that by difficulty with the direction and magnitude of the land use and
Ruitenbeek, insofar as the costs of damage from a recent flood are hydrological relationship. As irrigated agriculture and navigation
actually gathered to motivate the damage cost estimate. Assump- canals will clearly benefit from an increase in dry season baseflow
tions regarding flood frequency and intensity are then made under there is little doubt that the relationship between the hydrological
the ‘with’ and ‘without’ project cases, accounting for a gradual outputs (dry season baseflow) and economic activities is increas-
phase-in of project benefits. Straight multiplication is then used ing. However, if the direction or magnitude of the land use and
to arrive at yearly flood control benefits as the difference between hydrology relationship is misstated, the overall conclusions of the
the ‘with’ and ‘without’ project scenarios. By year five, the nom- studies regarding the hydrological externalities would be erro-
inal flood control benefits outweigh the project costs by a ratio of neous. As this concern is central to the interpretation of the results
3:1.4 While the benefits of flood control appear quite large, it is obtained by these studies, the hydrological analyses are explored
not clear to what degree they are a response to land use change below at some length.
in terms of on-farm soil conservation technologies as opposed to In the Sierra de las Minas Biosphere Reserve of Guatemala
the effect of hydraulic works and infrastructure located in gullies a comparison between dry season baseflow in a forested and a
and stream courses. partially cleared catchment was used to estimate the percentage
Interestingly, neither of the two studies mentioned above increase in baseflow associated with a forested catchment (Brown
attempts to apply the standard methodology for evaluating flood et al. 1996). Unfortunately, study limitations implied that only
damages as recounted by Gregersen et al. (1987). Under this four months of dry season data from 1996 were compared. As
methodology flood frequency curves (the probability that a given the two catchments were not calibrated prior to the change in
instantaneous streamflow level or stage height will be exceeded) land use it is not possible to rule out the possibility that the effect
are developed for the ‘with’ and ‘without’ project scenarios. A observed is a result of some other situational variable and not
damage function is then developed that relates peakflow levels to land use. For example, the forested catchment faces southeast and
damage costs. A practical difficulty in applying this technique in sits at an altitude of 1900–2400 metres. The cleared catchment
developing economies is the poor availability of historical data on faces southwest, is located some ten kilometers to the west of the
the damages of past flood events. This problem is exacerbated by forested catchment and sits at an altitude of 1400–2120 metres.
rapid urbanisation, industrialisation and population growth, which The forested catchment is known to be a cloud forest area and the
make the relationship between peakflow levels and damage costs study concerned reports on the capture of horizontal precipitation
unreliable over time. A further limitation of flood frequency analy- during the dry season in this catchment. Given the lack of calibra-
sis from a hydrological standpoint is that it rests on the assumption tion the higher level of baseflow in the forested catchment may
of stationarity: the analysis ignores changes in river and stream simply be attributable to climatic conditions such as the presence
discharge linked to climate variability and land use change over of cloud forest moisture or rainfall levels and not only to conver-
time. sion of the other catchment.5 Bruijnzeel (this volume) also notes
Quesada-Mateo (1979) develops a deterministic simulation
model that enables the user to determine the maximum amount of 4 The study does not give the present value of flood control but only the project
firm (reliable) power that could be produced from a hydroelectric internal rate of return.
5 The authors also do not provide data on yearly rainfall totals in the two
reservoir. What makes the model interesting is that it incorporates catchments, but indicate that rainfall levels will vary with elevation and that
explicitly the effect of the accumulation of sediment in the live at high elevations precipitation may vary greatly within short distances.
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 111

that the two catchments are of different size, which may also affect r climatological factors do not explain the reduction in water
baseflow levels. To make matters even more difficult, the cleared yield.
catchment is not in the basin in which the impact of baseflow
changes is valued, while the forested catchment is within one of The land use change that took place in Mae Teng during the
these basins. 1972 to 1991 period consisted of both an increase in irrigated
Brown et al. (1996) also note that high values for the capture of agriculture and an expansion of pine forestry plantations. As both
fog moisture were observed only in an elevation zone that occu- of these activities can be expected to increase water use, the authors
pied a very slight percent of the total catchment area and that the conclude that land use change has indeed led to the reduction in
lower catchment was well below this zone. Despite the intuition, water yield, particularly during the dry season. Unfortunately, the
then, that the existence of forest will serve to strip moisture from authors are unable to define clearly to what extent the conversion
clouds in the dry season thus adding to dry season baseflow as of land to agriculture, the use of water in irrigation or the growth
compared to a scenario in which forest conversion occurs, the of pine plantations were responsible for the observed decrease in
simulations undertaken in the study are not very well supported streamflow.
by the hydrological analysis. Pattanayak and Kramer (2001a, b) value ‘drought mitigation’
The study of the Panama Canal Basin relies on a similar ‘paired’ in a large number of catchments that lie below the Ruteng Park, on
catchment analysis that does not have an experimental basis (i.e. the island of Flores in eastern Indonesia. In the longer of the two
calibration followed by treatment) (Intercarib and Nathan Asso- papers, the authors estimate an explicit hydro-economic model of
ciates, 1996). Nevertheless, the data are more convincing as the how changes in baseflow lead to changes in profits received by
monthly streamflow for six forested and cleared catchments (three farmers from crops (Pattanayak and Kramer, 2001b). In the other
each) compared are based on 21 years of data. The data reveal that paper, the authors explore what farmers would be willing to pay to
monthly streamflow measured as a percent of total precipitation is obtain ‘drought mitigation’ services from forest areas in the Park
less responsive in the case of the forested catchments. The authors (Pattanayak and Kramer, 2001a).
use this information to substantiate the claim that land that remains The authors cite a number of sources as providing evidence
in forest stores a larger amount of water going into the dry sea- that forest in the Park plays a role in drought mitigation, with one
son. This capacity is then available to refill the dams that release consultancy report explicitly cited as claiming higher dry season
their stored water in the dry months, thereby augmenting reduced baseflow under forest. And clearly it seems logical that more water
streamflow during these months. in the dry season would increase farm productivity and, indeed, the
Once again, the potential existence of confounding variables has willingness-to-pay survey confirms this expectation (Pattanayak
not been ruled out in the analysis. Further, as annual water yield and Kramer, 2001a). The hydrological portion of the model, how-
from a cleared catchment can be expected to rise, even a lowering ever, weakens the meaningfulness of the hydroeconomic analysis.
in monthly streamflow in percentage terms during the dry season First, the authors actually do not include dry season baseflow
does not rule out an increase in streamflow in absolute terms. In in the model, but rather total annual baseflow. That agricultural
this regard it is worth noting that the Intercarib study ignores the production is related to total water availability is not in ques-
potential decrease in water yield that would presumably result tion, however it seemed that the intent of the paper was to get at
from reforesting the cleared areas of the Canal Basin. Thus, the the marginal benefit associated with increased flows when they
study emphasises one type of hydrological change and ignores presumably matter most, that is during dry periods.
another, in addition to falling short of providing firm evidence A second difficulty encountered by the authors, however, con-
of the hydrological effect that is subsequently included in the cerns their effort to develop a quantitative linkage between for-
valuation exercise. est cover and baseflow. The authors estimate a cross-sectional
The analysis of the Mae Teng Basin in Thailand by Vincent regression equation using data from 37 catchments and a series of
et al. (1995) resolves a number of the issues encountered above explanatory variables, amongst them three for forest cover: area of
by employing historical data on streamflow and precipitation. By forest cover, percent of forest cover, and the square of percent of
analysing data from periods before and during the period of land forest cover. As the squared term produces a negative coefficient,
use change the authors strengthen their case further. The authors the end result is that the simulation of increases in forest cover in
use regression analysis to demonstrate that: the catchments leads to a mixture of expected losses and gains in
farmers profits as a result of increases in forest cover (Pattanayak
r no change in streamflow is observed prior to land use change and Kramer, 2001b). The study illustrates the importance of multi-
(1952–1972); disciplinary cooperation as poor theoretical formulation and exe-
r dry season streamflow is reduced during the period in which cution of the hydrological portion of this study undermines an
land use change occurs (1972–1991); otherwise excellent economic analysis.
112 B . AY LWA R D

Richards (1997) values the aquifer recharge benefits of the same potential in the region. Thus, should the claimed hydrological
Bolivian soil conservation programme mentioned above. Appar- effect be substantiated, the authors have demonstrated a significant
ently, the intuition is that the project will increase infiltration but hydrological externality of forest conversion in these Guatemalan
without, the project infiltration rates will fall. There appears to catchments.
be some confusion, however, as the author first misrepresents the In the Bolivian case, the economic methodology employed is
direction of water quantity effects as found in the literature and fairly simple. Unit values for water are multiplied by the changes
then states that with the project ‘runoff would be reduced by 15– in aquifer storage (Richards, 1997). Again, this linkage is not well
25%’ (Richards, 1997:26). By year 50 the author calculates that demonstrated but as presented is significant.
aquifer recharge would be 80% higher with the project than with- In the Panamanian case, the valuation hinges on the prospects
out the project. Further, although the benefits of aquifer recharge for developing a third set of locks in the Canal, at which point
under the project are considerable, there is no discussion of sea- the current water storage capability would not be sufficient (Inter-
sonality of runoff or water storage and, thus it is not clear how the carib and Nathan Associates, 1996). The benefits of water storage
change in aquifer recharge is translated into water supply benefits. offered by 132 000 ha of existing forest are estimated to be an
The last of the studies is a valuation of the hydrological function additional 1500 m3 ha yr−1 based on the hydrological analysis.
provided by peat swamp forest in Malaysia by Kumari (1995). The costs of building additional capacity are US$0.185 m−3 . The
Unfortunately, insufficient detail of the hydrological basis for the study reports water storage benefits for these existing forest areas
analysis is provided in the paper to provide an informed content as US$277 ha−3 in present value terms. The same figure is cal-
and thus cannot be analysed further here. Interestingly, however, culated for the water storage benefits of reforesting an additional
the paper does refer to a controversy over the role of forests in the 100 000 ha in the Canal Basin.
production of dry season padi rice. The study apparently uses the Polestar software to generate dif-
The studies reviewed above demonstrate the difficulty of devel- ferent scenarios for how land use determines water and sediment
oping convincing hydrological analyses of the linkages between inflows to the dams and water supply to the system of locks is
specific land uses and dry season flows. This is particularly acute modelled over a 60-year planning horizon. According to results
when the study site does not have a history of hydrological mea- presented in the study, there is an anticipated water shortage only
surement or evaluation and points to the difficulty of undertaking if the third set of locks is built, an event projected for the year
short-term policy-oriented studies where long-term hydrological 2020. Unfortunately, it is not possible to come close to the per
research or calibration of process-based models to local conditions hectare calculation using a 10% discount rate (the exact discount
is probably necessary to guarantee the reliability of results. rate employed is not cited in the document). It is however, possi-
ble to calculate the US$36 million present value attributable to the
D RY S E A S O N F L OW A N D G RO U N DWAT E R 132 000 ha of existing forest, by simply multiplying the number
S T O R AG E : E C O N O M I C A NA LY S I S of hectares by the annual water storage figure and the per unit cost
The variation in sophistication of the economic modelling of building the new dam. However, assuming that the new dam
conducted for these studies also varies tremendously. In the would not need to be built until 2020, the present value of such a
Guatemalan, Indonesian and Thai studies detailed econometric figure would be more in the region of US$3 million than US$36
analysis of agricultural production are used to estimate the change million.7 Further, it has been estimated recently that sedimenta-
in revenue that would be associated with changes in flows. In the tion levels in the Canal Basin have dropped back to background
Thai case it is not possible to link the significant (roughly 50%) levels given that land use has stabilised in the last decade (Stallard,
loss in revenues to a particular causal factor. In the Indonesian 1997). In all likelihood then the hydrological benefits of engaging
case, the weakness of the hydrological analysis undermines the in massive reforestation of the Panama Canal Basin due to both
results provided by full hydro-economic model. water storage and erosion control are substantially overstated, if
In the Guatemalan study only aggregate figures are provided, they exist at all.
not estimates of the loss in irrigated agriculture in dollar per Whether as a result of questions regarding the hydrological
hectare or net present value terms. Employing data from the report, assumptions or modelling, or the economic interpretation of these
however, the loss of revenue accruing to the area in the two catch- relationships, the results of the Bolivian, Guatemalan, Indone-
ments that would be cleared under the simulation can be calcu- sian and Panamanian studies examined above must be regarded
lated to be US$7.5 ha−1 yr−1 and US$47 ha−1 yr−1 .6 If the effect
6 On page ii the percentage of remaining forest area that is cleared under the
is assumed to continue indefinitely and the money flows are con- simulation is presented and on pages 69 for the Jones catchment and page
verted into present value terms at a 10% discount rate, the figures 80 for the Hato catchment the remaining forest areas can be derived from
land use and area data (Brown et al., 1996).
may be multiplied tenfold to obtain approximate present values
7 Current intentions in Panama greatly exceed such marginal changes with
of from US$75 to US$470 ha−1 . Such values would be low to plans to build a series of three dams in order to double the water supply to
respectable values for land of presumably marginal productive the Canal by approximately 2010.
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 113

as highly questionable. The cautious stance taken by the Thai interannually while having little to no impact on a downstream
study simply reflects the inherent difficulties in undertaking such water treatment plant that is fed from such a reservoir. In other
an integrated hydrological and economic analysis of dry season words, profits (and eventually utility) may be either an increas-
flows. ing or decreasing function of these hydrological outputs and of
land use itself. This result is clearly at odds with the conventional
wisdom on the effects of changes in water quantity on productive
T H E D I R E C T I O N O F H Y D RO L O G I C A L activities.
E X T E R NA L I T I E S The situation with regard to consumptive values of water quan-
tity in developed countries is somewhat clearer. On the one hand,
The effects of changes in hydrological outputs on economic con- in cases where streamflow is already greatly diminished or altered
sumption and production will vary with different types of hydro- (for example due to abstraction, dams or levees), the benefits to
logical function and types of economic activities. For instance, an recreation activities of increases in these flows are clear. How-
additional unit of baseflow into an irrigation scheme during the dry ever, the restoration of original vegetation cover in the catchment
season will lead to additional output by raising water availability may only provoke a worsening of the situation if it means replac-
during a critical period. If baseflow is an increasing function of ing shallow-rooted vegetation (crops) with deep-rooted vegeta-
land use then the relationship between land use and agricultural tion (forest). A further consideration is that the extent to which
production will be increasing. On the other hand a rise in sedi- developed country consumers actually are aware of the nature
mentation of the irrigation canals will be associated with either a of original streamflow conditions is debatable, given the large
loss in production as the sediment impairs the ability of the canal modifications and extractions already made to most waterways in
to deliver water or an increase in, for example, labour expended developed countries. Thus, although a change back to the origi-
on dredging. In this case then, production will be a decreasing nal land use would alter the status quo, it is not clear that such a
function of land use. change would produce perceived improvement in aesthetic values.
In general, an increase in sedimentation, nutrification or leach- In other words the direct effects of land use change on utility as
ing can be expected to impact negatively the profits from activities experienced through hydrological functions may not be terribly
such as irrigation, hydroelectric power generation, water treatment large, nor may utility necessarily be a decreasing function of land
and navigation. Similarly, the effects of increases in these outputs use for these functions. Again, much will depend on the sever-
on developing country households may be negative. However, it ity of the problem posed by current streamflow and hydrological
is at least conceivable that on occasion they may also have posi- conditions at the site.
tive elements, as in the case in South East Asia where sediment is An added difficulty to the process of unravelling the implication
actually harvested (Enters, 1995; van Noordwijk, 1998). The aug- of downstream hydrological change is that a single hydrological
mentation of natural processes of renewing soil fertility cannot be output may affect a series of productive or consumptive activities.
assumed to be negative. In addition, it should be noted that there A study in the Philippines demonstrates that logging of a coastal
is no general intuition that requires a given change in chemical catchment may lead to an increase in sedimentation of a coral
or nutrient outflows to have a negative impact on the household. reef downstream (Hodgson and Dixon, 1988). This sedimentation
Much will depend on how ideal the starting point is with respect subsequently has negative effects on both coral cover (biomass
to desired water quality characteristics and what thresholds or production) and coral diversity. As coral cover and diversity are
discontinuities exist in the relationship. Finally, it is reasonably assumed implicitly by the authors to enter into an ecotourism pro-
clear that reduction in water quality of waterways and lakes has duction function, the knock-on effect of the change in hydrology
a negative impact on recreation opportunities. In other words, the is negative. At the same time the loss in coral cover has a negative
conventional wisdom with regard to the sign of the water quality impact on the biological production function for fish in the area.
effect is likely to be correct, though questions remain regarding Fish in turn are a key input in the fishing production function,
the magnitude of the problem. which is also affected adversely by the logging and subsequent
The case with the different measures of water quantity is much change in catchment hydrology.
less certain and will depend on the hydrological functions that This example demonstrates the need to clearly specify the intri-
are germane to the production technology and end use demand. cate relationships that may exist between the outputs of the hydro-
For example, an increase in land use that leads to soil compaction logical production function and their subsequent impacts on eco-
and an increase in peakflows will affect profits adversely from nomic production functions. This impact may occur directly, as
a run-of-stream hydroelectric plant, whilst having no affect on inputs into economic production functions, or indirectly, as inputs
an annual storage reservoir used for irrigation, hydroelectricity affecting other biophysical production functions that subsequently
or navigation control. An increase in annual water yield may produce another level of outputs that in turn enter an economic
raise profits for a large hydroelectric reservoir that stores water production function (i.e. turbidity impacts on fish that are the
114 B . AY LWA R D

object of fisheries production). It is also the case that a single eco-

Box 7.1 Direction of hydrological externalities: an example
nomic production function may be affected in different ways by a
from Arenal, Costa Rica
number of hydrological outputs that are linked to a given land use
change. Aylward (1998) valued the hydrological consequences of land use
In sum, although hydrological function is more often than not practices in the Rio Chiquito catchment in Arenal, Costa Rica.
Lake Arenal is an enlarged natural lake found at the headwaters
an increasing function of land use (interpreting an increase in land
of the Arenal river that is of national importance to Costa Rica for
use as modification of original vegetation and intensification of
the production of electricity. Given that annual inflow is normally
land use), there may also be cases where it is a decreasing function
less than the live storage capacity, technical production levels are
of land use. On the other hand, utility (whether affected directly limited by water availability rather than by production capacity
or indirectly) may be either an increasing or decreasing function (362 MW) and the facility is operated as an interannual storage
of hydrological function. Increases in land use that lead to an reservoir to buffer the national grid during dry years.
increase in sedimentation, nutrification and leaching will generally Located on the Atlantic side of the continental divide in Costa
be related negatively to utility. Similarly, increases in peakflows Rica, the Rio Chiquito catchment occupies 8900 ha and makes up
that lead to increased and localised flooding may affect utility approximately one-fourth of the entire drainage system for Lake
negatively. However, increases in land use that lead to increase in Arenal. The dominant characteristics of the catchment are steep
downstream annual water or increased dry season baseflow will slopes with abrupt ridges and valleys with 90% of the area on
be related positively to utility. Thus, while in many cases utility slopes greater than 25%. Elevation ranges from lake level at 545 m
up to 1800 m. Four of the Holdridge life zones are present in Rio
will be a decreasing function of land use it will by no means be
Chiquito including Wet Premontane Forest, Premontane Rainfor-
the rule.
est, Lower Montane Rainforest and Wet Lower Montane Forest.
As a single land-use change will cause a series of changes in util-
The main categories of land cover in the Rio Chiquito catchment are
ity it is clear that the net impact will reflect the trade-off between pasture and forest with minimal amounts of agriculture. Between
the different functions and value changes that result. Added to the 1960 and 1992, pasture areas more than quadrupled, while the
complexity of understanding the net result is that an individual area under primary forest was cut in half, from almost 80% to just
hydrological output (for example sediment) may affect a number under 40%.
of economic activities (for example recreation and hydropower Three aspects of the change in hydrological function expected
production) and a given activity (for example hydropower pro- from pasture areas (as opposed to reforestation) were valued explic-
duction) may be affected by changes in a number of hydrolog- itly in the analysis, based on the scientific literature and available
ical outputs (for example sediment, water yield, and dry season hydrological studies and data for the site. These included the effect
flow). of sediment on the dead and live storage volumes of Lake Arenal
and the impact of a change in water yield on water inflows. In addi-
Thus, given the nature of hydrological function and the range
tion, sensitivity analysis explored what would be the effect if pasture
of economic activities that depend on this function it will not be
and livestock contributed to a loss of dry season flow through soil
possible to generalise regarding the sign of hydrological external-
compaction and a reduction in infiltration opportunities.
ities. A reduction in the intensity of land ‘use’ (i.e. reforestation Table 7.3 presents the value figures for one of the land use units
of pasture) may lead to a decrease in sedimentation, subsequently included in the analysis (Wet Premontane Forest) in a sequential
improving water storage capacity for hydroelectric production. format. What the table seeks to demonstrate is that the conclusion
At the same time, however, the increase in forest cover may also as to what is the direction of the externalities – positive or negative –
lead to a decrease in water yield thereby decreasing water inflow will depend on which effects are valued explicitly. If the dead stor-
to the reservoir. Aylward (1998) traces out these linkages in pro- age impacts of sedimentation is the only effect valued then the
viding a formal model linking land use to hydropower generation hydrological externalities of pasture would be seen to be positive.
for the case of large hydroelectric reservoirs. The model illus- If the impacts on live storage are included then the conclusion would
trates the effect of a change in land use on discharge from the be that pasture represents a loss of benefits had under forest. How-
ever, once the effects of the gain in water yield on the interannual
reservoir, power production and, hence, the marginal opportunity
hydropower facility are included it can be seen that these dominate
costs of power generation. As both streamflow and sediment yield
the costs of sedimentation, once again leading to the conclusion that
functions are increasing (i.e. increase with forest conversion), but
the hydrological impacts are positive in economic terms. Finally, the
have opposing effects on discharge it cannot be assumed that for- sensitivity analysis showed that if a strong loss of dry season flows
est conversion will not be unambiguously positive or negative. was associated with pasture, then it is possible (though remote)
Summarising the results from Aylward (1998) as cited earlier, that the net effect in terms of hydrological externalities could go
Box 7.1 provides an example of how the perception of the direction to zero – i.e. neither positive nor negative. While the hydrological
(and magnitude) of hydrological externalities can vary as addi- basis for the valuation exercise could be improved, the example
tional hydrological components are incorporated into the valuation demonstrates the potential for obtaining false results if the range of
exercise. hydrological impacts considered is restricted arbitrarily.
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 115

Table 7.3.. Valuation of hydrological externalities provided by pasture (as opposed to reforestation), Arenal, Costa Rica

Total hydrological
Sedimentation effect externalities
Water effect:
Values taken into account Loss of Dead storage Loss of live storage change in water inflow Direction Magnitude

1. Dead storage impact on $6 ha−1 Positive $6 ha−1

sedimentation only
2. Both dead and live storage impacts $6 ha−1 ($80 ha−1 ) Negative ($74 ha−1 )
of sedimentation
3. Sedimentation impacts plus $6 ha−1 ($80 ha−1 ) $1149 ha−1 Positive $1075 ha−1
reduction in annual water yield
4. Sensitivity analysis of a loss in dry $6 ha−1 ($80 ha−1 ) $74 Neutral $0 ha−1
season flow under pasture (i.e. all
water yield gained under pasture
accrue during wet season and 50%
again of the amount of the water
yield gain is redistributed from dry
to wet season)

In reality, then, there will often be a number of hydrological conventional wisdom that forests ‘conserve’ water and act like a
functions (sedimentation, water yield, water regulation, etc.) that ‘sponge’ persists in the face of a good deal of empirical evidence
need to be considered in determining the net impact (direct or of cases where this does not apply. The literature on forest hydrol-
indirect) upon a range of affected economic activities. Thus, the ogy reveals that a reduction in normal vegetation levels will likely
general statement that forest provides soil and water conserva- increase annual water yield and may either raise or lower dry sea-
tion benefits, or catchment protection benefits, is disingenuous in son baseflow. Intensification of land use that involves substantial
implying uni-directional effects, i.e. benefits only. soil compaction, will certainly lead to an increase in the flood
potential. Where such compaction is small in area (as is com-
monly the case), however, this effect will be localised and will
CONCLUSIONS not extend to the basin scale. Finally, there is evidence emerging
that forest cover could have a direct relationship with precipita-
The findings from research into land use-hydrological interac- tion at scales greater than 10 km2 and certainly at the scale of
tions suggest that the reduction or conversion of natural vegeta- the Amazon (see Pielke et al. (1999), as well as Heil Costa; and
tion accompanying land use change is likely to increase down- Bonell, Callaghan and Connor, this volume). Thus the relation-
stream sediment levels and lead to higher nutrient and chemical ship between land use and these hydrological variables is mixed
outflows. The empirical literature on this topic supports the con- with some positive and some negative effects and others for which
ventional wisdom that the end result will be a decrease in economic there is no generalisation.
welfare due to a myriad of downstream effects on production by Changes in water quantity will affect a large range of produc-
enterprises, the household production function and consumption tive and consumptive activities, often affecting the same activi-
by individuals. Although the general direction of the effect of ties influenced by sedimentation. Interestingly, however, few of
land use change on water quality can be surmised, there remain the empirical studies of sedimentation have also considered water
legitimate questions as to whether the literature available conveys quantity effects. In forest areas, land use change may lead to major
accurately the magnitude of these damages. In particular, conven- changes in rates of evapotranspiration and so it would appear indis-
tional wisdom that such adverse water quality effects must always pensable to combine both aspects into the analysis of externalities.
be of disastrous proportions and merit immediate attention across This concern may be less pressing in temperate grassland areas,
the board is probably flawed as the economic valuation studies however, the study by Kim (1984) suggests that even in a drier
reviewed in this chapter demonstrate that the magnitude of the grassland environment the choice of land management technique
effects will likely vary according to the economic and biophysical may have a large impact on water yield.
characteristics and conditions of the site. It is also the case that many of the studies appearing in the
With regard to the effects of land use change on water quantity literature are either extremely simplistic or flawed in their formu-
variables, the review of the hydrological literature reveals that the lation or implementation, limiting the reliability of their results
116 B . AY LWA R D

and at worst leading to the confusion of positive and negative existing literature cannot be taken as evidence that in practice the
externalities. As observed by Aylward (1998) there is also a large net effect is typically negative as most studies are either incomplete
methodological gap between the rudimentary valuations provided or unreliable. The small but growing number of studies reviewed
in the externalities literature (reviewed here) and the complex here sustain the theoretical conclusion that there will be cases
dynamic optimisation models employed in the design and opera- where the net result of the hydrological impacts of forest clearance
tion literature (as found for example in Water Resources Research (away from natural vegetation) may lead to increases in economic
and Journal of Water Resources Planning and Management). In welfare or produce only trivial losses in welfare. In such cases
this regard, it is worth noting that the failure to make a connection the initial production benefits (e.g. timber, livestock, agricultural
between these two larger sets of literature is mutual. The optimi- outputs) of the subsequent land use would have to decline sig-
sation of reservoir operation is not mentioned in the literature on nificantly (or be negative), before basin rehabilitation would be
economic valuation of watershed management. While sedimen- warranted.
tation and land use are occasionally mentioned in the operations This analysis argues for more emphasis on ‘catchment man-
literature, issues of land use and water quantity are effectively agement’ as opposed to ‘catchment protection’. In the case where
ignored (Howard, 2000). existing old-growth forest is under threat there will be a host of
The range of empirical studies reviewed in this chapter reveals other goods and services that need to be included in any evalua-
a heavy emphasis on the economic evaluation of sedimentation tion of land use alternatives. When combined with the prevalence
impacts with only a few studies examining water quantity and of uncertainty regarding the direction or magnitude of hydrologi-
water quality (excluding sedimentation) impacts. Given that water cal externalities associated with potential land use change (due to
quality and water quantity impacts may affect the same con- their site-specific biophysical and economic nature) a risk-averse
sumptive or productive activity, the exclusion of water quantity posture is likely to be the prudent approach. Water quality impacts
impacts from consideration implies that much of the literature is can be expected to be negative and water quantity impacts ambigu-
incomplete. ous, so it is natural to prefer to realise the other goods and ser-
Combining the analysis of hydrological effects and economic vices provided by forests and avoid potentially negative impacts
effects, a discussion of the sign (or direction) of the differ- on hydrological function.
ent impacts confirms that in most cases land use change (away Thus, the analysis in this chapter should not be taken as provid-
from natural vegetation) will affect economic welfare negatively ing support to relaxing the protection and management afforded
through its impact on water quality. However, it cannot be argued to forest areas. It does, however, point out that in developing mar-
a priori that all water quantity impacts will have a similarly nega- kets for environmental services, and in this case for hydrological
tive economic outcome. Review of the empirical evidence on sed- or catchment services, critical analysis is needed to avoid blithely
imentation impacts also suggests that these impacts may often be assuming that downstream water users should be willing to pay
of limited economic consequence. Meanwhile, empirical studies large amounts of money for hydrological ‘services’ provided by
of water quantity impacts often either misinterpret the direction of intact forest. It would be unfortunate indeed if large transfers of
hydrological change (based on erroneous conventional wisdom) this kind led to little or a perverse impact and it was clear that no
or rely on questionable hydrological and economic assumptions effort had been made to even consult the available scientific and
to demonstrate negative impacts. economic literature in designing the scheme.
Thus, the principal conclusion of the chapter is that both the- Further, the decision to take a risk-averse approach has impli-
ory and empirical evidence suggest that it would be incorrect to cations for the case of reforestation of already converted lands.
assume that the hydrological externalities resulting from land use In this case, the goods and services already being supplied will
change are necessarily negative. As a result it may be time to be the crop or livestock values of the land. The downside risk
reconsider the conventional wisdom that land use change away of reforesting will be the potential for negative impacts on dry
from natural vegetative states must always impair catchment pro- season baseflow. Clearly, the extent of land degradation and the
tection values – when these are narrowly defined as hydrological downstream effects on water quality are a prime consideration, but
in nature. Clearly in any comprehensive assessment of land use the point is that risk aversion works against land use change here
choices such hydrological externalities would be but one crite- as well. This is particularly true where land-use change involves
rion amongst many, contributing to the decision process alongside significant socio-economic dislocation or entails large up-front
biodiversity, timber and other values. costs in terms of retraining of rural workers and the direct costs
Notwithstanding the larger decision framework, this chapter of reforestation. In order to authorise large expenditures of this
has shown that on theoretical grounds the case can be made that, nature, decision-makers should logically place the burden on the
a priori, the net outcome of the effect of land use on the dif- proponent to present a clear and well-reasoned assessment of the
ferent hydrological functions is indeterminate. Empirically, the potential hydrological benefits and risks, as well as the costs of
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 117

such programmes, the losses in agricultural values and gains in The degree of involvement of stakeholders will vary with the
timber and non-timber forest product values. objectives and content of the research. For applied work that is
Of course, where reforestation leads to carbon sequestration aimed at policy or project development, stakeholders should be
these values must also be included and considered. Viewed from a consulted and involved in the project on a continual basis, from
macro-perspective, the potential for reforestation to reduce water assisting in the identification and prioritisation of research topics
availability raises the larger issue of the trade-off between two and sub-themes through to the dissemination, outreach and policy/
equally critical sets of ecological services. As carbon is a service project formulation phases. For basic research, stakeholder con-
of global value and hydrological services of local or national value, sultation will likely be more punctual, but nevertheless should be
the act of balancing these two sets of demands or needs may well used to ensure that the research design addresses local concerns
evolve into an important policy issue. and issues where feasible. The time and money costs of partic-
Future research priorities for the valuation of hydrological ipation will vary, but it is important that they be provided for
externalities will revolve primarily around efforts to encourage explicitly in project budgets and time schedules. In particular, it
multidisciplinary work. It is likely that effort needs to be devoted is important to avoid under-budgeting resources for outreach and
not to the development of new methods, per se, but rather that communication of research results.
an investment must be made in determining how models and From an economic standpoint the overriding concern has to
methods applied in each area can be joined into a comprehensive do with the economic contribution that such research can make
approach to the problem. Given the complexity of the interactions to local and national development goals. If it is true that such
involved, the investigation of hydrological externalities is likely research can greatly improve the productivity and efficiency with
to require participation by experts in land use/productive systems, which catchment resources are managed, then there is no better
forest hydrologists, engineers and economists. While economists way to ensure that such research is funded than by providing assis-
are conversing increasingly with hydrologists, engineers tend tance to the actors that will actually reap these benefits (or avoid
to be left out of the equation and land use aspects are simply the costs of poor management/investments). While water is a pub-
assumed. lic good, the distribution of the benefits of improved catchment
The literature review suggests that water quantity and water management may often be localised in a particular region. Achiev-
quality impacts are largely under-researched and that there is ing stakeholder buy-in to a research programme will thus not only
great scope for expanding our understanding of the relationships increase the likelihood that the research will lead ultimately to wel-
between the different variables. Additional case studies and more fare improvement but may open up new partnerships and funding
general theoretical work would greatly assist in the development avenues for researchers.
of a clear set of rules of thumb and shortcuts that could contribute
to better project and policy formulation. In this regard, a funda-
mental question to which hydrologists need to respond is whether, References
to what extent, and under what conditions, it is possible to develop Alvarez, M., Aylward, B., and Echeverrı́a, J. (1996). Management of Natu-
reliable predictive models for land use and hydrology interactions ral Forest for Off-Site Hydrological Benefits in the Magallenes National
Reserve. Santiago, Chile: Corporación Nacional Forestal, Environmental
in the absence of calibrated datasets for catchments. As the chap- Resources Management, Ltd., British Council.
ter notes, much of the policy-oriented studies are short-term when Aylward, B. (1998). Economic Valuation of the Downstream Hydrological
Effects of Land Use Change: Large Hydroelectric Reservoirs. Unpub-
compared to long or medium-term hydrological data collection lished Ph.D. diss., Fletcher School of Law and Diplomacy, Medford,
and research. Nor is it possible to guarantee that catchments that MA.
Aylward, B., and Echeverria, J. (2001). Synergies between Livestock Produc-
are to be the subject of policy or project interventions are those tion and Hydrological Function in Arenal, Costa Rica. Environment and
that have historically been metered. Development Economics, 6(3), 359–82.
Aylward, B., and Fernández González, A. (1998). Incentives and Institutional
There are many reasons, some more or less obvious, for advocat- Arrangements for Watershed Management: A Case Study of Arenal,
ing increased stakeholder participation in research programmes – Costa Rica, CREED Working Paper Series. London: International Insti-
whether academic or applied (Deutsch et al., this volume). Two tute for Environment and Development and Institute for Environmental
Studies, Free University, Amsterdam.
central objectives of stakeholder involvement are to ensure that the Barbier, E. B. (1998). Natural Capital and the Economics of Environment and
research responds to local conditions and concerns and to increase Development. In E. B. Barbier (Ed.), The Economics of Environment and
Development. Cheltenham, UK: Edward Elgar.
the likelihood of the practical uptake of research results in actual Barbier, E. B., and Bishop, J. (1995). Economic Values and Incentives Affect-
practice. Stakeholders will include both those who live and work ing Soil and Water Conservation in Developing Countries. Journal of Soil
and Water Conservation, 50(2), 133–37.
in the catchments as well as those who benefit at a distance from Barbier, E. B., Burgess, J. C., and Folke, C. (1994). Paradise Lost? The Eco-
the services provided by water resources. Policy-makers and tech- logical Economics of Biodiversity. London: Earthscan.
Barrow, P., Hinsley, A., and Price, C. (1986). The Effect of Afforestation
nical staff of relevant agencies and utilities are also an important on Hydroelectricity Generation. A Quantitative Assessment. Land Use
set of stakeholders. Policy(4), 141–51.
118 B . AY LWA R D

Bishop, J. (1992). Economic Analysis of Soil Degradation, LEEC Gatekeeper Crosson, P. (1984). New Perspectives on Soil Conservation Policy. Journal of
Series (Vol. 92–01, pp. 13). London: International Institute for Environ- Soil and Water Conservation, 39(4), 222–25.
ment and Development. (1995). Soil Erosion and its On-Farm Productivity Consequences: What
(1995). The Economics of Soil Degradation: An Illustration of the Change Do We Know?, Discussion Paper (Vol. 95-29, pp. 17). Washington:
in Productivity Approach to Valuation in Mali and Malawi, LEEC Dis- Resources for the Future.
cussion Paper (Vol. 95–02, pp. 79). London: International Institute for Crosson, P., and Miranowski, J. (1982). Soil Protection: Why, by Whom, and
Environment and Development. for Whom? Journal of Soil and Water Conservation, 37(1), 27–28.
Bosch, J. M., and Hewlett, J. D. (1982). A Review of Catchment Experiments Crowder, B. M. (1987). Economic Costs of Reservoir Sedimentation: A
to Determine the Effect of Vegetation Changes on Water Yield and Evap- Regional Approach to Estimating Cropland Erosion Damages. Journal
otranspiration. Journal of Hydrology, 55, 3–23. of Soil and Water Conservation, 42(3), 194–97.
Bouwes, N. W. (1979). Procedures in Estimating Benefits of Water Quality Cruz, W. D., Francisco, H., and Conway, Z. (1988). The On-Site and Down-
Change. American Journal of Agricultural Economics, 61(3), 535–39. stream Costs of Soil Erosion in the Magat and Pantabangan Watersheds.
Briones, N. D. (1986). Estimating Erosion Costs: A Philippine Case Study Journal of Philippine Development, 15(26), 85–111.
in the Lower Agno River Watershed. In K. W. Easter and J. A. Dixon Daubert, J., and Young, R. (1981). Recreational Demands for Maintaining
and M. M. Hufschmidt (Eds.), Watershed Resources Management: An Instream Flows: A Contingent Valuation Approach. American Journal of
Integrated Framework with Studies from Asia and the Pacific (pp. 205– Agricultural Economics, 63(4), 666–75.
18). Boulder: Westview Press. De Graaff, J. (1996). The Price of Soil Erosion: An Economic Evaluation of
Brooks, K. N., Gregersen, H. M., Berglund, E. R., and Tayaa, M. (1982). Soil Conservation and Watershed Development (Vol. 14). Wageningen:
Economic Evaluation of Watershed Projects: An Overview Methodology Wageningen Agricultural University.
and Application. Water Resources Bulletin, 18(2), 245–50. de Wit, M., Crookes, D. J., and van Wilgen, B. W. (forthcoming). Conflicts of
Brown, M., de la Roca, I., Vallejo, A., Ford, G., Casey, J., Aguilar, B., Interest in Environmental Management: Estimating the Costs and Bene-
and Haacker, R. (1996). A Valuation Analysis of the Role of Cloud fits of Black Wattle (Acacia mearnsii) in South Africa. Biological Inva-
Forests in Watershed Protection: Sierra de las Minas Biosphere Reserve, sions.
Guatemala and Cusuco National Park, Honduras. Guatemala: RARE, de Wit, M., Crookies, D., Scott, D., Visser, W., and Netshiluvhi, T. (2000).
Defensores de la Naturaleza and Fundación Ecologista Hector Rodrigo Environmental Impacts of the Forestry Sector in South Africa with Specific
Pastor Fasquelle. Reference to Water Resources (ENV-P-C 99016, Project: JN716). Preto-
Brown, T. C., Taylor, J. G., and Shelby, B. (1992). Assessing the Direct Effects ria: Division of Water, Environment and Forestry Technology, CSIR; PO
of Streamflow on Recreation: A Literature Review. Water Resources Bul- Box 395, Pretoria 0001.
letin, 27(6), 979–89. Duda, A. M. (1985). Environmental and Economic Damage Caused by Sed-
Bruijnzeel, L. A. (1990). Hydrology of Moist Tropical Forests and Effects of iment from Agricultural Nonpoint Sources. Water Resource Bulletin,
Conversion: A State of Knowledge Review. Paris: International Hydro- 21(2), 225–34.
logical Programme of the United Nations Educational, Scientific and Duffield, J. W., Neher, C. J., and Brown, T. C. (1992). Recreation Benefits of
Cultural Organization. Instream Flow: Application to Montana’s Big Hole and Bitterroot Rivers.
Bruijnzeel, L. A. (1997). Hydrology of Forest Plantations in the Tropics. In Water Resources Research, 28(9), 2169–81.
E. K. S. Nambiar and A. G. Brown (Eds.), Management of Soil, Nutrients Duisberg, P. (1980). Erosión y Conservación de Suelos. In CCT (Ed.), Estu-
and Water in Tropical Plantation Forests (pp. 125–67). Canberra/Bogor: dio Ecológico Integral de las Zonas de Afectación del Proyecto Arenal
ACIAR/CSIRO/CIFOR. (pp. Anexo 2:1-65). San José: Centro Cientı́fico Tropical.
(1998). Soil Chemical Changes after Tropical Forest Disturbance and Con- Enters, T. (1995). The Economics of Land Degradation and Resources Con-
version: The Hydrological Perspective. In A. Schulte and D. Ruhyat servation in Northern Thailand. In J. Rigg (Ed.), Counting the Costs:
(Eds.), Soils of Tropical Forest Ecosystems. Characteristics, Ecology and Economic Growth and Environmental Change in Thailand (pp. 90–110).
Management (pp. 45–61). Berlin: Springer. Singapore: Institute of South East Asian Studies.
(2002). Tropical forests and environmental services: not seeing the soil for (1998). Methods for the Economic Assessment of the On- and Off-site
the trees? Agriculture, Ecology and Environment. Impacts of Soil Erosion, Issues in Sustainable Land Management (Vol. 2,
Bruijnzeel, L. A., and Proctor, J. (1995). Hydrology and Biogeochemistry pp. 60). Bangkok: International Board for Soil Research and Manage-
of Tropical Montane Cloud Forest: What Do We Really Know? In ment.
L. S. Hamilton and J. O. Juvik and F. N. Scatena (Eds.), Tropical Mon- Epp, D. J., and Al-Ani, K. S. (1979). The Effect of Water Quality on Rural
tane Cloud Forests (Vol. Springer Ecol. Studies, pp. 38–78). New York: Non-Farm Residential Property Values. American Journal of Agricultural
Springer-Verlag. Economics, 61, 529–34.
Calder, I. R. (1992). The Hydrological Impact of Land-Use Change (with Forster, D. L., and Abrahim, G. (1985). Sediment Deposits in Drainage
Special Reference to Afforestation and Deforestation), Proceedings of the Ditches: A Cropland Externality. Journal of Soil and Water Conservation,
Natural Resources and Engineering Advisers Conference on Priorities 40(1), 141–43.
for Water Resources Allocation and Management, Southampton, July Forster, D. L., Bardos, C. P., and Southgate, D. (1987). Soil Erosion and Water
1992 (pp. 91–101). London: United Kingdom Overseas Development Treatment Costs. Journal of Soil and Water Conservation, 42(1), 349–52.
Administration. Fox, G., and Dickson, E. (1990). The Economics of Erosion and Sediment
(1999). The Blue Revolution: Land Use and Integrated Water Resources Control in Southwestern Ontario. Canadian Journal of Agricultural Eco-
Management. London: Earthscan. nomics, 38, 23–44.
CCT, and CINPE. (1995). Valoración Económico Ecológica del Agua. San Freeman, A. M. (1993). The Measurement of Environmental and Resource
José: Centro Cientı́fico Tropical and Centro Internacional en Polı́tica Values. Washington DC: Resources for the Future.
Económica para el Desarrollo Sostenible. Gregersen, H. M., Brooks, K. N., Dixon, J. A., and Hamilton, L. S. (1987).
Cheng, G. W. (1999). Forest Change: Hydrological Effects in the Upper Guidelines for Economic Appraisal of Watershed Management Projects,
Yangtze River Valley. Ambio, 28, 457–59. FAO Conservation Guide (Vol. 16, pp. 144). Rome: United Nations Food
Clark, E. H. (1985a). Eroding Soils: The Off-Farm Impacts. Washington, DC: and Agriculture Organization.
The Conservation Foundation. Gunatilake, H. M., and Gopalakrishnan, C. (1999). The Economics of Reser-
(1985b). The Off-Site Costs of Soil Erosion. Journal of Soil and Water voir Sedimentation: A Case Study of Mahaweli Reserviors in Sri Lanka.
Conservation, 40(1), 19–22. Water Resources Development, 15(4), 511–26.
Clark, R. (1996). Methodologies for the Economic Analysis of Soil Erosion Guntermann, K. L., Lee, M. T., and Swanson, E. R. (1975). The Off-site
and Conservation, CSERGE Working Paper (Vol. GEC 96-13). Norwich: Sediment Damage Function in Selected Illinois Watersheds. Journal of
University of East Anglia. Soil and Water Conservation, 30, 219–25.
Creedy, J., and Wurzbacher, A. D. (2001). The Economic Value of a Forested Guo, Z., Xiao, X., Gan, Y., and Zheng, Y. (2001). Ecosystem Functions, Ser-
Catchment with Timber, Water and Carbon Sequestration Benefits. Eco- vices and their Values – a Case Study in Xingshan County of China.
logical Economics, 38(1), 71–83. Ecological Economics, 38, 141–54.
L A N D U S E , H Y D RO L O G Y A N D E C O N O M I C VA L UAT I O N 119

Hamilton, L. S., and King, P. N. (1983). Tropical Forested Watersheds: Hydro- Protection in Eastern Indonesia. Water Resources Research, 37(3), 771–
logic and Soils Response to Major Uses or Conversions. Boulder: West- 78.
view Press. (2001b). Worth of Watersheds: A Producer Surplus Approach for Valuing
Hamilton, L. S., and Pearce, A. (1986). Biophysical Aspects in Watershed Drought Mitigation in Eastern Indonesia. Environment and Development
Management. In C. W. Howe (Ed.), Watershed Resources Management: Economics, 6, 123–46.
An Integrated Framework with Studies from Asia and the Pacific (1st ed., Perrings, C., Folke, C., and Maler, K.-G. (1992). The Ecology and Economics
Vol. 10, pp. 33–49). Boulder: Westview Press. of Biodiversity Loss: The Research Agenda. Ambio, 21(3), 201–11.
Hearne, R. R. (1996). A Review of Economic Appraisal of Environ- Pielke, R. A., Walko, R. L., Steyaert, L. T., Vidale, P. L., Liston, G. E., Lyons,
mental Goods and Services: With a Focus on Developing Coun- W. A., and Chase, T. N. (1999). The Influence of Anthropogenic Land-
tries, Environmental Economics Programme Discussion Paper (Vol. 96- scape Changes on Weather in South Florida. Am. Meteorol. Soc., (127),
03, pp. 13). London: International Institute for Environment and 1663–72.
Development. Pierce, F. J., Larson, W. E., Dowdy, R. H., and Graham, W. P. (1983). Produc-
Hitzhusen, F. J. (1992). The Economics of Sustainable Agriculture: Adding tivity of Soils: Assessing Long-Term Changes Due to Erosion. Journal
a Downstream Perspective. Journal of Sustainable Agriculture, 2(2), of Soil and Water Conservation, 38(1), 39–44.
75–89. Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair,
Hodgson, G., and Dixon, J. A. (1988). Logging Versus Fisheries and Tourism in M., Crist, S., Shpritz, L., Fitton, L., Saffouri, R., and Blair, R. (1995).
Palawan: An Environmental and Economic Analysis, Occasional Paper Environmental and Economic Costs of Soil Erosion and Conservation
(Vol. 7, pp. 95). Honolulu: East-West Center. Benefits. Science, 267(24 February), 1117–23.
Holmes, T. P. (1988). The Offsite Impact of Soil Erosion on the Water Treat- Quesada-Mateo, C. A. (1979). Effect of Reservoir Sedimentation and Stream
ment Industry. Land Economics, 64(4), 356–66. Flow Modification on Firm Power Generation. Unpublished Ph.D. diss.,
Howard, C. D. (2000). Operations, Monitoring and Decomissioning of Dams, Colorado State University, Fort Collins, CO.
WCD Thematic Review (Vol. IV.5). Cape Town: World Commission on Ralston, S. N., and Park, W. M. (1989). Estimation of Potential Reductions in
Dams. Recreational Benefits due to Sedimentation. Water Resources Bulletin,
Hubbard, R. K., and Lowrance, R. (1996). Solute Transport and Filtering 25(6), 1259–65.
Through a Riparian Forest. Transactions of the ASAE, 39(2), 477–88. Ribaudo, M. O., Young, C. E., and Shortle, J. S. (1986). Impacts of Water Qual-
Intercarib S. A, and Nathan Associates. (1996). Manejo Ambiental, Aspectos ity Improvement on Site Visitation: A Probabilistic Modeling Approach.
Institucionales, Económicos y Financieros: Volumen 2 de 2, Estudios Water Resources Bulletin, 22(4), 559–63.
Sectoriales. Panamá: Intercarib S.A. y Nathan Associates, Inc. Richards, M. (1997). The Potential for Economic Valuation of Watershed
Johnson, N., White, A., and Perrot-Maitre, D. (2001). Developing Mar- Protection in Mountainous Areas: A Case Study from Bolivia. Mountain
kets for Water Services from Forests: Issues and Lessons for Innova- Research and Development, 17(1), 19–30.
tors. Washington, DC: Forest Trends, World Resources Institute, The Rodrı́guez, R. (1989). Impactos del Uso de la Tierra en la Alteración
Katoomba Group. del Régimen de Caudales, la Erosión y Sedimentación de la Cuenca
Johnson, N. S., and Adams, R. M. (1988). Benefits of Increased Streamflow: Superior del rı́o Reventazón y los Efectos Económicos en el Proyecto
The Case of the John Day River Steelhead Fishery. Water Resources Hidroeléctrico de Cachı́, Costa Rica. Unpublished MSc thesis, Centro
Research, 24(11), 1839–46. Agronómico Tropical de Investigación y Eneñanza, Turrialba.
Johnson, S. H., and Kolavalli, S. (1984). Physical and Economic Impacts Rojas, M., and Aylward, B. (Forthcoming). Mechanisms for Upstream-
of Sedimentation on Fishing Activities: Nam Pong Basin, Northeast Downstream Cooperation in Rural Watersheds: The Case of a La Esper-
Thailand. Water International, 9(4), 185–88. anza: A Small, Private, Hydropower Producer in Costa Rica. In B. Kier-
Kim, I.-H. (1984). An Economic Analysis of Watershed Practices: Impact of sch (Ed.), Watershed-Based Integrated Land and Water Management:
Grazing on Watershed (Erosion, Sedimentation, Valuation). Unpublished Valuation of Land Use Impacts on Water Resources and Mechanisms
Ph.D. diss., Utah State University. for Upstream-Downstream Cooperation in Rural Watersheds. Rome:
Kumari, K. (1995). An Environmental and Economic Assesment of For- FAO.
est Management Options: A Case Study in Malaysia, Environmental Ruitenbeek, H. J. (1990). Economic Analysis of Tropical Forest Conservation
Economics Series, Environment Department Papers (Vol. 26, pp. 46). Initiatives: Examples from West Africa. Surrey, UK: World Wide Fund
Washington, DC: World Bank. for Nature.
Lant, C. L., and Mullens, J. B. (1991). Lake and River Quality for Recreation Santos, B. A. (1992). Cost Benefit Analysis of Soil Erosion Control: The Case of
Management and Contingent Valuation. Water Resources Bulletin, 27(3), Plan Sierra (Dominican Republic). Unpublished Ph.D. diss., University
453–60. of California, Berkeley.
Ledesma, H. R. (1996). Effects of Reservoir Sedimentation and the Economics Sheridan, J. M., Lowrance, R., and Bosch, D. D. (1999). Management Effects
of Watershed Management: Case Study of Aguacate Dam, Dominican on Runoff and Sediment Transport in Riparian Forest Buffers. Transac-
Republic. Unpublished Ph.D. diss., University of Florida. tions of the ASAE, 42(1), 55–64.
Lutz, E., Pagiola, S., and Reiche, C. (1994). The Costs and Benefits of Soil Con- Snyder, N. J., Mostaghimi, D. F., Berry, D. F., Reneau, R. B., Hong, S., McClel-
servation: The Farmer’s Viewpoint. The World Bank Research Observer, lan, P. W., and E. P., S. (1998). Impact of Riparian Forest Buffers on
9(2), 273–95. Agricultural Nonpoint Source Pollution. Journal of the American Water
Magrath, W., and Arens, P. (1989). The Costs of Soil Erosion on Java: A Nat- Resources Association, 34(2), 385–95.
ural Resource Accounting Approach, Environment Department Working Southgate, D. (1992). The Rationality of Land Degradation in Latin America:
Paper (Vol. 18, pp. 60). Washington DC: World Bank. Some Lessons from the Ecuadorian Andes, LEEC Gatekeeper Series
McConnell, K. (1983). An Economic Model of Soil Conservation. American (Vol. 92-04, pp. 9). London: International Institute for Environment and
Journal of Agricultural Economics, 65(1), 83–89. Development.
Mohd Shahwahid, H. O., Awang Noor, A. G., Abdul Rahim, N., Zulkifli, Y., Southgate, D., and Macke, R. (1989). The Downstream Benefits of Soil Con-
and Ragame, U. (1997). Economic Benefits of Watershed Protection and servation in Third World Hydroelectric Watersheds. Land Economics,
Trade-off with Timber Production: A Case Study in Malaysia, EEPSEA 65(1), 38–48.
Research Report Series. Singapore: Economy and Environment Program Stallard, R. F. (1997). Detailed Report of the Soil and Hydrology Component,
for South East Asia. Informe Anual del Program de Monitoreo del Canal. Panamá: Programa
Narayanan, R. (1986). Evaluation of Recreational Benefits of Instream Flows. del Monitoreo del Canal.
Journal of Leisure Research, 18(2), 116–28. Stocking, M. A. (1984). Erosion and Soil Productivity: A Review. Norwich:
Olson, K. R., Lal, R., and Norton, L. D. (1994). Evaluation of Methods to Overseas Development Group, University of East Anglia.
Study Soil Erosion-Productivity Relationships. Journal of Soil and Water van Noordwijk, M., and al, e. (1998). Erosion and sedimentation as multiscale,
Conservation, 49(6), 586–90. fractal processes: implication for models, experiments and the real world.
Pattanayak, S. K., and Kramer, R. A. (2001a). Pricing Ecological Ser- In F. W. T. Penning de Vries (Ed.), Soil Erosion at Multiple Scales.
vices: Willingness to Pay for Drought Mitigation from Watershed Wallingford: CABI Publishing.
120 B . AY LWA R D

Veloz, A., Southgate, D., Hitzhusen, F., and Macgregor, R. (1985). The Eco- Ward, F. A. (1987). Economics of Water Allocation to Instream Uses in a
nomics of Erosion Control in a Subtropical Watershed: A Dominican Fully Appropriated River Basin: Evidence from a New Mexico Wild
Case. Land Economics, 61(2), 145–55. River. Water Resources Research, 23(3), 381–92.
Vertessy, R. A., Watson, F., O’Sullivan, S., Davis, S., Campbell, R., Benyon, White, W. C. (1994). The Study of Sustainable Hydrological Development in
R., and Haydon, S. (1998). Predicting Water Yield from Mountain Ash Lao PDR. Unpublished Ph.D. diss., Fletcher School of Law and Diplo-
Forest Catchments. Canberra: CRCCH. macy, Tufts University, Medford, MA.
Vincent, J. R., and Kaosa-ard, M. (1995). Policy Brief: The Economics of Young, C. E. (1984). Perceived Water Quality and the Value of Seasonal
Watershed Protection: A Case Study of the Mae Teng River. Cambridge: Homes. Water Resources Bulletin, 20(2), 163–66.
Harvard Institute for International Development. Zalewski, M., Thorpe, J. E., and Naiman, R. J. (2001). Fish and Riparian
Walker, D. J. (1982). A Damage Function to Evaluate Erosion Control Eco- Ecotones: A Hypothesis. Ecohydrology and Hydrobiology, 1(1–2), 11–
nomics. American Journal of Agricultural Economics, 64(11), 690–98. 24.