See discussions, stats, and author profiles for this publication at: [Link]

net/publication/264784928

Priority areas for watershed service conservation in the Guapi-Macacu Region

of Rio de Janeiro, Atlantic Forest, Brazil

Article in Ecological Processes · August 2014

DOI: 10.1186/s13717-014-0016-7

CITATIONS READS

25 660

6 authors, including:

Vanesa Rodríguez Osuna Jan Börner

University of Bonn University of Bonn
13 PUBLICATIONS 52 CITATIONS 187 PUBLICATIONS 9,616 CITATIONS

SEE PROFILE SEE PROFILE

Udo Michael Nehren Rachel Bardy Prado

TH Köln - University of Applied Sciences Brazilian Agricultural Research Corporation (EMBRAPA)
136 PUBLICATIONS 1,746 CITATIONS 165 PUBLICATIONS 729 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Rachel Bardy Prado on 11 April 2022.

The user has requested enhancement of the downloaded file.

Rodríguez Osuna et al. Ecological Processes 2014, 3:16
[Link]

RESEARCH Open Access

Priority areas for watershed service conservation

in the Guapi-Macacu region of Rio de Janeiro,
Atlantic Forest, Brazil
Vanesa Rodríguez Osuna1,2*, Jan Börner1, Udo Nehren2, Rachel Bardy Prado3, Hartmut Gaese1 and Jürgen Heinrich4

Abstract
Introduction: Land use intensification and urbanisation processes are degrading hydrological ecosystem services in
the Guapi-Macacu watershed of Rio de Janeiro. A proposal to pay farmers to restore natural watershed services
might be an alternative to securing the water supply in the long-term for the around 2.5 million urban water users
in the study region. This study quantifies the costs of changing current land use patterns to enhance watershed
services and compares these costs to the avoided costs associated with water treatment for public supply.
Methods: We use farm-household data to estimate the opportunity costs of abandoning current land uses for the
recovery of natural vegetation; a process that is very likely to improve water quality in terms of turbidity due to
reduced inputs from erosion. Opportunity cost estimates are extrapolated to the watershed scale based on remote
sensing land use classifications and vulnerability analysis to identify priority zones for watershed management
interventions. To assess the potential demand for watershed services, we analyse water quality and treatment cost
data from the main local water treatment plant.
Results: Changing agricultural land uses for watershed services provision generally comes at high opportunity costs
in our study area near to the metropolis of Rio de Janeiro. Alternative low cost watershed conservation options do
exist in the livestock production sector. These options have the potential to directly reduce the amount of
sediments and nutrients reaching the water bodies, and in turn decrease the costs of treatment needed for
drinking water. Land cover changes at the scale needed to improve water quality will, nonetheless, likely exceed
the cost of additional investments in water treatment.
Conclusions: The state water utility company’s willingness to pay for watershed services alone will not be enough
to induce provision of additional watershed services. We conclude that monetary incentives conditioned on specific
adjustments to existing production systems could still have a complementary role to play in improving watershed
services. However, we note that our willingness to pay analysis focusses on only one of the potentially wide range
of ecosystem services provided by natural vegetation in the Guapi-Macacu watershed. Factoring these ecosystem
services into the willingness to pay equation is likely to change our assessment in favour of additional conservation
action, be it through PES or other policy instruments.
Keywords: Water services; Drinking water treatment; Water quality; Incentive-based watershed management

* Correspondence: vanesa@[Link]
1
Center for Development Research (ZEF), Department of Economic and
Technological Change, University of Bonn, Walter-Flex-Str. 3, 53117 Bonn,
Germany
2
Institute for Technology and Resources Management in the Tropics and
Subtropics (ITT), Cologne University of Applied Sciences, Betzdorfer Str. 2, 50679
Cologne, Germany
Full list of author information is available at the end of the article

© 2014 Rodriguez Osuna et al.; licensee Springer. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License ([Link] which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 2 of 21
[Link]

Introduction Stanton et al. 2010). PWS in Latin America typically

One of the biggest challenges of our time is to revert the involve compensating upstream rural producers for pro-
ongoing degradation of ecosystems while meeting the in- tecting and/or restoring natural forest ecosystems and
creasing demand for food and biomass (Thomas and highland natural pastures (páramo) (Pria et al. 2013; Veiga
Callan 2010; MEA 2005). Population and economic growth and Galvadao 2011; Grieg-Gran et al. 2005). Such ecosys-
are increasing the demand for water resources and, at the tems are mostly located in strategic water production
same time, amplifying the pressure on ecosystems that de- areas, such as headwaters, riparian forest or water intake
liver watershed services (Porras et al. 2008). points for water supply (Pria et al. 2013; Veiga and
The degradation of ecosystem services represents a Galvadao 2011). There is, however, little evidence on the
loss of widely undervalued natural capital assets (TEEB effectiveness of such schemes in meeting conservation
2009; Montes 2007); MEA 2005. While there is general and development goals in the peer-reviewed literature (see
agreement that land use choices influence watershed ser- for example Arriagada et al. 2012; Pattanayak et al. 2010).
vice provision, the magnitude and nature of the effects In Brazil, PWS are expanding and there are already 848
are highly context dependent and poorly understood in ecosystem service providers in the Atlantic Forest region
tropical forest environments (Porras et al. 2008; Veiga alone, mainly organised or supported within the National
2008; Calder 2005; van Noordwijk 2005). Water Agency’s Water Producer Programme (Veiga and
Land and water linkages are challenging to manage as Galvadao 2011; Santos et al. 2010). As of 2012, 41 PWS
watershed services have a common good characteristic projects had been implemented or were in preparation,
(Porras et al. 2008) and are generally unaccounted for in covering an area of around 40,000 hectares. A project in
monetary terms. As a result, they are often degraded the state of Rio de Janeiro is currently in development, in
and this is not reflected in traditional economic mea- the Três Picos State Park, Municipality of Cachoeiras de
sures, such as GDP (TEEB 2009; Costanza et al. 1997; Macacu (Guedes and Seehusen 2011). This State Park is
FAO 2007). Land use patterns and land management located in the mountainous area of the Guapi-Macacu
practices by upstream landholders in a watershed deter- watershed, which contributes to the supply of drinking
mine, to a great extent, quality and quantity of water water for almost 2.5 million inhabitants of five municipal-
bodies (Porras et al. 2008). For example, unsustainable ities, including the city of Niteroi (Pedreira et al. 2009)
land use and agricultural practices can result in negative and this watershed is the focus of the present study. The
hydrological side-effects, or externalities, such as sedi- main drivers of degradation of water resources in this
mentation (Veiga 2008; MEA 2005). As a result, the de- watershed are urbanisation processes, intensive agriculture
sign of policy measures that encourage farmers to adopt and conversion of riparian vegetation.
watershed conservation measures and, at the same time, Managing watershed services requires solid knowledge
maintain or even increase productivity has become a about (1) the costs of providing an additional unit of
major research issue. water quality or quantity (supply side) and (2) water
Several policy options are available to enhance incen- users or intermediaries’ willingness to pay for an add-
tives for the supply of ecosystem services. Command- itional unit of water quality or quantity (demand side).
and-control regulations, such as bans on forest clearing, The economic aspects of watershed services supply and
have long been the preferred policy option to control en- demand are particularly poorly studied and many PWS
vironmentally harmful land use changes (Boerner and initiatives begin operating without quantitative know-
Vosti 2012; Porras et al. 2008; FAO 2007). More re- ledge of such parameters (Martin-Ortega et al. 2012). To
cently, mechanisms based on economic incentives or address this knowledge gap, we focus on two research
markets for ecosystem services are being proposed as questions within our study area of the Guapi-Macacu
cost-effective complementary tools to promote ecosys- watershed in the state of Rio de Janeiro:
tem service conservation (Porras et al. 2008; Engel et al.
2008; FAO 2007). These incentive-based mechanisms in- 1. What are the costs associated with shifting land uses
clude payments for ecosystem services (PES) and pay- to foster improved and enhanced watershed services
ments for watershed services (PWS), water quality (services supply)?
trading markets and reciprocal or in-kind agreements 2. What are the treatment costs for the water supply
(Bennett et al. 2013). company to improve water quality parameters such
Latin America registered a transaction value of USD as turbidity?
528.9 million in such investments between 2001 and 2011
corresponding to 3.4 million hectares (Bennett et al. Our focus is on understanding the supply and environ-
2013). PWS programmes are considered to be the “fastest mental conditions that determine watershed service
growing” and most mature among all PES schemes in provision. Particularly, we concentrate on the watershed
Latin America (Bennett et al. 2013; Balvanera et al. 2012; service related to water quality maintenance in terms of
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 3 of 21
[Link]

turbidity reduction for supplying drinking water. However, 2013; Hanson et al. 2011; TEEB 2009; Postel et al. 2005).
we also analyse factors that influence water treatment costs Some examples are included in Table 1. For example, in the
(related to services demand) based on land use changes. United States of America, 27 water suppliers showed that
The remainder of this paper is structured as follows, sec- in watersheds with at least 60% forest cover, treatment costs
tion “Assessing watershed services supply and demand” were 50 percent less than those with 30% forest cover
presents our approach, addresses relevant literature and (Postel et al. 2005). Accordingly, several North American
concepts related to watershed service supply and demand, cities have decided to invest in watershed management to
section “Methods” describes the study area and methods avoid the expense of water treatment plants.
used for this research, section “Results” presents our find- In the literature regarding water quality, off-site effects
ings, section “Discussion” discusses these findings and con- of soil erosion are frequently referred to as sedimentation
textualises them with other studies and reviews the main (Veiga 2008; Holmes 1988). A summary of economic
caveats. Finally, section “Conclusions” presents our main activities most affected by loss of watershed services is
conclusions and the policy implications for watershed presented in Table 2.
service management. Different authors have classified ecosystem services in
distinct ways (Haines-Young and Potschin 2013 Daily
Assessing watershed services supply and demand and Matson 2008; FAO 2007; Farber et al. 2006; MEA
The supply of watershed services 2005; Postel et al. 2005; Hawkins 2003; De Groot et al.
Watershed service provision is to a large extent deter- 2002; Costanza et al. 1997). In the current study, we
mined by land use and land cover. For example, unsus- followed the TEEB (2010b) classification of watershed
tainable land use is frequently linked to a high surface services, whereby maintenance of water quality for hu-
runoff and an elevated concentration of suspended and man consumption is considered a provisioning service.
soluble loads in water bodies (Batchelor et al. 1998). We use this terminology throughout this article.
Changes in land cover, such as forest to agriculture con- It is often argued that a major challenge of market
version, tend to increase superficial runoff and sediment mechanisms relates to the “packaging” of ecosystem ser-
flux in rivers (FAO 2007; MEA 2005). vices into commodities that are tradable or subject to a
Healthy forest and wetland ecosystems are considered contract (Porras et al. 2008). However, most PWS
very effective at regulating water flow and improving schemes in developing countries are guided by a “land-
water quality (Russi et al. 2013; TEEB 2010a,b). Main- based” approach, whereby suppliers are paid to improve
taining water quality includes the control of sediment, their land management practices, which are in turn con-
nutrient (in particular phosphorous and nitrogen) and sidered highly likely to result in watershed service
chemical loads, as well as salinity (TEEB 2010b). In provision, rather than being paid for the actual service
addition, forest ecosystems can remove pathogenic mi- delivery (Porras et al. 2008). Regardless of the approach
crobes, sequester and convert inorganic ions and trans- chosen, watershed services supply is inevitably linked to
form persistent organic pollutants (TEEB 2010b). farmers’ land use decisions (FAO 2007). Consequently,
Watersheds with an important extension of forest tend watershed services supply analyses often require cost-
to offer better water quality than those subject to other benefit analyses of agricultural production systems.
land uses, such as agriculture, pasture, industry and The concept of opportunity costs (OCs) is most fre-
urban infrastructure, since the latter are associated with quently used to express the costs of additional watershed
higher discharges of diverse types of pollutants into soils service provision. In the context of watershed services,
and water. In this way, the presence of forest could sub- OCs represent any benefits foregone by a farmer upon
stantially reduce the cost of water treatment for drinking converting from their current land use to an alternative
water in most cases, thereby reducing the related costs form of land use that is more watershed service-friendly.
for water supply (Medeiros et al. 2011; Reis 2004). Ideally, PWS are designed in such a way that payments
Cities such as Rio de Janeiro, Johannesburg, Tokyo, compensate for at least the OCs of additional service
Melbourne, New York and Jakarta all depend on pro- provision or the foregone benefits of the land use promoted
tected areas with forests to provide drinking water for in order to provide the service. The extent to which such
their residents (Dudley and Stolton 2003). Moreover, a payments are or are not appropriate depends on the alter-
third of the hundred largest cities worldwide take a sig- native land uses in each given area (Pagiola et al. 2010).
nificant proportion of their drinking water from pro-
tected forested areas (Dudley and Stolton 2003). The demand for watershed services
Worldwide several examples show that well-managed When dealing with demand, we refer to those who are
forests, wetlands and protected areas are very likely to sup- currently benefiting from the delivery of watershed ser-
ply clean water at significantly lower costs than man-made vices and to the resources available for protecting and
replacements, such as water treatment plants (Bennett et al. conserving these services (Guedes and Seehusen 2011).
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 4 of 21
[Link]

Table 1 Examples of the economic values of watershed services

Action Examples Source
Payments to maintain water purification services in the Payments for maintaining watershed purification represent 1 to Hanson et al. 2011;
Catskills watershed, New York avoid the costs of launching a 1.5 billion USD in comparison to the much higher estimated WRI 2011; Pagiola
filtration plant for water treatment cost of a filtration plant (6–8 billion USD plus 300–500 million et al. 2004
USD yearly for operating costs)
The cost of removing nitrate contamination from water in 2.8-4.8 million USD per year Hawkins 2003
Rochester, Minnesota
WTP for water quality improvement in the levels of iron, 2.4, 2.0, 6.6 and 2.6 million USD yearly respectively Hawkins 2003
sulphate, hardness, and copper in South-western, Minnesota (via contingent valuation method)
Yearly revenues in public utility resulting from natural water 94 USD per acre Hawkins 2003
supply in Mud Lake, Minnesota/ South Dakota
Annual WTP from residents living close to the Minnesota 14.1 USD via taxes or 19.6 USD via water bills for a 40% Hawkins 2003
River for reducing phosphorous levels reduction in a nearby river (contingent valuation method)
Increased reservoir water quality and surface area in St. Yearly net benefit of 25 million USD (travel cost method) Hawkins 2003
Louis, Missouri
Value of water supply in Milesburg, Pennsylvania Between 14 and 36 USD per household (avoided cost) Hawkins 2003
In Venezuela, a national protected area system avoids Without the provision of this service by the national protected Pabon-Zamora et al.
sedimentation area, unavoided sedimentation could reduce farmer’s income 2008
by approximately 3.5 million USD yearly
The current provision of ecosystem services related to 79,000 EUR per year savings in water treatment costs for the Honey-Roses et al.
existing stream vegetation along the Llobregat River, Spain residents of Barcelona 2013
In New Zealand, the value of water provision from the Now it is “free” but would cost 136 million NZDa to bring it in TEEB 2009
Otago region coming from the Te Papanui Conservation Park from somewhere else (total benefit of the service)
a
1 USD is equivalent to 1.30 NZD at current exchange rate ([Link]

Table 2 Examples of economic activities most affected by the loss of watershed services
Economic activities and Examples of incurred costs/damages due to the loss of watershed services Source
damage caused by the loss
of watershed services
Activities that depend on reservoirs Capacity loss for energy generation Veiga 2008; Reis 2004
Capacity loss of irrigated production
Loss of benefits related to flood control
Capacity loss of navigation channels
Increase of suspended material/siltation of water bodies resulting in a reduction
of their storage capacity
Drainage and maintenance Costs related to (a) irrigation and drainage of channels, (b) hydroelectric plant Veiga 2008; Reis 2004
operations reservoirs, (c) ports and (d) eutrophication (increase concentration of N and P
in water bodies)
Increased water treatment One example are watersheds in São Paulo with lower costs related to Reis 2004; Dearmont et al.
costs related to augmented turbidity consumption of chemical products for the water treating process (less than 1998; Holmes 1988,
20 BRLa per 1,000 m3 of treated water) were those with a forest cover higher Moore and McCarl 1987
than 15%. The two water treatment units with less percentage of forest cover
(Piracicaba and Atibaia rivers with less than 10% forest cover) show considerably
higher specific costs of chemical products
Monetary damages related Additional annual costs from 35 to 661 million USD for a US water treatment López 1997; Holmes 1988
to sedimentation plant due to soil erosion. These values relate strongly to agricultural production
causing sedimentation
Additional annual operational cost of 3.2 BRL per hectare and 0.10 BRL per ton
of eroded soil for the water treatment of the river Corumbataí in São Paulo.
These estimates were made assuming that the only externality caused by soil
erosion were increased water treatment costs but not taking into account
siltation of river banks, flooding of areas close to river banks and loss of
navigation capacity
a
1 USD is equivalent to 2.38 BRL at current exchange rate ([Link] This rate will be used onwards in this document when
comparison among monetary units is required.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 5 of 21
[Link]

Besides PWS, which are carried out by national states development and implementation of well-known PES pro-
in Latin America, a user’s fee system can be an effective grammes in Mexico and Costa Rica (FAO 2007). Involve-
approach to efficiently use water resources. The water- ment of the private sector in paying for ecosystem services,
sheds of the rivers Piracicaba, Capivari and Jundiaí (PCJ) including in the context of corporate social responsibility,
in the state of São Paulo have implemented such a user’s is growing (TEEB 2010c).
fee system with around 8.8 million of beneficiaries of the Considering these factors, supply and demand analysis
Cantareira system (Veiga and Galvadao 2011). In this of watershed services provides essential information to
situation, an Inter-Municipal Basin Committee was assess the economic and financial viability of PWS
formed to manage a watershed protection fund and con- schemes (IIED 2012a). Figure 1 illustrates our approach
tributions to the fund come from the municipal water in identifying the economic preconditions for a potential
utility budgets. PWS scheme in the Guapi-Macacu watershed.
As a further alternative, some programmes in the Atlantic As the figure suggests, we expect PWS to be a viable
Forest region are subsidised by the government, for ex- policy option only if the willingness of water users to
ample: “Bolsa Verde”, “ProdutorES de Água” and “Mina pay (i.e. the demand of the water company) exceeds
D’água” in the states of Minas Gerais, Espírito Santo and the opportunity costs (OCs) of additional watershed
São Paulo, which pay rural producers for conservation ac- service provision (supply) incurred by land users in the
tivities on their properties (Guedes and Seehusen 2011). watershed.
These include the protection or restoration of native vege-
tation areas with a focus on headwaters and riparian forests Methods
(Guedes and Seehusen 2011; Veiga and Galvadao 2011). The Guapi-Macacu watershed
Another significant source of finance for PWS in devel- The study area lies within the Guapi-Macacu watershed
oping countries comes from the international public sector (1,263 km2), which is located in the Serra do Mar bio-
funding or development assistance (Porras et al. 2008). A geographical region (Ribeiro et al. 2009) and is a priority
key provider to this funding is the Global Environmental conservation target within the Atlantic Forest biodiver-
Facility (GEF), which acts as buyer on behalf of service sity hotspot in the state of Rio de Janeiro (CEPF 2001),
users for conserving global public services. Around 108 see Figure 2.
million USD and 52 million USD have been made available The Atlantic Forest biome is characterised by historically
as World Bank (WB) loans and GEF grants respectively for high deforestation rates (Dean 1997), resulting in a highly
WB/GEF-supported PWS projects (FAO 2007). For ex- fragmented area with numerous isolated and disperse for-
ample, the World Bank has given loans to support the est fragments in a landscape dominated by agricultural

Figure 1 Conceptual framework for assessing watershed services. This framework considers water quality maintenance for public supply
considering supply is at the farming system level and demand is at the watershed level.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 6 of 21
[Link]

Figure 2 The Guapi-Macacu watershed within the state of Rio de Janeiro.

production systems (Nehren et al. 2013; Ribeiro et al. The water intake point of the water supply is located
2009). This biome supplies 135 million people with water in the lower part of the watershed and managed by
(Pria et al. 2013); however, a mere 11 to 16 percent of its the Drinking Water and Wastewater State Company
original forest cover is left (Ribeiro et al. 2009). (CEDAE). Due to the good water quality coming from
Forest cover in the state of Rio de Janeiro and more the springs of the rivers Macacu and Guapiaçu, several
specifically in the Guapi-Macacu watershed, is compara- mineral water companies have been established in the
tively high due to its rugged topography. Unfortunately, region, as well as enterprises for which water is an
continuous forest cover occurs mostly on steep slopes essential input (Strobel et al. 2007).
that are inappropriate for agriculture, while the foothills In the Guapi-Macacu watershed, various demands in
and lowlands are dominated by agriculture and pasture water supply to domestic, industrial and agricultural
(Nehren et al. 2013; Strobel et al. 2007). In 2008, the consumption are taking place as a result of the good
land cover/land use of the Guapi-Macacu watershed water quality from the water sources of the main rivers
consisted of forests in all stages (48.8%), pastures (Strobel et al. 2007). Moreover, the ongoing construction
(41.4%) and agriculture (4.4%). The remaining 5.4% were of a new petrochemical complex (COMPERJ) in the
covered by urban areas, water bodies, bare soil, rock out- downstream area of the watershed is expected to further
crop, wetlands and mangrove (Fidalgo et al. 2008). increase both water demand and pressure on natural re-
Rivers and creeks in the Guapi-Macacu watershed origin- sources in the study area (Pedreira et al. 2009).
ate mainly in the State Park “Três Picos”, but some stem Within the watershed, three sub-watersheds (see Figure 3):
from the National Park “Serra dos Órgãos” and in the State (I) Manuel Alexandre, (II) Batatal and (III) Caboclo were se-
Ecological Station “Paraíso”. The main rivers Macacu and lected for this study. Each of these sub-watersheds repre-
Guapiaçu originate within the State Park and constitute the sents the different types of land cover mix found in the
main components of this watershed (Strobel et al. 2007). area, namely “forest”, “agriculture” and “pasture”, which
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 7 of 21
[Link]

Figure 3 Sub-watershed landscapes of three rivers: Manuel Alexandre (left), Batatal (upper right) and Caboclo (lower right).

account altogether for around 95% of the total watershed (III) In the Caboclo sub-watershed, the predominant
area. By selecting three sub-watersheds with greatly varying land use type is forest (81%), followed by pasture
land use patterns and intensities, we were able to compare (14%) and agriculture (3%) according to Naegeli
them with respect to their differences on agricultural profit- (2010) and Fidalgo et al. (2008). Agricultural
ability and their distinctive impact on water quality parame- systems with a considerably higher intensity than
ters: in this case turbidity levels. in Batatal are found in this sub-watershed, mainly
along the floodplain. The most common annual
(I) The sub-watershed Manuel Alexandre is located in crops are maize, cassava, beans and vegetables,
the Ecological Reserve of Guapiaçu (REGUA) and whereas perennial crops are rare. The higher
represents a well-preserved landscape with a high elevation area is within the boundaries of the State
proportion of forest land (87%) (Fidalgo et al. Park Três Picos, where agricultural activities are
2008). It therefore served as the reference for a carried out within the buffer area of this Park.
nature-near, less disturbed forest ecosystem. This
area includes a private reserve with low human Both sub-watersheds, Batatal and Caboclo have under-
impact resulting primarily from ecotourism in the gone historical exploitation cycles, resulting in high de-
form of birdwatchers. Most of the area within this forestation, forest fragmentation and degradation, as
sub-watershed is protected under the REGUA well as intense soil erosion (Nehren et al. 2013).
Association, which is financially supported by the
Brazilian Atlantic Forest Trust (BART) with the Methods to assess supply of watershed services
stated objective of protecting the Atlantic Rainforest To calculate the opportunity costs (OCs) related to the
of the upper Guapiaçu river watershed. This is provision of watershed services under varying land use
mainly done by enlarging protected areas through systems, we carried out a cost-benefit analysis of repre-
land purchase. sentative farming systems in the region. For this, we de-
veloped detailed individual budgets for all activities
(II) In contrast, the sub-watershed Batatal represents a within a given farming system (for definition see Beets
mixed system of the most relevant land uses with a 1990). Activity budgets summarised revenue and cost
mosaic of forest fragments (69%), pastures (28%) information and were finally aggregated to calculate
and agriculture (4%) considering the land use the average rate of return for each land use type (WBI
classification by Fidalgo et al. (2008). Predominantly, 2011). Crop budgets were compared for coherence with
banana (perennial) is found in higher elevation areas, official current production costs used by the Rural
followed by annual crops mainly found in flat areas Extension and Technical Assistance Agency (EMATER)
or lowlands (cassava, maize, beans and vegetables). of the state of Rio de Janeiro.
Forest fragments in different stages of ecological Our target population was households practicing some
succession are found in high elevation areas. degree of agriculture at the sub-watershed level considering
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 8 of 21
[Link]

the sub-watersheds of Batatal and Caboclo. A household located at ca. 83 m.a.s.l. Agricultural production in the
survey was carried out in two field campaigns in 2011 and lowlands of Batatal is comparable to those sub-watersheds
2012 with the permanent support of key local producers, located along the Macacu River. The same occurs in
EMATER of the municipality of Cachoeiras de Macacu, the Caboclo, which is representative for sub-watersheds along
City Council of this municipality and Embrapa Soils scien- the Guapiaçu River (Figure 2).
tists. Expert interviews were carried out with the Director Subsequently, a cost-benefit analysis was carried out
of the REGUA Reserve and other staff members in Manuel for each farming system. The occurrence of each farm-
Alexandre sub-watershed, to better understand land use ing system was estimated and validated through local ex-
history and recent management practices in the region. pert consultation as suggested by Angelsen et al. (2011).
To define the sample universe, we created an inventory Our sampling strategy focused on capturing the diver-
using the indirect census technique following Forero sity of smallholder production systems in the region, yet
(2002). This process consisted of a participatory mapping our total sample size was too small to obtain a represen-
exercise based on recent aerial imagery provided by the tative sub-sample of the large cattle operations that
City Council of Cachoeiras de Macacu. This enabled us to dominate in the lower part of the watershed. For the
assemble a list of all farm units within the sub-watersheds, cost-benefit analysis of cattle production systems, we
which was confirmed by extensive field visits and thus relied on additional in-depth interviews with a
supported thoroughly by local experts including a member group of livestock producers deemed representative by
of the Agricultural Department of the City Council officers of EMATER.
(Cachoeiras de Macacu), the President of the Faraó Based on interviews with selected livestock producers
Farmer’s Association (A.L.A.F.) in the sub-watershed of and secondary data on livestock systems in this area (see
Batatal, a member of the Rural Workers Union (in the Quintana 2012), we calculated livestock activity budgets
sub-watershed of Caboclo) and the Director of the REGUA for three slope categories: 1) ≤15°; 2) 16-25° and 3) >25°.
Reserve (Manuel Alexandre sub-watershed). As a result, a These budgets were calculated under the assumption
total of 32 households in Batatal and 60 in Caboclo were that profits for livestock production decreased with in-
identified, of which 78 households within the two populated creasing slope, because of lower productivity of pasture,
sub-watersheds were interviewed using a semi-structured among other factors. This assumption is based on inter-
survey. No interviews were made in the reference site of views with local farmers.
Manuel Alexandre sub-watershed. The sample size ob- Once profits for agricultural and livestock systems were
tained is supported by Angelsen et al. (2011), who suggest a obtained, they were extrapolated to the watershed level
minimum sample size of 25–30 households from each using a Landsat based land cover classification that identi-
community. This is valid for communities with 100 to 500 fied “agriculture” and “pasture” areas (Pedreira et al. 2009;
families. The designed questionnaire was based on various Fidalgo et al. 2008). The agricultural profit calculated in
scientific publications and reports (see Rodriguez Osuna the selected sub-watersheds (only considering lowland
2013; Angelsen et al. 2011; WBI 2011; Gaese 2009; Instituto areas) was applied to all sub-watersheds in the same river
Terra Mater 2009; Forero 2002). network. In the sub-watershed of Batatal, we divided
Throughout the course of our fieldwork, two survey farming systems located in uplands and lowlands since
rounds were carried out. The first survey round included a this division influences significantly production patterns.
random sampling of farm units within each sub-watershed This was not necessary in the Caboclo sub-watershed,
to define “representative farming systems”. Important se- where all farming systems are located in the lowlands.
lection criteria for these farming systems as suggested by The profit derived from farming systems is equal to
Zimmer et al. (2009) and local experts included mainly: the opportunity cost of converting agricultural or pas-
farm size, land tenure, production programme and agricul- toral lands to forest, thereby reducing turbidity. For ex-
tural management practices, and average location of farms ample low OCs are associated with low profits from
in terms of metres above sea level (m.a.s.l.). current land use. Per hectare OC estimates for each sub-
Once such farming systems were defined, a second watershed thus represent the weighted average per hec-
survey round was launched to explore in-depth charac- tare profits from the respective land cover types.
teristics of farming systems with special attention to the We relied on a spatial analysis of the vulnerability of
inputs and outputs that are relevant to profitability water resources in the study area (Ferreira 2012), which
among such systems. was understood as the likelihood of watershed service loss.
In the sub-watershed of Batatal, we divided the farm- In this case, our assessment considered state and pressure
ing systems by location in upland versus lowland, since indicators following the Driving Forces, Pressure, State,
this division significantly affects production patterns. Impact and Response (DPSIR) framework (see Borja et al.
Farm units in the uplands are located at an average alti- 2006), where 50% were state indicators including: geo-
tude of 344 m.a.s.l. in contrast to those in the lowlands morphology, hydrogeology, drainage density, soils, index
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 9 of 21
[Link]

of circularity, index of areas of permanent protection is based on the assumption that environmental goals are
(APP) fragments and slope. APP areas are established by more important than cost criteria. However, this could
the Brazilian Forest Code (Federal Law 4771/1965) and be changed when there are budget restrictions and when
are defined as “protected areas, both covered or not with there is intent to increase efficiency of payments in com-
native vegetation, that have the environmental functions pensation schemes for watershed services. The lowest
of preserving water resources, landscapes, geological sta- priority areas are those sub-watersheds with high OCs
bility, biodiversity, and genetic fluxes of flora and fauna, as and low vulnerability of water resources.
well as protection of the soil and securing the wellbeing of
human populations” (Ministry of Environment 2005). Methods to assess the potential demand for watershed
These areas include a minimum vegetation area to protect services
riverbanks and headwaters. The other 50% included pres- Demand for watershed services was assessed by identify-
sure or anthropogenic indicators such as phosphorous (P) ing costs related to the end-user of those services (Honey-
and nitrogen (N) production, land use and road density Roses et al. 2013). For this particular study case, it in-
(Ferreira 2012). All mentioned factors (state and pressure) cluded quantifying water treatment costs incurred by the
were weighed by hydrological expert consultation by main watershed user, the state water utility (CEDAE).
Ferreira 2012 (Figure 4). Single indicators were based Water treatment costs that may be avoided if forests are
mostly on published official maps predominantly gener- restored can be translated as the potential willingness to
ated by Embrapa Soils. pay (WTP) for watershed services. For example, if a one
Our next step was to identify priority areas for water- unit reduction in turbidity levels implies 10 additional
shed service provision. For this, we overlaid environ- monetary units in treatment costs, the water company’s
mental and economic criteria, i.e. vulnerability of water maximum WTP for watershed services will be 10.
resources and spatial OCs within the Guapi-Macacu We applied the avoided cost method focusing on the
watershed. This allowed us to identify the areas where annual operational costs of chemical products for the
low OCs of shifting land use towards improving water- treatment of raw water in the period between 1998 and
shed services can result in high watershed service pay- 2011 from the local state water utility company. This ap-
offs. These sub-watersheds were given the highest proach required identifying key water quality indicators
priority for intervention with watershed service improve- related to the main operational cost categories of the
ments. Second and third priorities were given to sub- water utility company.
watersheds with high OCs and vulnerability and those Based on expert interviews and as suggested by Medeiros
with low OCs and vulnerability, respectively. The latter et al. (2011), Reis (2004) and Dearmont et al. (1998), we

Figure 4 State and Pressure Indicators to assess water resources vulnerability. These indicators were weighed by hydrological expert
consultation to assess the vulnerability of water resources in the Guapi-Macacu watershed (Modified from Ferreira 2012).
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 10 of 21
[Link]

identified turbidity as the key indicator amongst all water use. In Batatal, most of the land is obtained by heritage
quality parameters, since an increase in turbidity implies or with a provisionary land title (66%), followed by ba-
an increased concentration of suspended solids in surface nana share-croppers (19%), who are generally entitled to
waters and is likely to reduce the quality of raw water to be half the harvest. In contrast, in Caboclo most of the
treated by the water utility. The conversion of forest to population lives within a settlement that keeps a com-
other land uses such as agriculture or pasture caused by mon forest area called a “private reserve”. This reserve is
farming systems can result in higher turbidity values, protected and is restricted from cultivation. Therefore,
which in turn correlate with higher water treatment costs 85% of the farm-households in Caboclo are located in
to reduce the concentration of suspended solids for public the lowlands, while the rest are found on the hillsides. In
water supply (Medeiros et al. 2011; Reis 2004; Dearmont both sub-watersheds, one farm unit does not surpass 14
et al. 1998). hectares, which is considered typical for family-oriented
In addition Medeiros et al. (2011) and Reis (2004) sug- agriculture for the municipality of Cachoeiras de Macacu
gest that chemical products account for close to 60% of (MDA 2010). After considering the specific differences
annual operational costs in treatment plants in São Paulo among farming systems in Batatal and Caboclo, four
(Reis 2004). These products are used to flocculate the sus- types of farming systems (FS) were classified in Batatal
pended particles, measured as turbidity, that are found in and two in the sub-watershed of Caboclo (Table 3).
raw water to fulfill regulations on drinking water quality In Batatal, we found that 80% of FS in the uplands are
for human consumption (maximum value of 5 NTU). specialised in banana (Musa sp.) production (classified as
FS 1) and 20% had a mixed system of cassava (Manihot sp.)
Results and banana (FS2). In the lowlands of Batatal, two add-
Vulnerability of water resources itional FS where agriculture is more intensive were classi-
Water resource vulnerability is a function of both an- fied as FS3 and FS4 and are equally distributed. FS3 has a
thropogenic impact/pressure indicators and of environ- production system composed of cassava, green maize (Zea
mental state indicators, so the assessment of vulnerability mays), yams (Colocasia sp.) and courgette (Cucurbita sp.),
must account for this set of indicators. When only consid- while FS4 has the same mix of cassava, green maize and
ering environmental state indicators, sub-watersheds lo- yams, but banana instead of courgette (Table 3).
cated in higher areas of the Guapi-Macacu watershed tend Cassava is the dominant crop in the Caboclo sub-
to have higher vulnerability to anthropogenic pressure watershed, followed by green maize, yams and common
than those in the lowlands (Ferreira 2012). Pressure indi- beans (Phaseolus vulgaris). Cassava is the most culti-
cators in a watershed are highly influenced by population vated crop, mainly due to low investment requirements
density, land use practices, presence of urban settlements, and because of relatively stable returns after a cropping
road density and other factors previously mentioned. period of 8 to 9 months. Green maize requires higher in-
Therefore, the sub-watersheds with relatively high an- vestments; however, it provides relatively rapid returns
thropogenic impacts are those with a high density of after only 90 days. Common beans are used to improve
urban settlements and rural population nuclei. Sub- soil fertility (nitrogen fixation) and as an alternative to
watersheds found in higher areas of the watershed have the other products. Other relevant short-cycle products
considerably lower impact values (Ferreira 2012). are okra (Hibiscus esculentus) and gilo (Solanum gilo).
Figure 5 shows the sub-watersheds with higher vulner- The most common farming system in the Caboclo
ability in the darker tones and those with lower vulner- sub-watershed (FS5) combines cassava, yams, common
ability in the lighter tones. The less vulnerable sub- beans and green maize (70%), while the rest (FS6-FS6a)
watersheds are found in the lower areas of the watershed combine cassava, yams, common beans, green maize
and one of these is considered a natural protected area and -additionally- okra or gilo (Table 3). Typically, green
with limitations and restrictions in land use, despite its maize, courgette and beans are planted twice in one
lower vulnerability. cropping cycle.
Agriculture in the uplands of Batatal is clearly less inten-
Agricultural production and opportunity costs sive in fertilizer use than in the lowlands, especially given
Field surveys carried out in the Batatal and Caboclo sub- that banana production in the uplands is carried out with-
watersheds showed differences in production patterns out fertilization. The remoteness of these locations makes
and differences in specific environmental factors that re- intensive production less attractive than in the lowlands.
flect both the effect of farming systems on the provision In the lowlands of Batatal, vegetable producers applied on
of watershed services as well as distinct farming systems’ average of 240 kg of fertilizer per hectare each year. In
profitability. Caboclo, more than 70% of households used fertilizers for
Relevant characteristics that differ considerably across their agricultural production, at an average of 547 kg per
these two sub-watersheds include land tenure and land hectare each year.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 11 of 21
[Link]

Figure 5 Vulnerability of water resources in the Guapi-Macacu watershed (Ferreira 2012).

In the uplands of Batatal, all households used herbi- In Batatal, the per hectare average annual profit was
cides as their basic approach to weed control, whereas a estimated at 4,115 BRL and in Caboclo at 5,052 BRL
variety of agrochemicals were used in the lowlands. The (Table 3). Returns for agriculture tend to be higher in
high use of herbicides for weed control could be the re- the lowlands, where the intensity of production is higher
sult of the shortage of and high cost of labour that than in the uplands, and there is a higher use of agricul-
would otherwise undertake this activity. It should also tural inputs especially fertilizers for cash crops.
be noted that the Atlantic Forest Law (Law 11.428/1986) According to local expert interviews small livestock
and the Brazilian Forest Code (Law 12.651/2012) bring farmers are considered those with farm size less than 20
certain limitations to agricultural production. The Atlantic hectares, while big-scale producers are considered those
Forest Law bans the conversion of secondary forest into with more than 400 hectares. According to official cattle
land uses such as agricultural land. As an example, a col- vaccination data in the Municipality of Cachoeiras de
lapse of the market price for banana around fifteen years Macacu in 2011, an estimation of the herd size can be
ago left many banana plantations uncultivated and sec- given on the base of vaccinated animals. This resulted in
ondary forests developed and expanded, leading to aban- 27,995 animals in all three districts of this Municipality
donment of these plantations. In addition, the Brazilian (Secretary of Agriculture, Cattle Farming, Fisheries and
Forest Code defines for the Atlantic Forest Biome that Supply 2011). Most livestock farmers (90%) have less
20% of rural properties need to be maintained as a than 500 animals and small-scale producers are consid-
permanent forest reserve “Reserva Legal”. The Brazilian ered in this municipality those with less than 20 animals.
Forest Code also prohibits the clearing of primary vegeta- Mostly, animals are distributed in paddocks without
tion on steep slopes (>45°), along the margins of rivers dividing fences.
and streams and in headwater (source) areas, which are Livestock production systems achieved profits of 20,
classified as areas of permanent protection (APPs) (Ministry 40 and 100 BRL per hectare annually, depending on
of Environment 2005). slope class (Table 3).
 [Link]
Rodríguez Osuna et al. Ecological Processes 2014, 3:16
Table 3 Summary of annual farming systems profits in the study area
Main crops Unit Batatal sub-watershed Caboclo sub-watershed Pasture land systems
oriented to beef production
Upland Lowland Lowland
considering slopea
16° to >25°
FS1 FS2 FS3 FS4 FS5 FS6a FS6b ≤15°
25°
Banana 972 972 983
Cassava −1,334 3,950 3,950 5,231 5,232 5,232
Green maize
2,371 2,371 2,066 2,067 2,067
(summer)
Green maize (winter) 3,438 3,438 2,371 2,371 2,371
Yam (BRL) 2,861 2,861 6,199 6,199 6,199
Courgette 3,633
Gilo 10,941
Okra 10,092
Beans (summer) 1,419 1,419 1,419
Beans (winter) 1,522 1,523 1,523
Occurrence (%) 80 20 50 50 70 15 15
Typical area and crop 4.0 (1.0 cassava, 4.0 (1.0 cassava, 1.5
10.3 (8.3 banana, 6.0 (2.0 cassava, 2.0 green 6.0 (2.0 cassava, 2.0 green
distribution for each (ha) 10.3 1.5 green maize, green maize, 0.5 yam,
2.0 cassava) maize, 1.0 yam, 1.0 banana) maize, 1.0 yam, 1.0 courgette)
farming system (FS) 1.0 yam, 0.5 beans) 0.5 okra or gilo, 0.5 beans)
Per hectare profit (BRL ha−1) 972 526 3,894 4,336 4,889.7 5,376.3 5,482.5 20 40 100
Average value for Macacu River lowland Guapiaçú River
extrapolation at (BRL ha−1)
the watershed level 4,114.8b 5,051.6 20 40 100
a
According to expert interviews, there is a gain of 40 to 60 kg of live animal weight in high slope areas in contrast to those in lowlands with a gain in live animal weight of 120 to 150 kg in this particular watershed;
b
Since upland agriculture in the Batatal sub-watershed (FS1 and FS2) was not detected by the available land use classification (Naegeli 2010; Pedreira et al. 2009; Fidalgo et al. 2008), we were limited to using average
per hectare annual profits for lowland agriculture in Batatal (FS3 and FS4) resulting in an average annual per hectare value of 4,114.8 BRL.

Page 12 of 21
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 13 of 21
[Link]

Area-weighted OCs per sub-watershed were spatially extent, in protected areas such as the State Park “Três
mapped, ranging from 14 to 1,660 BRL per hectare Picos”, the National Park “Serra dos Órgãos” and the
(Figure 6). This reflects that extensive pasture is the State Ecological Station “Paraíso”.
most important land use in the Guapi-Macacu water-
shed according to area, which is an activity with com- Analysis of environmental and economic criteria for
paratively low per hectare profits. watershed service conservation and improvement
Figure 5 shows that quite a number of high OC areas Analysis of environmental (vulnerability of water re-
are located close to the main two river beds, where in- sources) and economic criteria (OCs of watershed ser-
tensive agriculture predominates. Resulting OCs from vice provision) in the Guapi-Macacu watershed allowed
agriculture occur only in lowlands and close to the river us to identify priority areas for watershed service conser-
plain, which have higher nutrient concentration than in vation and improvement. These areas are the land use-
higher slope areas. based management options (i.e. conversion of pastoral
Low OCs areas are often either located in the steep or agricultural lands to forest) with the highest potential
slope areas of the upper watershed or in the valleys, where of improving water quality and lowest OCs (Figure 7).
extensive pasture areas dominate. However, these areas Our results highlight the sub-watersheds where the vul-
are also found in the lower parts of the watershed, where nerability of water resources are highest and OCs of con-
a small-scale settlement promoting family-oriented cattle verting land uses to foster watershed service provision are
ranching was launched a decade ago by the government lowest. The high priority areas (in dark red) are in most
of Rio de Janeiro. cases found in sub-watersheds with steeper average slope
Forests predominate mainly in higher slope areas levels where impact on water resources was found to be
(mostly in white in Figure 6), which originate, to a great high (Figure 7).

Figure 6 Spatial per hectare opportunity costs in BRL for the land uses “agriculture” and “pasture” within each sub-watershed in
the Guapi-Macacu watershed. The white areas within the map correspond to those with land uses other than agriculture in the
Guapi-Macacu watershed.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 14 of 21
[Link]

Figure 7 Priority areas for improving or ensuring watershed services of the Guapi-Macacu watershed. Priority areas for improving or
ensuring watershed services (water quality) where 1 is considered high and 4 represents low priority.

Demand for watershed services: water treatment metres (Table 5). The Pearson correlation coefficient for
According to micro-economic theory, we interpret the average annual turbidity levels and costs for chemical prod-
water utility company’s demand for chemicals to treat ucts in this water utility for the period between 1998 and
water as its willingness to pay for a desired water quality 2011 was 0.4. This correlation value is relatively low com-
level (in this case turbidity under 5 NTU for human pared to other studies carried out in Brazil. For example,
consumption). The treatment cost of an additional tur- Reis (2004) calculated a correlation value of 0.7 for seven
bidity unit is thus equivalent to the company’s potential water treatment units in São Paulo and Cabral de Sousa
willingness to pay for any measure that reduces turbidity (2011) found a coefficient of 0.9 when comparing 10 differ-
by the same amount (see avoided cost method, for ex- ent sub-watersheds and their treatment costs in São Paulo.
ample, in Perman et al. 2003). Reis (2004) found that chemical treatment costs in
Based on expert interviews and relevant studies (Medeiros water utility units on the Piracicaba River are 12.7 times
et al. 2011, Reis 2004 and Dearmont et al. 1998), which higher than the cost of treating water from the Cantareira
identified turbidity as the key water quality parameter system. This author argues that this finding can be ex-
relevant for water treatment costs, we applied the avoided plained by the considerably lower forest cover in the
cost method to the local water utility company. In Table 4, Piracicaba watershed (4.3%) compared to the Cantareira
the main characteristics of the water treatment are watershed (27.2%).
presented. Therefore, the geographical location of forests, as well
Based on data from CEDAE (treatment unit Laranjal) for as land use, soil type, geomorphology and predominant
the period between 1998 and 2011, we estimated the geology are considered relevant factors that influence
avoided costs of a reduction of 1% turbidity at 15,510 BRL, water quality from headwaters (springs) and water treat-
considering an average volume of treated water of 174,545 ment costs for public supply. Nevertheless, Reis (2004)
cubic metres at an average cost of 22.2 BRL per 1,000 cubic showed that the percentage of forest cover is often a
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 15 of 21
[Link]

Table 4 Characteristics of water treatment utility in the lower catchment part of the Guapi-Macacu watershed in 2011
Treatment unit: Laranjal - CEDAE Municipality: São Gonçalo
Captivation area from the GMW 1,263 km2
Forest cover in the watershed 48.8%
Treatment type Conventional
General treatment phases Captivation, sedimentation, coagulation, flocculation, decantation, filtration,
disinfection, water fluoridation and pH correction
Treated water flow Average flow in 2011, 5.35 m3 s−1
Population supplied with treated water 2,000,000 inhabitants
Aluminium sulphate Al2(SO4)3, polyelectrolyte, hexafluorosilicic acid H2SIF6,
General chemical products used
chlorine, calcium oxide CaO
Costs with chemical products for the treatment of raw water 2.31 × 106 BRL (year 2011)
Turbidity of raw water, average values 17.10 NTU
Treated water characteristics Colour, 2.50 uH
Turbidity, 0.34 NTU
For the treatment of 6 m3 s−1,
Total costs with chemical products and electricity in a year for the • 300,000 [BRL] (monthly expenditure for chemical products)
water treatment unit
• 100,000 [BRL] (monthly expenditure for electricity in the captivation and
production area within the unit)

sufficiently informative indicator of watershed health watershed based on avoided costs is 15,510 BRL per
and thus water quality. additional 1% reduction in turbidity levels at the water
Water quality monitoring obtained in seven moni- intake point of the water utility.
toring campaigns in the years 2010 and 2011 along Ideally, a full-scale hydrological model for the water-
the whole watershed demonstrated lower turbidity levels shed would provide us with the potential effect of alter-
in the highest parts of the sub-watershed, where forest native land use scenarios on turbidity at the water intake
cover is higher. For example, our reference sub-watershed point of the water utility. In the absence of such a
Manuel Alexandre presented an average value of 0.8 NTU model, we can only provide an informed estimation with
at the outlet as compared to 17.4 NTU at a lower region regard to the viability of PWS in the watershed. At the
close to the water intake point of the water utility high OCs end, it is clearly unrealistic to expect that the
(Paiva et al. 2011). conversion of less than 3 hectares of intensively used
cropland in the whole watershed (the avoided cost of
Is paying for land use changes that foster watershed water quality reduction: 15,510 BRL / maximum per
services cheaper than treating water? hectare OC: 5,482.5 BRL per ha = 2.83 ha) will result in
The estimated OCs of converting land use/land cover to a 1% reduction of turbidity levels. In contrast, carefully
the benefit of water quality can range between 4,000 to selected pasture and low intensity agricultural sites could
5,000 BRL per hectare each year for agricultural systems potentially be converted into forest (162 to 814 hectares)
and less than 100 BRL for pasture land. However, the ac- in the case of pastures. Land use changes at that scale
tual OCs per hectare in many sub-watersheds are likely are more likely to bring about measurable changes in
to be much lower, especially if land close to rivers and turbidity levels if located in zones with a large impact on
headwaters is covered by extensive pastures. On the river water quality.
demand side, we estimate that the water company Unfortunately, given the spatial distribution of pas-
CEDAE’s maximum WTP for land use change in the tures and high intensity agriculture in the watershed,
there are likely to be limited opportunities to convert
large tracts of land at low costs. Payments for forest re-
Table 5 Annual avoided costs from 1998–2011 for the cuperation may thus likely remain a complementary
local water utility company watershed management measure in our study area. As
Average Average cost Total cost Avoided cost an alternative to full scale conversion, some simple pas-
quantity (1,000 BRL m−3) (BRL) for 1% of turbidity ture management techniques, such as limiting access of
(m3yr−1) reduction (BRL)
cattle to the riverbed in lowland pastures (see dark red
174,545.3 22.2 3,877,482.0 15,509.9
areas in the centre of Figure 7), could prove to be
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 16 of 21
[Link]

comparatively low cost and highly effective measures to after heavy rainfalls. These accelerated slope processes that
reduce turbidity levels close to the water intake point of are associated with high sediment loads and turbidity can
CEDAE. even take place under closed forest cover, as seen for
Inadequate livestock grazing practices can compromise example during the 2011 mudslides and floods (Nehren
water quality to the point where is considered degraded et al. 2014). The nature of soils and geology in the water-
and highly polluted and not able to meet water quality shed (e.g. erosion of the riverbed) also determines how
standards (EPA 2013). Therefore, excluding livestock from easily erosion might occur (ECI 2014).
streams and improving range management practices can We found a lower correlation between water treatment
contribute to reduce turbidity on streams (EPA 2013). costs and turbidity (0.4) than the 0.7 found by Reis
Although the water supply company’s WTP does not (2004) in her analysis of 7 treatment utility plants in São
match the estimated OCs, payments from other water Paulo. The use of average values for turbidity and the
users are an additional option that can be taken into ac- costs of chemicals used for treatment would most likely
count in the design of a PWS scheme in this watershed. mask a higher correlation. However, our correlation is
Particularly important would be the demand of water by relevant since it supports our conclusion that the WTP
COMPERJ (the Rio de Janeiro petrochemical complex). In from the water supply company is likely to be rather
addition, this assessment was solely carried out on the low. Yet, turbidity levels monitored at the water intake
basis of water quality improvements in terms of turbidity point of the water supply facility do not exceed 50 NTU
levels, since it was found this service to be relevant for the (Paiva et al. 2011).
demanders downstream. However, accounting for add- Comparing these turbidity levels for water treatment
itional ecosystem services provided by forest ecosystems with other cases in Brazil, we found mean turbidity
would increase the potential WTP for ecosystem services levels of 16 NTU and 64 NTU in São Paulo (the first
in this region. value between April to October and November to March
respectively) at the water intake point of the River
Discussion Piracicaba. In this particular case, it was found that this
This study estimated the costs involved in both supply treatment plant had the highest costs related to the use
and demand of water quality maintenance and improve- of chemical products for water treatment, which in
ment; where we identified priority areas for supply in addition might have been correlated with the low forest
order to target watershed management measures or sup- cover of 4.3% (Reis 2004). During these same periods
port the launch of compensation schemes such as PES mean turbidity in the Cantareira system (dam) varied
or PWS. We estimated demand by assessing the willing- from 3–9 NTU respectively with the lowest costs for
ness to pay for a water quality improvement in terms of chemical products with a more elevated forest cover
turbidity by the main watershed user, the state water 27.1% (Reis 2004).
utility company (CEDAE). This approach showed to be However, with regards to the use of chemical products
helpful for the design of watershed payment schemes in for water treatment and their relative costs, it should be
other local contexts (Pagiola et al. 2010, Martinez de noted that not all chemical products are used to tackle
Anguita et al. 2011, Garcia-Nieto et al. 2013, Martin- turbidity. The main product used to flocculate suspended
Ortega et al. 2012). Adjusting off-the-shelf modelling material and reduce turbidity is aluminum sulphate, which
packages, such as SWAT, for our purposes (see for ex- has the highest share of the cost of chemicals for the treat-
ample, Martinez de Anguita et al. 2011 and Quintero ment process. The use of the other single chemical prod-
et al. 2009), was deemed inappropriate by local hydrolo- ucts is also correlated to the performance of the treatment
gists. Attributing water quality dynamics to land use process, where, for example the better the quality of the
changes is one of the most challenging issues in ecosys- final product (i.e. treated/filtered water), the lower the cost
tem services research (De Groot et al. 2010). Since we with the rest of the chemical products, namely hexafluoro-
lack an appropriate hydrological model, we have left the silicic acid, chlorine and calcium oxide.
quantification of the effects of specific land use changes Agricultural land use clearly represents the most im-
on water quality improvement in our study area for fu- portant entry point for watershed managers in the stud-
ture research. ied area, which is why we focused on quantifying the
Some limitations of this study are related to the use of OCs of changing rural land cover. It should be noted
turbidity as a water quality indicator. Land use is most that Lorz et al. (2011) demonstrated that urban settle-
likely the most relevant factor that influence turbidity ments also negatively impact water quality in terms of
changes in streams (ECI 2014). However, turbidity is turbidity, ammonium (NH+4 ) and Chemical Oxygen
subject to natural physical and biological variations even Demand (COD). Therefore, future studies should also
if the watershed is forested. The physical factors relate par- consider the costs of changes to waste water treatment
ticularly to erosion processes, landslides, and mudslides in upstream urban settlements.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 17 of 21
[Link]

Moreover, we note that the land cover classification 2010). The Pinampiro PWS programme pays for forest and
used in our analysis does not distinguish between agri- páramo protection and regeneration (Wunder and Alban
cultural production systems and underestimates the total 2008), while the Los Negros programme compensates for
land under production due to low resolution (Pedreira avoided deforestation of cloud forests (IIED 2012b).
et al. 2009; Fidalgo et al. 2008). Land use classifications Our results may prove useful for ongoing watershed con-
with higher resolution (based on SPOT 5) only exist for servation initiatives, such as the World Bank funded Rio
the upper part of the Guapi-Macacu watershed, but such Rural carried out by the State Secretary for Agriculture and
data would permit our OC analyses to be extrapolated in Livestock in the state of Rio de Janeiro (RioRural 2013).
more detail. This and many other initiatives in the Atlantic Forest re-
Comparing our OC estimates to actual transfers in gion have shown great interest in incentive-based water-
existing PWS schemes in the Atlantic Forest, we find shed management approaches. Furthermore, Brazilian
that per hectare payment values range from 10 BRL an- legislation has been supporting such approaches, such as
nually to 566 BRL monthly (Veiga and Galvadao 2011). the Brazilian National Law on Water Resources (Law
These include annual per hectare payments of 176 BRL 9433/1997), which allows for example the establishment of
in the municipality of Extrema, state of Minas Gerais; watershed service markets and permits charging for water
25–125 BRL in the watershed of PCJ in the state of São use as it is taking place in some watersheds in Rio de
Paulo; 10–60 BRL in the watershed of Guandu, state of Janeiro (Veiga 2008).
Rio de Janeiro; 80–340 BRL in the watershed Benevente, Equally relevant is the more recent Law 9985/2000
state of Espírito Santo; 80–340 BRL in the Guandu “SNUC” that adopts the protector-receiver principle,
watershed, states of Espírito Santo and Rio de Janeiro; which allows for rewards in exchange for good natural
75–563 BRL paid by the OASIS Foundation, states of resource stewardship (Strobel et al. 2007). Funding for
Espírito Santo and Paraná; and 175–577 BRL monthly incentive-based watershed management may be derived
payments in the state of Santa Catarina (Veiga and from Brazil’s “ecological” tax system, which compensates
Galvadao 2011). municipalities for conservation costs based on the value-
At the regional level in Latin America, we can com- added tax (ICMS) (Marques 2009; Veiga 2008).
pare our estimates to well-known schemes, such as the In addition, the decree No. 42029/2011 established the
Costa Rican PES programme (FAO 2007). Here, annual Payment for Environmental Services mechanism under
per hectare payments vary from 98 BRL for natural re- the Rio de Janeiro State Programme for Conservation
generation to 152 BRL for forest preservation and more and Revitalization of Water Resources (PROHIDRO).
than 2,332 BRL for new forest plantations for a time This decree has given legal status to the PES Pro-
period over five years (Ecosystem Marketplace 2010). grammes already underway in the state of Rio de Janeiro,
The National Programme for Hydrological Environmen- such as the Lagos São João Committee through Good
tal Services in Mexico (PSAH) pays between 57–90 BRL Practices Fund (Funboas) and Guandu Committee
per hectare annually depending on the forest type con- through Water and Forest Programme implemented in
sidering the OCs of land conversion (Ecosystem Market- the municipality of Rio Claro.
place 2010; Muñoz-Piña et al. 2008). The Ecuadorian This study can potentially contribute to implementation
PWS Programme in Pinampiro pays landowners around of these laws in the Guapi-Macacu watershed in a more
14–29 BRL per hectare annually (Wunder and Alban effective way with respect to social, economic and envir-
2008), while the Los Negros programme in Bolivia pays onmental aspects. Given that our study results take into
in-kind with beehive boxes for honey production account these different aspects and are based on primary
(Asquith and Wunder 2008). Agricultural OCs in our data obtained in the watershed, our study can contribute
study area thus clearly exceed the average annual payment by identifying limitations and opportunities related to eco-
in existing PWS schemes by more than a factor of 10. nomic watershed compensation programmes.
However, we note that payments made under the above
mentioned schemes may focus in other actions other than Conclusions
conversion from agriculture or pastoral lands into forest. Assessing the economic scope of incentive-based
In the case of the Atlantic Forest schemes, mostly pay- watershed management requires knowledge of both
ments are directed to protect headwater areas and restore the costs of providing additional watershed services
and conserve riparian forests. However, actions other than through land use and land cover change (service sup-
conversion from agriculture or pastoral lands into forest ply) and the willingness to pay (demand) for such ser-
can be paid for (eg. soil conservation practices). The price vices. This study has quantified these two necessary
paid to farmers in the PSAH Mexican scheme has been de- inputs to enable an informed decision-making process
cided based on the assumption that corn production would in the context of the Guapi-Macacu watershed in the
be an alternative to conservation (Ecosystem Marketplace state of Rio de Janeiro, Brazil.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 18 of 21
[Link]

With regard to supply, we found that land users’ OCs Competing interests
range between 972–5,482 BRL per hectare for cropland The authors declare that they have no competing interests.

and <100 BRL per hectare for pastures. At the sub-

Authors’ contributions
watershed level, zones that use a high proportion of pas- VRO undertook the field work, carried out the analysis and wrote the
ture but low proportion of cropland in the land use mix manuscript under her PhD thesis at the University of Leipzig. JB and RBP
supported the conception of the research design and data analysis. JB, RBP
will accordingly have low area-weighted average OCs
and UN reviewed and commented on various versions on this manuscript.
and vice versa. With regard to demand, we assessed UN provided insights on ecosystem management aspects. HG and JH
water treatment costs and found a 1% reduction in water contributed to the overall research design and gave valuable comments on
aspects related to farming systems analysis and landscape ecology
turbidity levels to be worth 15,510 BRL on average to
(respectively). All authors read and approved the final manuscript.
the water company, CEDAE.
In absolute terms, only a relatively small share of land Authors’ information
in the watershed is cropland (5,560 hectares); the area Vanesa Rodríguez Osuna is an Environmental Engineer with a [Link]. Degree
in Natural Resource Management and Technologies in the Tropics and
covered by pasture (52,374 ha) and forest (61,665 ha) is Subtropics, currently finalising her PhD in Geography at the University of
much larger. However, cropland covers a considerable Leipzig, Germany. She is a Senior Researcher at the Center for Development
amount of land with potential to provide watershed ser- Research (ZEF, University of Bonn, Germany) and the Cologne University of
Applied Sciences. Her research areas include ecosystem services, climate
vices and, as a result, high OCs become a critical limita- change and integrated land and water resources management.
tion on strategies that aim to change land use for the Jan Börner is an agricultural and environmental economist with a Masters in
benefit of water quality. International Agricultural Sciences and a PhD in Agricultural Science. He is a
Professor for Economics of Sustainable Resource Use at ZEF (University of
Nevertheless, given the spatial distribution of pastures Bonn, Germany), where he leads a research group focusing on the analysis
and high intensity agriculture in the watershed, we find of environmental and conservation policy tools for tropical forest
that payments for forest recuperation are likely to be conservation. Udo Nehren is a physical geographer with a Masters in
Engineering and a PhD in Geography. He is a senior researcher and lecturer
cost-effective in sub-watersheds with a high proportion in landscape ecology at Cologne University of Applied Sciences, Germany.
of pasture land. Some of these sub-watersheds also hap- His research areas include quaternary landscape evolution, landscape
pen to exhibit high levels of water resources vulnerability vulnerability and resilience and ecosystem management with a strong focus
on tropical environments. Rachel B. Prado is a Biologist with a PhD in
and thus represent potential priority intervention zones Environmental Engineering Science. She is currently working as a Research
for watershed management (Figure 6). In these areas, Scientist, focused on geotechnologies applied at environmental monitoring
watershed planners may reap considerable benefits from and management at Embrapa Soils, Rio de Janeiro, Brazil. Her research areas
include water resources monitoring, watershed management and water
active pasture management options, such as silvopas- ecosystem services modeling. Hartmut Gaese is an Agricultural Engineer with
toral systems or strategic fencing of riparian areas. In specialisation on farming systems in the tropics and subtropics and irrigation
addition, sustainable agricultural and soil conservation systems. He is a former Director of the Institute of Natural Resource
Management and Technologies in the Tropics and Subtropics (ITT) at
practices can bring additional benefits in terms of redu- Cologne University of Applied Sciences, Germany. Jürgen Heinrich is a
cing and avoiding sedimentation, erosion, which in turn Physical Geographer with a PhD in Geography. He is a Professor in
result in improving water quality. Geography at the University of Leipzig, Germany. His research focuses on
geomorphology, quaternary geology, soil science, plant geography,
To the extent that smallholders in the study area also geoecology and environmental research.
depend on subsistence production for their livelihoods,
land use based watershed management strategies will Acknowledgement
have to also consider food security issues, especially in The authors acknowledge financial support of the IPSWAT Programme
(financed from the German Federal Ministry of Research-BMBF) and the
the upper part of the watershed. DINARIO Project: “Climate Change, Landscape dynamics, Land use and Natural
Our study area reflects watershed management is- Resources in the Atlantic Forest of Rio de Janeiro”. We thank the support of the
sues that are typical across many parts of the Atlantic Team of Researchers of MP2 Embrapa Soils of Rio de Janeiro and DINARIO. In
our study region, we specially thank Lenilson Biazatti (from the Rural Workers
Forest region, which often have diverse agricultural Union), Nicholas Locke (REGUA Association Director), Thabta Matos de
mosaics and thus highly variable OCs. Wherever in- Mata from the City Council of Cachoeiras de Macacu, Jocemir Da Silva
tensively used cropland dominates in the vulnerable (Chief technical extension officer of EMATER in Cachoeiras de Macacu),
Demerval Pereira de Sousa (President of A.L.A.F.), technical staff from the
zones of large watersheds, land use planners will find state water utility company (CEDAE – Imunana Laranjal) and all farmers
it difficult to rely on the use of payments for water- who enriched this study with their valuable time. We are also very grateful
shed services alone. Effective watershed management for the proofreading of this text by Dr. Kylie Quinn and Meghan Doiron.
will then have to be combined with enhanced moni- Author details
toring and enforcement activities to ensure that the 1
Center for Development Research (ZEF), Department of Economic and
Brazilian Forest Law is complied with, particularly Technological Change, University of Bonn, Walter-Flex-Str. 3, 53117 Bonn,
Germany. 2Institute for Technology and Resources Management in the Tropics
with regard to riparian forests. Research designed with and Subtropics (ITT), Cologne University of Applied Sciences, Betzdorfer Str. 2,
a bottom-up approach that quantifies the potential 50679 Cologne, Germany. 3Empresa Brasileira de Pesquisa Agropecuária
compliance costs of land users can clearly help policy (EMBRAPA Soils), Brazilian Enterprise for Agricultural Research, Geotechnologies
and Environmental Monitoring, Rua Jardim Botânico, CEP 22460-000 Rio de
makers to target both incentives and disincentives in a Janeiro, Brazil. 4Institute for Geography, Department of Physical Geography,
cost-effective manner. University of Leipzig, Johannisallee 19a, 04103 Leipzig, Germany.
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 19 of 21
[Link]

Received: 5 February 2014 Accepted: 22 May 2014 php?section=water_market&page_name=crwb_market. Accessed

13 Sep 2011
Engel S, Pagiola S, Wunder S (2008) Designing payments for environmental services
in theory and practice: An overview of the issues. Ecol Econ 65(4):663–674
References EPA (2013) Total Maximum Daily Loads (TMDLs) at Work: Arizona. Committed
Angelsen A, Larsen HO, Lund J (2011) Measuring livelihoods and environmental Landowner Implements and Advocates Improved Grazing Practices in
dependence: Methods for research and fieldwork. Routledge, London Nutrioso Creek, Reducing Turbidity Levels to Attain Water Quality Standards.
Arriagada RA, Ferraro PJ, Sills EO, Pattanayak SK, Cordero-Sancho S (2012) Do United States Environmental Protection Agency, [Link]
payments for environmental services affect forest cover? A farm-level lawsregs/lawsguidance/cwa/tmdl/[Link]. Accessed 5 May 2014
evaluation from Costa Rica. Land Econ 88(2):382–399
FAO (2007) The State of Food and Agriculture: Paying Farmers for Environmental
Asquith N, Wunder S (2008) Payments for Watershed Services. The Bellagio
Services. Food and Agriculture Organization of the United Nations, Rome
Conversations. Fundación Natura Bolivia, Bolivia
Farber S, Costanza R, Childers DL, Erickson J, Gross K, Grove M, Hopkinson CS,
Balvanera P, Uriarte M, Almeida-Lenero L, Altesor A, DeClerck F, Gardner T, Hall J,
Kahn J, Pincetl S, Troy A (2006) Linking ecology and economics for
Lara A, Laterra P, Pena-Claros M, Silva Matos DM, Vogl AL, Romero-Duque LP,
ecosystem management. Bioscience 56(2):121–133
Arreola LF, Caro-Borrero AP, Gallego F, Jain M, Little C, de Oliveira XR, Paruelo
Ferreira CEG (2012) Sistema de Suporte a Decisao Espacial Aplicado a Analise da
JM, Peinado JE, Poorter L, Ascarrunz N, Correa F, Cunha-Santino MB,
Vulnerabilidade dos Recursos Hidricos na Bacia Guapi-Macacu, Master Thesis.
Hernandez-Sanchez PA, Vallejos M (2012) Ecosystem services research in
Universidade Estadual do Rio de Janeiro, Rio de Janeiro, Brazil
Latin America: The state of the art. Ecosystem Services 2:56–70
Fidalgo E, Pedreira B, Abreu MD, Moura ID, Godoy M (2008) Uso e Cobertura da
Batchelor C, Cain J, Farquharson F, Roberts J (eds) (1998) Improving Water
Terra da Bacia Hidrografica do rio Guapi-Macacu, Serie Documentos 105.
Utilization from a Catchment Perspective, vol 4. SWIM Paper. International
Embrapa Solos, Brazil, ISSN 1517–262
Water Management Institute, Colombo, Sri Lanka
Forero FA (2002) Sistemas de producción rurales en la Región Andina
Beets WC (1990) Raising and Sustaining Productivity of Smallholder Farming
colombiana: análisis de su viabilidad económica, ambiental y cultural. Grupo
Systems in the Tropics. A Handbook of Sustainable Agricultural
Sistemas de Producción y Conservación/ Instituto de Estudios Rurales (IER),
Development. A Handbook of Sustainable Agricultural Development. AgBe
Bogota, Colombia
Publishing, Holland
Bennett G, Carroll N, Hamilton K (2013) Charting New Waters. State of Watershed Gaese H (2009) Demands for Interdisciplinary Research in Human-Ecological
Payments. Forest Trends, Washington Systems. In: Gaese H, Torrico JC, Wesenberg J, Schluter S (eds) Biodiversity
Boerner J, Vosti S (2012) Managing Tropical Forest Ecosystem Services: An and Land Use Systems in the Fragmented Mata Atlantica of Rio de Janeiro.
Overview of Options. In: Muradian R, Rival L (eds) Governing the Provision of Cuvillier Verlag, Gottingen, Germany
Ecosystem Service, vol 4. Springer, pp 21–46, [Link] Garcia-Nieto AP, Garcia-Llorente M, Iniesta-Arandia I, Martin-Lopez B (2013)
environment/environmental+management/book/978-94-007-5175-0 Mapping forest ecosystem services: From providing units to beneficiaries.
Borja A, Galparsoro I, Solaun O, Muxika I, Tello EM, Uriarte A, Valencia V (2006) Ecosystem Services 4(Special Issue on Mapping and Modelling Ecosystem
The European Water Framework Directive and the DPSIR. A methodological Services):126–138
approach to assess the risk of failing to achieve good ecological status. Grieg-Gran M, Porras I, Wunder S (2005) How can market mechanisms for forest
Estuar Coast Shelf Sci 66(1):84–96 environmental services help the poor? Preliminary lessons from Latin
Calder I (2005) Blue revolution: Integrated land and water resources America. World Dev 33(9):1511–1527
management 2nd edition. Earthscan, London Guedes FB, Seehusen SE (2011) Pagamentos por Servicos Ambientais na Mata
CEPF (2001) Perfil do Ecossistema Mata Atlantica, Hotspot de Biodiversidade – Brazil. Atlanica: licoes aprendidas e desafios, vol 2, Serie Biodiversidade 42.
Critical Ecosystem Partnership Fund Ministerio do Meio Ambiente-MMA, Brasilia, Brazil
Costanza R, d’Arge R, de Groot R, Farber S, Grasso M, Hannon B, Limburg K, Haines-Young R, Potschin M (2013) Proposal for a Common International
Naeem S, O’Neill RV, Paruelo J, Raskin RG, Sutton P, van den Belt M (1997) Classification of Ecosystem Goods and Services (CICES) for Integrated
The value of the world’s ecosystem services and natural capital. Nature 387 Environmental and Economic Accounting, Report to the European Environment
(1):253–260 Agency., Available via CICES. [Link]
Daily GC, Matson PA (2008) Ecosystem services: From theory to implementation. CICES-V43_Revised-Final_Report_29012013.pdf Access 21 Jul 2011
Proc Natl Acad Sci 105(28):9455–9456 Hanson C, Talberth J, Yonavjak L (2011) Forests for Water: Exploring Payments for
de Cabral Sousa WJ (2011) Pagamento por Servicos Ecossistemicos: Mata Ciliar, Watershed Services in the U.S. South. World Resources Institute, WRI Issue
Erosao, Turbidez e Qualidade de Agua: Produtos Tecnicos 1. Projeto de Brief. Available via WRI. [Link]
Recuperacao de Matas Ciliares, Brazil., Available via Governo do Estado São Accessed 04 Jul 2014
Paulo [Link]/mataciliar Accesed 12 Jun 2010 Hawkins K (2003) Economic Valuation of Ecosystem Services. University of
De Groot RS, Wilson MA, Boumans RM (2002) A typology for the classification, Minnesota, [Link]
description and valuation of ecosystem functions, goods and services. Ecol 2008/04/valuation%2520of%[Link]. Accessed 04 Jul 2014
Econ 41(3):393–408 Holmes TP (1988) The offsite impact of soil erosion on the water treatment
De Groot RS, Alkemade R, Braat L, Hein L, Willemen L (2010) Challenges in industry. Land Econ 64(4):356–366
integrating the concept of ecosystem services and values in landscape Honey-Roses J, Acuna V, Bardina M, Brozovi N, Marce R, Munne A, Sabater S,
planning, management and decision making. Ecol Complex 7(3):260–272 Termes M, Valero F, Vega L, Schneider DW (2013) Examining the Demand fo
Dean W (1997) With Broadax and Firebrand: The Destruction of the Brazilian Ecosystem Services: The Value of Stream Restoration for Drinking Water
Atlantic Coastal Forest. University of California Press, United States of America Treatment Managers in the Llobregat River, Spain, vol 90:196.205. Elsevier BV
Dearmont D, McCarl BA, Tolman DA (1998) Costs of water treatment due to IIED (2012a) Paying for Watershed Services: An Effective Tool in the Developing
diminished water quality: a case study in Texas. Water Resour Res World, IIED Briefing Papers., Available via IIED [Link]
34(4):849–853 Accessed 27 Aug 2012
Dudley N, Stolton S (2003) Running Pure: The Importance of Forest Protected IIED (2012b) Compensation for Hydrological Environmental Services in Los
Areas to Drinking Water, The Arguments for Protection Series. World Negros Cloud Forest. Payments for Watershed Markets-Information on
Bank/WWF Alliance for Forest Conservation and Sustainable Use, United Schemes in Developing Countries., Available via IIED [Link]
Kingdom, [Link] [Link]/casestudies/Bolivia_Los_Negros_E.html. Accessed 24 Mar 2014
Accessed 10 Jan 2011 Instituto Terra Mater (2009) Relatório da Análise de Percepção Ambiental e sobre
ECI (2014) Chapter 3 – Streams. In: Total Suspended Solids and Turbidity in Pagamentos por Serviços Ambientais nas Microbacias do Moinho e Cancan.
Streams. Department of Ecology, State Washington, [Link] Produto 3. Serviços de monitoramento sócioconômico e de percepção
programs/wq/plants/management/joysmanual/[Link] accessed ambiental em microbacias piloto – “Projeto de pagamento de serviços
7 may 2014 ambientais”. Contrato no PRMC/GEF/BIRD, Piracicaba, Brazil
Ecosystem Marketplace (2010) Water Market: Mexico Payment for Hydrological López A (1997) Análise dos Custos Privados e Sociais da erosão do solo- o caso
Services. Ecosystem Marketplace: A Forest Trends Initiative. The Katoomba da bacia do Rio Corumbatai, Doctoral Thesis. Escola Superior de Agricultura
Group., [Link] “Luiz de Queiroz” (ESALQ), University of São Paulo (USP), Piracicaba
Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 20 of 21
[Link]

Lorz C, Abbt-Braun G, Bakker F, Borges P, Bornick H, Frimmel FH, Gaffron A, Pattanayak SK, Wunder S, Ferraro PJ (2010) Show me the money: Do payments
Hebben N, Hofer R, Makeschin F, Neder K, Roig HL, Steiniger B, Strauch M, supply environmental services in developing countries? Rev Environ Econ
Walde D, Weis H, Worch E, Wummel J (2011) Die Bedeutung von Policy 4(2):254–274
Landnutzungsanderungen fur ein Integriertes Wasserressourcen- Pedreira B, Fidalgo ECC, Abreu MB, Epiphanio JCN, Galvao LS (2009)
Management. Eine Fallstudie aus dem westlichen Zentral-Brasilien. . Mapeamento do uso e cobertura da terra da bacia hidrografica do rio
Fachberichte Wasserversorgung. Fachberichte Wasserversorgung, Vulcan Guapi-Macacu, RJ. Anais XIV Simposio Brasileiro de Sensoriamento Remoto.
Verlag, Essen, Germany, gwf-Wasser Abwasser Natal, INPE. pp 2111–2118
Marques FM (2009) Valoracao dos Servicos Ambientais da Floresta de Mata Perman R, Ma Y, McGilvray J, Common M (2003) Natural Resource and
Atlantica associados a Qualidade e Quantidade da Agua na APA do Sana. Environmental Economics, Third Editionth edn. Pearson Education Limited,
Curso de Pos-Graduacao em Agronomia Ciencia do Solo, Thesis. Seropedica, Available via WordPress. [Link]
Universidade Federal Rural do Rio de Janeiro (UFRRJ), Rio de Janeiro, Brazil 01/roger-perman-yue-ma-michael-common-david-maddison-james-
Martinez de Anguita P, Rivera S, Beneitez JM, Cruz F, Espinal FM (2011) A GIS mcgilvray-natural-resource-and-environmental-economics-3rd-edition-2003.
cost-benefit analysis-based methodology to establish a payment for pdf. Accessed 04 Jul 2014
environmental services system in watersheds: application to the Calan River Porras I, Grieg-Gran M, Neves N (2008) All that Glitters: A Review of Payments for
in Honduras. J Sustain For 30(1–2):79–110 Watershed Services in Developing Countries, vol Natural Resource Issues 11.
Martin-Ortega J, Ojea E, Roux C (2012) Payments for Water Ecosystem Services in International Institute for Environment and Development, Available via IIED.
Latin America: Evidence from Reported Experience. In: TEEB Conference [Link] Accessed 04 Jul 2014
2012. Mainstreaming the Economics of Nature: Challenges for Science and Postel B, Barton H, Jr T (2005) Watershed protection: Capturing the benefits of
Implementation, Leipzig, Germany., p ID143 nature’s water supply services. Nat Res Forum 29:98–108
MDA (2010) Programa Nacional de Crédito Fundiário (PNCF). Linha de Pria DA, Diederichsen A, Klemz C (2013) Pagamento por Servicos Ambientais.
Financiamento combate à pobreza rural, Manual de operações. Ministry of Uma estrategia para a conservacao ambiental nas regioes produtivas do
Agrarian Development (MDA). Secretaria de Reordenamento Agrario, Brasilia, Brasil? Sustentabilidade em. Debate 4(1):317.340
Brazil Quintana B (2012) Native Tree Species in Silvopastoral Systems: A Bioeconomic
MEA (2005) The Millenium Ecosystem Assessment. Island Press, Washington D.C Assessment in Cachoeiras de Macacu, RJ-Brazil, Master Thesis. Master of
Medeiros R, Young CEF, Pavese HB, Araujo FFS (2011) Contribuição das unidades Science awarded by Universidad Autónoma de San Luis Potosí and Cologne
de conservação para a economia nacional: Sumário Executivo. UNEP-WCMC, University of Applied Sciences. Cologne University of Applied Sciences.
Brasilia, Brazil Available. [Link]
Ministry of Environment (2005) Rio Floresta. Assistencia Tecnica e extensao [Link]%3FidT%3D258&rct=j&frm=1&q=&esrc=s&sa=U&ei=z9q2U8K-
florestal aos agricultores familiares da Mata Atlantica do Estado do Rio de DLOTmywPbw4CgCA&ved=0CCQQFjAD&usg=AFQjCNFldcTMmPE0Err5jbaX-
Janeiro. Bases Legais e tecnicas para implantacao de projetos florestais. kOJOA3FeJA. Accessed 04 Jul 2014
Ministry of Environment Niteroi, Rio de Janeiro, Brazil Quintero M, Wunder S, Estrada R (2009) For services rendered? Modeling
Montes C (2007) Del desarrollo sostenible al servicio de los ecosistemas. hydrology and livelihoods in Andean payments for environmental services
Asociacion Espanola de Ecologia Terrestre. Ecosistemas Revista Cientifica y schemes. For Ecol Manag 258(9):1871–1880
Tecnica de Ecologia y Medio Ambiente 16(3). Available via RUA.[Link] Reis LVS (2004) Cobertura Florestal e Custo do Tratamento de Águas em Bacias
es/dspace/handle/10045/7641?locale=en. Accessed 04 Jul 2014 Hidrográficas de Abastecimento Público: Caso do Manancial do Município de
Moore W, McCarl B (1987) Off-site costs of soil erosion: A case study in the Piracicaba, Doctoral Thesis. University of São Paulo (USP), Piracicaba,
Willamette Valley. West J Agric Econ 12(1):42–49 Accessed via USP. [Link]
Muñoz-Piña C, Guevara A, Torres JM, Braña J (2008) Paying for the hydrological teses/disponiveis/11/11150/tde-14122004-113308/publico/[Link]&
services of Mexico’s forests: Analysis, negotiations and results. Ecol Econ 65 rct=j&frm=1&q=&esrc=s&sa=U&ei=29u2U8e9FKviywPXg4H4CA&
(4):725–736 ved=0CBQQFjAA&usg=AFQjCNEomEW_8Ve3hF6ct4kADwlw013bCA.
Naegeli F (2010) Evaluation of Forest Fragmentation and Land Use Change Accessed 04 Jul 2014
Patterns using Remote Sensing Techniques and Field Methods, Master Ribeiro MC, Metzger JP, Martensen AC, Ponzoni FJ, Hirota MM (2009) The
Thesis. Cologne University of Applied Sciences, Abstract available via Brazilian Atlantic Forest: How much is left, and how is the remaining forest
DINARI[Link] distributed? Implications for conservation. Biol Conserv 142(6):1141–1153
20forest%20fragmentation%20and%20land%[Link]. Accessed 04 Jul 2014 RioRural (2013) Rio Rural. Programa de Desenvolvimento Rural Sustentavel em
Nehren U, Kirchner A, Sattler D, Turetta AP, Heinrich J (2013) Impact of natural Microbacias Hidrograficas. Governo do Rio de Janeiro. Secretaria de
climate change and historical land use on landscape development in the Agricultura e Pecuaria, Rio de Janeiro, Brazil, Available via [Link]
Atlantic Forest of Rio de Janeiro, Brazil. An Acad Bras Cienc 85(2):497–518 [Link]/. Accessed 15 Jul 2012
Nehren U, Sudmeier-Rieux K, Sandholz S, Estrella M, Lomarda M, Guillén T (eds) Rodriguez Osuna V (2013) Smallholder Production and Climate Risk: The Lower
(2014) The Ecosystem-based Disaster Risk Reduction Case Study and Exercise Amazon Region, Brazil. LAP Lambert Academic Publishing, Germany
Source Book. UNEP/CNRD, [Link] Russi D, ten Brink P, Farmer A, Badura T, Coates D, Forster J, Kumar R, Davidson
CASE_STUDIES_&_EXERCISES.pdf. Accessed 04 Jul 2014 N (2013) The Economics of Ecosystems and Biodiversity for Water and Wetlands.
Pabon-Zamora L, Fauzi A, Halim A, Bezaury-Creel J, Vega-Lopez E, Leon F, Gil L, Institute for European Environmental Policy (IEEP) & Ramsar Secretariat. Available
Cartaya V (2008) Protected Areas and Human Well-Being: Experiences from via TEEB. [Link]
Indonesia, Mexico, Peru and Venezuela, vol CBD Technical Series No. 36. and-biodiversity-teeb-for-water-and-wetlands/. Accessed 04 Jul 2014
Secretariat of Convention on Biological Diversity, Montreal Santos DG, Domingues AF, Gisler CVT (2010) Gestão de recursos hídricos na
Pagiola S, von Ritter K, Bishop J (2004) Assessing the Economic Value of agricultura: O Programa Produtor de Água. In: Prado RB, Turetta AP, Andrade
Ecosystem Conservation, Environment Department Paper No. 101. The World AG (eds) Manejo e Conservação do Solo e da Água no Contexto das
Bank Environment Department, Washington, United States of America, Mudanças Ambientais. Brazil, Rio de Janeiro, pp 353–376
Available via CBD. [Link] Secretary of Agriculture, Cattle Farming, Fisheries and Supply (2011) Government
[Link]. Accessed 04 Jul 2014 of Rio de Janeiro. Campanha de Vacinalção contra a febre aftosa. Cachoeiras
Pagiola S, Zhang W, Colom A (2010) Can payments for watershed services help de Macacu Region. Núcleo de Defesa Sanitária- Escritorio Cachoeiras Macacu,
finance biodiversity conservation? A spatial analysis of Highland Guatemala. Rio de Janeiro, Brazil
J Nat Resour Pol Res 2(1):7–24 Stanton T, Echavarria M, Hamilton K, Ott C (2010) State of Watershed Payments.
Paiva M, Penedo S, Kuenne A, Prado RB, Schuler AE (2011) Qualidade da An Emerging Marketplace. Executive Summary. Ecosystem Marketplace: A
Agua e Exportacao de Sedimentos em Sub-bacias dos rios Guapi- Forest Trends Initiative., [Link]
Macacu- Bioma Mata Atlantica- RJ. In: XXXIII Congresso Brasileiro de publicationID=2440. Accessed 04 Jul 2014
Ciencia do Solo. Solos nos Biomas Brasileiros: Sustentabilidade e Strobel JS, De Sousa JR WC, da Motta RS, Amend MR, Goncalves DA (2007)
Mudancas Climaticas. Embrapa, Uberlandia /Minas Gerais, Brazil, Available Critérios Econômicos para a Aplicação do Princípio do Protetor – Recebedor:
via Embrapa. [Link] Estudo de Caso do Parque Estadual dos Três Picos Serie Tecnica, vol Edicao
1/[Link]. Accessed 04 Jul 2014 11. Conservation Strategy Fund, Brazil
 Rodríguez Osuna et al. Ecological Processes 2014, 3:16 Page 21 of 21
[Link]

TEEB (2009) The Economics of Ecosystems and Biodiversity for National and
International Policy Maker. Summary: Responding to the Value of Nature.
The Economics of Ecosystems and Biodiversity (TEEB), Wesseling, Germany
TEEB (2010a) The Economics of Ecosystems and Biodiversity (TEEB). Ecological
and Economic Foundations, Earthscan, London and Washington
TEEB (2010b) Mainstreaming the Economics of Nature: A Synthesis of the
Approach, Conclusions and Recommendations of TEEB. The Economics of
Ecosystems and Biodiversity (TEEB), Malta
TEEB (2010c) TEEB for Business Report, Executive Summary. The Economics of
Ecosystems and Biodiversity (TEEB), Malta
Thomas JM, Callan SJ (2010) Economia Ambiental: Aplicacoes, Politicas e Teoria.
Cencage Learning, São Paulo
van Noordwijk M (2005) RUPES typology of environmental service worthy of
reward, RUPES Working Paper. ICRAF-Southeast Asia, Bogor
Veiga FC (2008) A Construcao dos Mercados de Servicos Ambientais e suas
Implicacoes para o Desenvolvimento Sustentavel no Brasil, Doctoral Thesis.
UFRRJ, Rio de Janeiro, Brazil
Veiga FC, Galvadao M (2011) Iniciativas de PSA de conservacao dos recursos
hidricos na Mata Atlantica. In: Guedes BF, Seehusen ES (eds) Pagamentos por
Servicos Ambientais na Mata Atlanica: licoes aprendidas e desafios, vol Serie
Biodiversidade 42, Secondth edn. Ministerio do Meio Ambiente, Secretaria de
Biodiversidade e Florestas, Departamento de Conservacao da Biodiversidade,
Brasilia, Brazil
WBI (2011) Estimating the opportunity costs of REDD+, vol Version 1.3, A training
manual. The World Bank Institute, Washington DC
WRI (2011) Mainstreaming Ecosystem Services Initiative (MESI). World Resources
Institute, Washington, United States of America, Available via WRI. http://
[Link]/project/mainstreaming-ecosystem-services
Wunder S, Alban M (2008) Decentralized payments for environmental services:
The cases of Pimampiro and PROFAFOR in Ecuador. Ecol Econ 65(4):685–698
Zimmer Y, Deblitz C, Seifert K (2009) Die globale Landwirtschaft besser verstehen
: Agrarokonomen betreiben internationales Netzwerk agri benchmark,
Forschung Report. Agribenchmark, Braunschweig, Germany

doi:10.1186/s13717-014-0016-7
Cite this article as: Rodríguez Osuna et al.: Priority areas for watershed
service conservation in the Guapi-Macacu region of Rio de Janeiro,
Atlantic Forest, Brazil. Ecological Processes 2014 3:16.

Submit your manuscript to a

journal and benefit from:
7 Convenient online submission
7 Rigorous peer review
7 Immediate publication on acceptance
7 Open access: articles freely available online
7 High visibility within the field
7 Retaining the copyright to your article

Submit your next manuscript at 7 [Link]

View publication stats