mapeo sa cuenca amazonica tnc

8/14/2019 Mapeo SA Cuenca Amazonica TNC

1/20

Project Report

Mapping Spatial Patterns of Supply andDemand of Ecosystem Services in the

Amazon Basin

The Ecosystem Services and Poverty Alleviation (ESPA) Project

A situation analysis to identify challenges to sustainable management ofecosystems to maximize poverty alleviation: Securing biostability in the

Amazon/Andes

March 2008


2/20


3/20


4/20

findings are summarized in section 3, as well as the methodological and conceptualchallenges found in the implementation of the methodology.2. Methods

The methodology focused on mapping spatial patterns of ES provision and use. Allthe analyses were performed on a grid basis using a spatial resolution of 1 km2. Themethodology has two man components: 1) mapping patterns of provision of ES and 2)identification of areas of special interest regarding the use of specific types of ES. Allthe analyses were performed on a grid basis using a spatial resolution of 1 km2. Themodels were implemented using spatial analysis tools available in ArcGIS V9.2 (Annex1).

2.1 Provision of ES

2.1.1 Terrestrial biodiversity

An index of habitat quality was constructed as a proxy of all the goods and services

provided by biodiversity at a given place. The underlying assumption is that biodiversityis related with habitat quality. Better habitat quality means better biodiversity serviceprovision. This index was based on the analysis of biodiversity threats performed byThe Nature Conservancy, South American Region (Jarvis et al. 2006). Seven sourcesof threat were considered in the analysis: grazing pressure, recent conversion,accessibility, infrastructure, conversion to agriculture, oil and gas exploration and fire(Table 1). The analysis combines these sources of threat to evaluate the degree ofdegradation or, conversely, quality for a given habitat type (Polasky et al. 2007).

Threat to biodiversity Description Data

Grazing PressureGrazing of different livestock on naturalpastures

Density of livestock

cattle, sheep andgoats

Recent Conversion Vegetation loss levelsNDVI changes 1998 2005

AccessibilityCombination of Time to populatedplace and population density

Roads, railways,rivers, land cover

InfrastructureElements of infrastructure not capturedin other layers

Dams, airports

Conversion to agricultureNatural areas surrounding existingagricultural systems

Distribution of 22major crops

Fire Frequency of fires

Satellite derived fire

events 2000 -2005Oil and gas exploration Drill sites Points of drill sites

Table 1. Sources of threat used in the analysis (Adapted from Jarvis et al. 2006).

For each threat, an index of potential degradation was calculated using a distancedecay function applied to the map depicting the current distribution and intensity of thethreat (Eq. 1). This index was corrected afterwards to take into account the type ofresponse of each ecosystem to the source of threat, and the severity of the affectationto the ecosystem given the nature of the threat.

( ) ==Y

y yrxyrrrxr DdistfwD 1 , Eq. 1

4


5/20

where xrD is the index of the overall threat to habitat in cellx, x= 1,,Xindexes all

cells on the landscape, rindexes the roster of sources of threat, wr is the weight ofthreat rvis--vis all other threats r(if theat r is a greater threat to habitat than threat r~

then rr ww ~ > ), distxy is the Euclidean distance betweenxand the place of origin of

threat y. The function frvaries across each r(e.g. it can be an exponential function of

the form ( )xyr distexp ). Finally Dyrrepresents the current intensity of threat rfromplace y(e.g. population density, number of fires occurred). Very often Dyr= 0, i.e., theorigin ydoes not produce threat r. The combined index of threat was transformed into aproxy of habitat quality (Qx) by assuming that areas under high values of potentialthreat would present a low degree of provision of ES (Eq. 2).

( )

( )( )xrxrxrxr

xDD

DDQ

maxmin

max

= Eq. 2

Two additional analyses were conducted to analyze patterns of provision of ES by

biodiversity. First, the database of species distributions hosted by NatureServe wasprocessed to create maps of species richness for birds, amphibians and vascularplants. These maps served to compare the general patterns of habitat quality withinformation on biodiversity at the species level. Second, three maps of diversity ofecological systems were created using a lattice of grids at 20km, 50 km and 100 km.The diversity of ecological systems in each grid cell of the lattices was evaluated usinga Shannon's diversity index (Eq.3). This index equals zero when the grid cell containsonly one patch, and increases as the landscape contains more types of ecologicalsystems, or the distribution among types of ecological systems becomes moreequitable (McGarigal and Marks 1994).

( )= =m

i

ii ppSDI1

ln Eq. 3

wherepi is the proportion of the landscape occupied by ecological system i. Similar tothe analysis with species distributions, these maps were used to compare patterns ofhabitat quality with regional patterns of ecosystem diversity.

2.1.2 Timber and non-timber forest products (NTFP)

A literature review was conducted to identify a set of species with importance forthe livelihoods of different users in the Amazon basin and the Andean region. Patternsof current distribution for these species were estimated by comparing points ofoccurrence available in online collection databases with The Nature Conservancy map

of ecological systems for South America (Sayre et al. 2005). In this way, a binary mapdepicting presence/absence of each species was generated. In addition, the specieswere grouped according to their main use and target group of beneficiaries into: 1)medicinal species used by local communities (subsistence), 2) medicinal species thatare being commercialized, 3) timber species for local use (subsistence), 4) timberspecies that are being commercialized, 5) fruits for local use (subsistence), 6) fruits forcommercialization, 7) fibers for local use (subsistence) and 8) game species locallyused (subsistence). The numbers of species belonging to each group are listed inTable 2. More information regarding the selection of species and collection ofdistributional data can be found in Pineda (2008) and Chiriboga (2008).

Conceptually, the provision of environmental services related to the species in each

group was estimated using a relative index ( [ ]1,0xAT ). This index represents thecurrent existence of the species (and their associated ES) as a function of past

5


6/20


7/20

Use groupNumber of

speciesMin hours Max hours

Fruits and nuts - subsistence 39 6 15 0.5

Fruits and nuts - market 18 8 20 0.5

Game spp - subsistence 30 12 15 0.5

Fibers - subsistence 3 6 15 0.5

Wood - subsistence 5 6 15 0.5

Wood - market 10 12 20 0.5

Medicinal - subsistence 37 6 15 0.5

Medicinal - market 12 8 15 0.5

Table 2. Parameters used in the model of provision of timber and NTFPs.

2.1.3 ES related to tourism and recreation

Due to time limitations, it was not possible to obtain consistent data for the wholeAmazon basin related to recreational services provided by natural ecosystems.Available information was obtained on visitation rates to Protected Areas, whichconstitute one of the main target areas for nature based recreation. A map wasproduced using the patterns of visitation within each protected area.

2.2 Demand of ES

The team developed a conceptual model to estimate spatial patterns of demand forthe timber and NTFPs identified. The model estimates the potential demand for aproduct zat spatial unitxby a harvesters based in market h (Eq. 5):

( ) ( )( )xzxzxhhxzzhxhsz ATrsadjgatherhrsadjtravelhcATIp +>= 0 Eq. 5

and

x

S

s

Z

z xhszR= = =1 1 Eq. 6

wherepzh is the market price for product zin market h, ch is the hourly wage in the nextmost profitable line of work (e.g., construction, wage labor, etc.), adjtravelhrsxhsindicates the number of hours that it will take s to go from h to sitexand back and

( )xzxz ATrsadjgatherh is a function that indicates the number of hours that it will take s

to gather product zinxand is a function of the relative supply ofzin spatial unitx(ATxz). Finally, Rxrepresents the total demand for the set of products zassociated tomarket h.

Data constraints discussed in the results section prevented full implementation ofthe model. Instead, the analysis focused on the patterns of provision of ES related tobiodiversity, timber and NTFPs in two categories of areas of special interest: 1)indigenous territories and 2) protected areas (See Section 2.4).

2.3 Future scenarios of provision of ES

Future scenarios for the provision of ES were generated based in two main sources

of data: 1) a deforestation scenario developed by IPAM for the year 2020 and 2) dataregarding road development in the Amazon region (Soares - Filho et al. 2006). These

7


8/20

two data sources were used to recalculate the following maps under the projectedscenarios of change:

Habitat quality in the year 2020 (Qx2020) calculated using the projected threatlayers of accessibility and areas of recent conversion.

Provision of timber and NTPFs using habitat quality in the year 2020 (Qx2020)and the new layer of accessibility based on projected land use and new roads.

To ensure that the scenarios of habitat quality for 2020 were comparable to thecurrent maps, the threat layers used to calculate them were normalized to themaximum values in the 2000 maps.

2.4 Provision of ES in main regions of beneficiaries

The study area was regionalized into protected areas, indigenous territories, urbanareas, and background areas (zones not belonging to any of the preceding categories)(Fig. 2a). In the case of overlap of two or more categories, priority was given to urban

areas and indigenous territories. The rationale was that each of these areas representsbroad groups of beneficiaries of the services provided by biodiversity in theAmazon/Andes region. Additionally, a map of main habitat types was generated byreclassifying the map of Terrestrial Ecoregions of the World created by the WorldWildlife Fund (Olson et al. 2001; Fig. 2b). The main habitat types identified were: 1) dryforests, 2) Guayanan ecosystems, 3) mangroves, 4) moist forests, 5) montane forests,6) montane grasslands, 7) savannas, 8) swamp forests, 9) varzea and 10) otherecosystems (marginal areas of desert and pantanal present in the region). Meanvalues within each of the combinations of habitat types and regions of use werecreated for: 1) Qx2000, 2)ATx2000(for the eight groups defined of timber and NTFPs), 3)expected change in Qx between 2000 and 2020 and 4) expected changes in theprovision of timber and NTFPs between 2000 and 2020.

Figure 2. Regions defined to create summaries of habitat quality and provision oftimber and NTFPs: a) map of protected areas, indigenous territories, urban areas, andbackground values, b) main habitat types in the study area.

8


9/20

3. Results and discussion

3.1 Current and future patterns of habitat quality in the Amazon basin

Current and future estimated patterns of habitat quality (Qx) in the study area arepresented in Figure 3. When the study area is disaggregated into regions ofbeneficiaries, it becomes apparent that habitat quality is greater in indigenousterritories for most of the general habitat types defined (Table 3). Additionally, thehabitat type that presents on average the highest values of habitat quality correspondsto the Guayanan ecosystems. On the other hand, the habitat types with the lowestvalues of habitat quality are dry forests and savannas (Table 3).

Figure 3. Current patterns and projected changes in habitat quality.

Protected

area

Indigenous

territoryBackground

Dry forest 0.789 0.809 0.662

Guayanan ecosystems 0.867 0.878 0.840Mangroves 0.763 0.879 0.766Moist forest 0.842 0.851 0.808Montane forest 0.804 0.785 0.767Montane grasslands 0.776 0.750 0.851

Savanna0.782 0.765 0.703Swamp forest 0.795 0.864 0.815

Varzea 0.807 0.850 0.813

Other ecosystems 0.719

Urban areas 0.689

Table 3. Mean values of habitat quality ( [ ]1,0xQ ) for the year 2000.

The main expected losses in habitat quality are associated to deforestation areasspatially related to new roads projected in the Amazon Basin (Fig. 3). When thesepatterns are disaggregated, the region that would experience more loss in habitatquality corresponds to areas outside indigenous territories and protected areas(background region, Table 4) along with urban areas. At the habitat type level, dry

9


10/20

forests, Guayanan ecosystems, and moist forests would experience more habitatquality loss than the other habitat types. In contrast, montane grasslands wouldexperience low levels of loss of habitat quality (Table 4).

Protected

area

Indigenous

territoryBackground

Dry forest -3.947 -2.462 -6.234Guayanan ecosystems -1.471 -1.835 -5.567

Mangroves -0.599 -0.007 -2.607Moist forest -2.917 -1.764 -5.094

Montane forest -1.017 -2.753 -1.874Montane grasslands -0.014 -0.009 -0.014Savanna -1.901 -2.581 -2.674Swamp forest -2.060 -0.147 -6.587Varzea -4.507 -0.633 -3.036

Other ecosystems -0.252

Urban areas -5.391

Table 4. Mean expected values of change in habitat quality between the years 2000and 2020.

These patterns can be explained in the context of the location of indigenousterritories, generally in areas with difficult access and still relatively isolated from themain network of transportation in the Amazon Basin. In contrast, the areas classified asbackground are closer to the main fronts of deforestation in the region. Even thoughprotected areas present intermediate values in terms of projected loss of habitat qualitycertain ecosystems in these areas (varzea, dry forests) still could experience significantimpacts due to the influence of interrelated processes of increase in accessibility and

deforestation in the future.

3.2 Current and future patterns of provision of ES related to timber and NTFPs

A map of the index of provision of ES related to timber and NTFPs was created foreach of the eight groups of species defined (See Section 2.1.2). The maps depict boththe current patterns of the index and the estimated changes under the scenarios ofdeforestation and road development used in the present assessment. The completeseries of maps is presented in Annex 3. Figure 4 is an example of the maps producedfor the group of fruit species locally used for subsistence. A common pattern across theeight groups considered is the spatial concentration of the provision of ES in moistforests and more specifically in the ecological systems present in the western portion of

the Amazon Basin (Fig. 4).

To facilitate the interpretation of the results, the mean values of the index ofprovision were summarized for all the groups destined to subsistence (Table 5) and forthe market (Table 6). The results disaggregated by region of beneficiaries and type ofhabitat are similar for both groups of species. In both cases, the highest values of theindex of supply are associated to moist forests, closely followed by areas of varzea.Montane grasslands present the lowest values relative to the rest of habitat typesconsidered. At the level of regions of beneficiaries, the index of provision presents thehighest values in indigenous territories, followed by protected areas, and finally thebackground region. From all the combinations of habitat types and regions, the highestvalues of the index are found in areas of varzea within indigenous territories for both

groups of species (Tables 5 and 6).

10


11/20

Figure 4. Current patterns and projected changes in the provision of ES ( [ ]1,0

xAT )related to fruit and nut species for local use (subsistence).

Protected

area

Indigenous

territoryBackground

Dry forest 0.118 0.080 0.020Guayanan ecosystems 0.112 0.112 0.104Mangroves 0.013 0.034 0.005Moist forest 0.369 0.367 0.314

Montane forest 0.171 0.235 0.056

Montane grasslands 0.015 0.006 0.006Savanna 0.047 0.074 0.015

Swamp forest 0.002 0.031 0.023Varzea 0.225 0.442 0.316


Urban areas 0.004

Table 5. Mean values of the index of provision of ES ( [ ]1,0xAT ) related to speciesused for subsistence.

Protected

area

Indigenous

territoryBackground

Dry forest 0.130 0.083 0.019

Guayanan ecosystems 0.153 0.151 0.118Mangroves 0.011 0.022 0.005Moist forest 0.335 0.342 0.277


Montane grasslands 0.019 0.007 0.002

Savanna 0.050 0.063 0.011Swamp forest 0.008 0.022 0.020Varzea 0.205 0.408 0.276


Urban areas 0.002

Table 6. Mean values of the index of provision of ES ( [ ]1,0xAT ) related to speciesdestined to the market.

11


12/20

The patterns of loss in the index of provision of ES are summarized in Table 7 forspecies used for subsistence, and in Table 8 for species destined to the market. Again,the patterns between these two groups are similar. Most of the expected loss in theprovision of ES is found in moist forests, followed by montane forests in the case ofspecies for subsistence, and dry forests in the case of species commercialized.Looking at regions of beneficiaries, most of the expected loss is concentrated inbackground areas, followed by protected areas and indigenous territories for bothgroups of species. The highest values of loss correspond to moist forests inbackground areas (Tables 7 and 8).

Protected

area

Indigenous

territoryBackground

Dry forest -0.825 -0.377 -0.448Guayanan ecosystems -0.094 -0.248 -0.521Mangroves -0.072 0.000 -0.023Moist forest -2.123 -1.018 -2.819Montane forest -0.339 -1.089 -0.649

Montane grasslands -0.006 -0.024 -0.009

Savanna -0.234 -0.355 -0.291Swamp forest -0.046 0.000 -0.123Varzea -0.322 -0.552 -0.814


Urban areas -0.024

Table 7. Mean values of change in the index of provision of ES related to species usedfor subsistence.

Protected

area

Indigenous

territoryBackground

Dry forest -1.420 -0.631 -0.408

Guayanan ecosystems -0.110 -0.301 -0.706

Mangroves -0.075 0.000 -0.037

Moist forest -2.145 -1.098 -2.485

Montane forest -0.407 -1.015 -0.585

Montane grasslands -0.014 -0.012 -0.003

Savanna -0.235 -0.343 -0.235

Swamp forest -0.337 -0.002 -0.202

Varzea -0.180 -0.638 -0.637


Urban areas -0.013

Table 8. Mean values of change in the index of provision of ES related to speciescommercialized.

These results provide broad regional patterns for the identification of priority areasin terms of the management of biodiversity with the purpose of securing the futureprovision of its associated ES. Even though background areas present the lowestcurrent levels of provision of ES, these areas are expected to experience the highestloss under processes of deforestation in the short term. A related argument can beconstructed for moist forests which currently present high levels of provision of ES andhigh levels of expected loss. The human groups that depend of biodiversity-related ES

in these areas represent appropriate targets for the development and implementationof research and management initiatives aimed at reducing the loss of habitat qualityand securing the flows of goods and services from natural ecosystems.

12


13/20


14/20


15/20

The values of ecosystem diversity disaggregated by habitat type and regions ofbeneficiaries confirm the patterns observed above (Table 10). The highest levels ofecosystem diversity are associated with montane grasslands, followed by montaneforests and savannas. The lowest levels are present in moist forests. Consideringregions of beneficiaries, the differences are less evident, with indigenous territoriespresenting higher values than protected areas and background regions.

Protected

area

Indigenous

territoryBackground

Dry forest 0.661 0.631 0.760

Guayanan ecosystems 0.679 0.591 0.697

Mangroves 0.517 0.628 0.633

Moist forest 0.391 0.411 0.371



Savanna 1.055 0.896 0.813

Swamp forest 0.521 0.756 0.398

Varzea 0.431 0.637 0.573


Urban areas 0.788

Table 10. Mean values of habitat diversity (measured using the Shannon's diversityindex) for the different combinations of habitat type and region of beneficiaries.

Patterns of species richness in the study area are depicted in Figure 7. Thepatterns of species richness vary per group of species (Tables 11 13). For example,while the richness of amphibian species is highest in varzea and moist forestecosystems, mammal species are concentrated in Guayanan ecosystems. A consistent

pattern for the three groups of species is that they present the lowest richness inmontane ecosystems.

Figure 7. Species richness

15


16/20

Protected

area

Indigenous

territoryBackground

Dry fores t 43.180 46.755 36.023Guayanan ecosystems 55.042 29.828 34.300

Mangroves 52.871 31.149 33.293

Moist forest 66.595 63.852 66.924Montane forest 14.368 22.750 7.845Montane grasslands 6.570 6.432 3.834

Savanna 44.708 42.314 36.499Swamp forest 52.385 37.568 42.562

Varzea 64.596 80.255 79.671


Urban areas 40.958

Table 11. Mean values of amphibian species richness

Protected

area

Indigenous

territoryBackground

Dry fores t 137.101 134.772 126.759Guayanan ecosystems 174.824 177.146 176.435

Mangroves 171.621 148.225 126.676Moist forest 165.813 161.684 163.568

Montane forest 157.544 156.732 122.216Montane grasslands 125.291 126.824 79.607

Savanna 144.067 152.169 132.946Swamp forest 182.810 159.201 174.747

Varzea 157.249 169.130 167.294

Other ecosystems 93.497Urban areas 141.277

Table 12. Mean values of mammal species richness

Protected

area

Indigenous

territoryBackground

Dry forest 396.072 382.541 341.607

Guayanan ecosystems 396.551 449.427 442.948

Mangroves 483.596 314.718 348.261

Moist forest 444.143 429.453 446.771

Montane forest 395.371 422.389 262.091

Montane grasslands 201.696 189.776 139.023Savanna 387.651 418.810 346.640

Swamp forest 513.241 347.485 444.704

Varzea 443.135 469.363 472.431


Urban areas 363.405

Table 13. Mean values of bird species richness

Total species richness is higher in varzea, swamp forest, and moist forest ecosystems(Table 14). The lowest values are present in montane ecosystems. These resultingpatterns are highly influenced by the distribution of bird species, since this groups has

16


17/20

more species than the other two combined (mammals and amphibians) in the datasetused.

Protected

area

Indigenous

territoryBackground

Dry forest 576.353 564.067 504.420

Guayanan ecosystems 626.416 656.401 653.683Mangroves 708.088 494.092 508.229

Moist forest 676.551 654.989 677.263

Montane forest 567.283 601.871 392.154


Savanna 576.427 613.293 516.085

Swamp forest 748.436 544.254 662.013

Varzea 664.980 718.748 719.396


Urban areas 545.641

Table 14. Mean values of total species richness

3.5 Methodological issues and knowledge gaps

The development of this assessment presented important conceptual andmethodological challenges. In this context, the methodology and results presentedshould be regarded more as a framework to identify further goals for research than asa guide for intervention. The following discussion specifies the main knowledge gapsidentified during the execution of the study.

3.5.1 Evaluation of the state of biodiversity

Indexes that synthesize the influence of several factors on the status of biodiversityhave been used to provide spatial perspectives at global to regional scales (e.g.Sanderson et al. 2002, Alkemade et al. 2006). This strategy combines the advantagesof being relatively easy to implement with the presentation of results in an intuitiveformat. However, these measures of threat, pressure or habitat quality also share anumber of important limitations that should be taken into account when using andinterpreting the results of the analyses.

The calculation of habitat quality performed in this study takes into account sevenmajor sources of threat to the integrity of natural ecosystems. Even though thesethreats cover major processes that affect biodiversity, it is not difficult to think in

additional factors that are likely to have important effects in the study area. These couldinclude the effects of climate change on the structure and composition of existingecosystems, the effects of landscape fragmentation, and associated biotic processessuch as local extinctions or the spread of invasive species. The impacts of thesesources of habitat degradation should be evaluated, especially as they may affectspecies and communities that are critical for the sustainment of local livelihoods (e.g.game species).

An additional issue is the static nature of the methodology implemented. Most ofthe threat factors considered in the calculation of habitat quality present some level ofinteraction (e.g. recent conversion and conversion to agriculture) which is not explicitlyconsidered in the model. Furthermore, the projection of the index of habitat quality to

the year 2020 only included the threats from deforestation and the expansion of theroad network. The estimation of future patterns for the other sources of threat is

17


18/20

complicated due to the paucity of information, mismatch in the scale of data sources,and lack of appropriate methodological frameworks to generate future scenarios forthese factors. In spite of these limitations, the maps of habitat quality generated doprovide a robust overview of the status of biodiversity in the Amazon/Andes region.

3.5.2 ES related to timber and NTFPs

The key source of data for the generation of patterns of provision of ES related totimber and NTFPs is the distribution of species across the study region. In this study asimple approach was used to estimate the distribution of the species selected. It wasassumed that if a species had been registered in a given ecological system, thespecies would be potentially present in the whole extension of that ecosystem. In otherwords, the basic assumption is that the ecological system becomes a valid proxy of thefundamental niche of the species. This approach can lead to commission errors inareas where a given species is considered present but it is not, or omission errorswhen species are present in an ecological system but have not been registered.Significant improvements to the methodology could be attained by applying niche

modeling techniques to generate more accurate estimations of the distribution of keyspecies in the study area (Guisan and Zimmermann 2000).

The selection of used species in the Amazon basin also presented importantchallenges. The list of species selected should be regarded as a sample of theuniverse of species actually being used in the region. Data was especially difficult tocompile for some of the defined groups of use (Table 2) such as fibers or game speciesused for subsistence. Future assessments should consider generating more exhaustivelists of species and to incorporate regional variations in terms of which species areused, the intensity and purpose of their use, and their main markets of destination.

The model used to map the index of provision of ES related to timber and NTFPs

required the estimation of a set of parameters for which empirical data was not readilyavailable (Eq.4, Table 2). For example, the thresholds of accessibility used in themodel represent more referential values used to generate regional patterns rather thanabsolute values defined on an empirical basis. Other studies have acknowledged thesubjective nature associated with the definition of such thresholds, as accessibility is arelative concept that varies among cultural groups, types of ecosystems (e.g. montaneecosystems vs. lowland moist forests), and patterns of resource use (Sanderson et al.2002).

The model proposed in Eq. 5 has the advantage of providing a spatially explicitestimation of the demand associated with a species using a consistent scale (i.e.monetary value). However, during the implementation of the methodology it became

apparent that the data gaps needed to run the model were too big to be filled within thetimeframe of the project. However, the model still can be used to identify two importantknowledge gaps related to the demand of ES in the Amazon basin. First, it is required abetter discrimination of groups of users and their spatial distribution. This would lead toa better discrimination of patterns of resource use in heterogeneous areas such as thebackground region defined in this study (Fig. 2). In addition, this could facilitate theidentification of important market centers for the species identified as critical in theprovision of ES. The original idea of the model was to generate data about accessibilityand the related costs of transportation associated to species used locally, destined toregional markets, and exported to national or international markets.

A second knowledge gap identified is related to the specific areas that are providingES for these groups of users. A useful analog would be the delineation of a watershedproviding water for a defined group of users (e.g. a city). Of special importance are

18


19/20

areas critical for groups of users which are highly dependent on the ES provided bybiodiversity (e.g. through game, food or medicinal species used for subsistence).Furthermore, given global trends of economic integration it is important to consider theeffects of ex-situ users of goods and services provided by Amazonian biodiversity. Theregions of beneficiaries used in this study (Fig.2) constitute a first step towards thedisaggregation of spatial patterns of supply and demand of ES in the Amazon basin.

3.5.3 ES related to tourism and recreation

Mapping ES related to tourism and recreation was difficult due the scarcity ofconsistent data for the countries in the study region. The only dataset availableconsisted of visitation rates for protected areas in the Amazon basin. It becomesapparent the necessity to generate a more sophisticated analytical framework to relatethe ES provided by biodiversity with current or potential visitation patterns in theseareas. Such model would require basic data related to:

a. Visitation rate data, parsed to domestic and international.

b. Travel times. Travel times should by adjusted to represent the country oforigin for different groups of visitors.

c. Number of activities available in the area.

d. Site quality (e.g. overall landscape diversity, presence of charismaticmegafauna).

e. Cultural attractions (e.g. village visits, community management)f. Infrastructure (accommodations, trails, etc.).

g. Competition (e.g. distance to other parks, degree of similarity withneighboring protected areas).

Furthermore, the potential of regions outside national systems of protected areasshould be taken into account. This becomes an important goal given the recent trendtowards the implementation of integrated conservation and development projects thatpromote tourism as a key management strategy to attain the conservation ofbiodiversity and the alleviation of poverty at the community level. The relationshipbetween provision of tourism and recreation, biodiversity conservation, andimprovement of local livelihoods should be defined empirically for the Amazon basinand within each country.

3.5.4 Relationships between biodiversity and the provision of ES

It becomes evident from the results that patterns of habitat quality, provision of ES,species richness and habitat diversity do not necessarily coincide in space. In certain

cases, areas with high biological diversity in terms of species richness (e.g. moistforests) also present high levels of provision of ES related to timber and NTFPs.However, these areas do not necessarily present high diversity at the ecosystem level.This condition could be different for other types of ecosystem services such aspollination, water provision, and hydrological regulation. It is likely that the structure andcomposition of the landscape becomes a critical factor for the sustenance of theseservices. Establishing the links between biodiversity at different levels (i.e. species,ecosystems) and the provision of ES in the Amazon basin requires further researchand the collection of appropriate datasets.

4. References

Alkemade R., M. Bakkenes, R. Bobbink, L. Miles, C. Nellemann, H. Simons, T.Tekelenburg. 2006. GLOBIO 3: Framework for the assessment of global terrestrial

19


20/20

biodiversity Pp. 171 - 186 in: MNP (2006) Integrated modeling of globalenvironmental change. An overview of IMAGE 2.4. MNP, Bilthoven, TheNetherlands.

Chiriboga, C. 2008. Levantamiento de informacin sobre el uso, destino y distribucinde epsecies maderables y no maderables en la cuenca Amaznica (y reginAndina adyacente). Reporte Tcnico. TNC-Ecuador.

Fa, J.E., C.A.Peres and J.Meewig. 2002. Bushmeat Exploitation in Tropical Forests: anIntercontinental Comparison. Conservation Biology16:232-237.

Guisan A. and N. E. Zimmermann. 2000. Predictive habitat distribution models inecology. Ecological Modelling 135:147-186.

Jarvis, A., S.E. Castano, G. Hayman, S. Gebhardt, E. Guevara, M. Castro, J. Touval, L.Sotomayor, et al. 2006. Project Report: TNC Threats Assessment Version 1.1. TheNature Conservancy and Centro Internacional de Agricultura Tropical.

McGarigal, K., and B. J. Marks. 1994. Fragstats: Spatial pattern analysis program forquantifying landscape structure. Reference manual. For. Sci. Dep. Oregon StateUniversity. Corvallis Oregon 62 p.+ Append.

Millennium Ecosystem Assessment. 2005. Ecosystems and Human Well-being:Synthesis. Island Press, Washington, DC.

Olson, D. M., and E. Dinerstein. 2002. The global 200: Priority ecoregions for globalconservation.Annals of the Missouri Botanical Garden 89:199-294.

Olson, D. M., E. Dinerstein, E. D. Wikramanayake, N. D. Burgess, G. V. Powell,E. C.Underwood, J. A. D'amico, I. Itoua, H. E. Strand, J. C. Morrison, C. J. Loucks, T. F.Allnutt, T. H. Ricketts, Y. Kura, J. F. Lamoreux, W. W. Wettengel, P. Hedao, and K.R. Kassem. 2001. Terrestrial ecoregions of the world: a new map of life on earth.Bioscience 51:933-938.

Pineda, J. 2008. Georeferenciacin de especies tiles de la Amazona en los SistemasEcolgicos del mapaAmazon LAC Ecological Systems. Reporte Tcnico. TNC-Ecuador.

Polasky, S., G. Daily, P. Kareiva, T. Ricketts, E. Nelson, J. Regetz, G. Mendoza, H.Tallis, R. Naidoo, V. Adamowicz, D. Cameron, K. Chan, and R. Shaw. 2007. TheIntegrated Valuation of Ecosystem Services and Tradeoffs (InVEST) Tool: A

modeling suite developed by the Natural Capital Project to support environmentaldecision making. Natural Capital Project, Stanford University. Stanford, CA.

Sayre, R., J. Bow, and C. Josse. 2005. Modeling NatureServes Latin America andCaribbean Ecological Systems throughout South America. The NatureConservancy and NatureServe.

Soares-Filho, B.S., D.C. Nepstad, L.M. Curran, G.C. Cerqueira, R.A. Garcia, C.A.Ramos, A. Mcdonald, P. Lefebvre and P. Schlesinger. 2006. Modeling conservationin the Amazon basin. Nature 440:520-523.

mapeo sa cuenca amazonica tnc

Documents