Functional Value of Biodiversity – Overlay Analysis



Pantropic Analysis of the

Functional Value of Biodiversity

Phase I

Final Draft Report

(version 2)

Kate Sebastian1

Stanley Wood2

Report produced as the final Phase I output of the Project “Functional Value of Biodiversity Pantropic Analysis” undertaken through a partnership of the World Bank (DECRG) and the ASB (Alternatives to Slash and Burn) Consortium of CGIAR-affiliated research organizations and their local partners and funded in part by the Bank Netherlands Partnership Programme (BNPP).

1 GIS Analyst/Consultant

2 Senior Scientist, International Food Policy Research Institute

BNPP Project

Pantropic Analysis of the Functional Value of Biodiversity

The primary objective of the BNPP Project is to look at the means, at various scales, of assessing the functional contribution of biodiversity conservation to rural livelihoods. The operational focus is on locating those areas where one might expect to find synergies between conservation efforts and hydrological benefits for local populations.

The project goals comprise 1) to critically review existing and available global datasets that could be used to identify the extent of a pantropical area of biodiversity interest (or ‘Focus Area’); 2) to map and analyze the relationships among the defined Focus Area and other globally available data to define areas of hydrological concern as well as to assess the magnitude of populations living in these areas; 3) to provide a framework and collection of databases for broadly identifying critical or priority areas and 4) to work with DECGR’s GIS unit, to determine the value of performing such analyses at a pantropic scale.

The Phase I activities described here involved three clusters of activities:

1. Delineation of the Pantropical Focus Area based on global climatological data.

The Focus Area was first defined using climate variables. The delineation of the Focus Area allowed some basic measures of its relative importance to be addressed:

i. What share of the area and population of the tropics falls within the Focus Area?

ii. What is the regional and national breakdown of the Focus Area by area and population?

2. Characterization of the Pantropic Focus Area.

These activities defined the intrinsic properties of the Focus Area and their spatial variability in terms of land cover, hydrology and biodiversity. These activities addressed the following questions:

i. What is the distribution of land cover within the Focus Area?

ii. How would the Focus Area change if it were defined on hydrologically-relevant spatial units? What share of the area and population of the tropics falls within such units?

iii. What is the coincidence of the Focus Area and tropical forest biomes? To what extent have these forests been converted?

iv. Can we improve our representation of the physiographic characteristics of the Focus Area? What is the distribution of physiographic characteristics within the tropical forest biomes?

v. How well do the WWF Ecoregion boundaries correspond with the boundaries of the Focus Area?

3. Identification of “Critical Areas”

Having assessed the broad heterogeneity of biophysical conditions within the Focus Area, questions of critical sub-areas were addressed. This involved identifying hydrological and biodiversity measures representing value or threat, that might therefore be of priority in assessing the potential benefit of interventions. This is tricky at the Pantropic scale, and much of the planned Phase II activity will be to validate and improve such measures through 1) the incorporation of improved and/or new global datasets (e.g. rural populations) and 2) comparative assessments with richer regional datasets for Central America and Mountainous South-East Asia.

The questions posed in Phase I regarding potentially “critical” areas were:

i. What urban populations might be at risk from significantly altered hydrological function as a consequence of upstream land use change?

ii. Where, within the Focus Area, are the most important areas of biological diversity?

iii. What practical measures of biodiversity “value” are available for these critical areas?

iv. Can we identify forest margin areas using existing coarse resolution satellite data?

The remainder of the report contains question-by-question sections grouped within each of these three activity areas. The final section summarizes the major findings and recommendations with regard to proposed Phase II activities.

Section 1

Delineation of the Pantropic Focus Area

What share of the area and population of the tropics falls within the Focus Area?

The first stratification of the BNPP/ASB pantropical region was ecological. This stratification was based on agroclimatic factors derived from temperature and rainfall (Figures 1-3 show the progress from Agroclimatic Zones to the Focus Area). Using these data we defined the Focus Area as the Warm humid and subhumid tropics based on the following criteria:

Tropics – areas with year-round average monthly temperature adjusted to sea level of greater than 18 degrees Celsius

Humid – average length of growing period at least 270 days per year

Subhumid – average length of growing period 180 – 269 days per year

Warm – mean daily temperature during the growing season greater than 20 degrees Celsius

As shown in Table 1, the humid and subhumid tropics comprise 29% and 25% of the tropics, respectively. The majority of these regions have warm climates (92% and 90%, respectively). By eliminating the cooler areas from the focus area we are able to focus on regions that exhibit similar biophysical potential as well as greater spillover potential related to policies and technical innovations. The resulting area, defined here as the Focus Area, covers approximately 24 million square kilometers, just under 50 percent of the tropics and approximately 19 percent of total global land area (excluding Antarctica). Approximately 1.1 billion people live in the area – this represents over 56 percent of those living in the tropics and approximately 19 percent of total global population.

Data sources:

Tropics: Agroclimatic Zones from FAO/IIASA (2000) Global Agro-ecological Zones CD v1.0 based on University of East Anglia (UEA) monthly time series climate data (UEA 1998 ).

Length of Growing Period: Agroclimatic Zones from FAO/IIASA (2000) Global Agro-ecological Zones CD v1.0 based on University of East Anglia (UEA) monthly time series climate data (UEA 1998 ).

Temperature: Modeled using FAO/IIASA Thermal Zone surface from FAO/IIASA (2000) Global Agro-ecological Zones CD v1.0 and rules for defining major climate zones (FAO 1981 Report on the Agro-Ecological Zones Project)

Population: Two global population datasets were used in this analysis. The Global Population of the World (GPW) dataset (CIESIN 2000) is a raster dataset at a resolution of 2.5 arc minutes (approximately 5x5 km) which contains population count and density for 1995 (Figure 4). The GPW provides estimates of the population of the world in 1990 and 1995, both population counts (raw counts) and population densities (per square km) are available. The gridded data is drawn from over 127,000 administrative units (approximately 60,000 outside of the US). The national totals of these disaggregated data have been harmonized with United Nations estimates for those years. The LandScan Population dataset (ORNL 2000) is a raster dataset compiled at a resolution of 30 arc seconds (approximately 1x1 km) containing population estimates for 2000 (Figure 5). LandScan 2000 was developed as part of Oak Ridge National Laboratory (ORNL) Global Population Project for estimating ambient populations at risk. Census counts (mainly at sub-national level) were apportioned to each grid cell based on probability coefficients, which are based on proximity to roads, slope, land cover and nighttime lights. Although the Landscan data has a higher output resolution than the GPW, it is based on significantly lower resolution input data (approximately 69,000 administrative units with only 10,000 outside of the US) (Deichmann 2001).

Note: Most of the population estimates for this study were derived from GPW. In some cases, however, estimates were made using both datasets and a range was given or an average reported. This is noted on each the table.

*per GPW (CIESIN 2000)

What is the regional and national breakdown of the Focus Area by area and population?

Since there are many regional differences in agroclimatic conditions and since many policy and investment decisions are made at the regional or country level the Focus Area was further divided into sub-regions based on country level administrative boundaries. This perspective helps us get a sense of the relative importance of the Focus Area by region (and country) in terms of the share of total areas as well as the share of total population. The 5 subregions are (see Figure 6):

Mesoamerica and the Caribbean

South America

Sub-Saharan Africa

Mainland Asia

Insular Southeast Asia (Peninsular Malaysia, Indonesia, Philippines & PNG)

The majority of the focus area is in South America (44 percent) with 33 percent falling in Sub-Saharan Africa (Table 2 and Figure 7). The share of total population in the focus area is much lower for South America (11 percent) compared to the more densely populated regions of Mainland Asia (42 percent), Insular Southeast Asia (21 percent) and Sub-Saharan Africa (20 percent). The large discrepancy in area versus population share for South America is primarily because a large portion of the focus area falls in the sparsely inhabitated tropical forests of Brazil which alone accounts for 28 percent of the total focus area (see Table 3). The warm humid tropics follow this same pattern with South America and Sub-Saharan Africa containing most of the area (51 and 25 percent, respectively) but Insular Southeast Asia having a majority of the population (41 percent) followed by Sub-Saharan Africa with approximately 22 percent. The warm sub-humid tropics are equally as skewed with most of the area in Sub-Saharan Africa or South America (43 and 35 percent respectively) but most of the population living in Mainland Asia (56-58 percent). Even at a highly aggregated regional level, these statistics highlight the need to look at both the area and populations affected by policy decisions and conservation measures.

Table 3 shows all countries with five percent or greater area within the focus region. As noted, Brazil accounts for 28 percent of the focus area with the next largest share being Indonesia with 7 percent.

* Note: The population estimates by region are given for both GPW and LandScan data. The difference in the estimates (at this level of aggregation) are minimal (0-2 percent) thus instilling a level of confidence in the use of either data set for regional or possibly sub-regional analysis. Further analyses at a more disaggregated level will more than likely increase the differences but at the pan-tropic level either dataset is appropriate for analysis.

Data sources:

Country boundaries – ESRI 1999.

Regional boundaries – derived from the country boundaries.

Population 1995: Global Population of the World (GPW) CIESIN 2000.

Population 2000: LandScan Population Project (ORNL 2000).

Section 2

Characterization of the Pantropic Focus Area

What is the distribution of land cover within the Focus Area?

An intersection of land cover data and the ecologically defined Focus Area by region highlighted that, as expected, the majority of the Focus Area is classified as forest (51 percent) particularly in the humid regions (69 percent) with substantially less forest cover in the sub-humid regions (29 percent). The next largest shares are in the grassland and agriculture/other mosaic classes with 19 and 15 percent, respectively (see Table 4). As shown in Figure 8, the sub-humid areas of each sub-region have significantly higher shares of mosaic classes than the humid. This is particularly true in South America where 51 percent of the subhumid region is classified as agricultural mosaics compared to 12 percent for the sub-humid areas. In Asia there was not a significant difference between the humid and subhumid regions with the mosaics comprising around 20-25 percent of each for Mainland Asia and Insular Southeast Asia.

The land cover data used for this analysis was the Global Land Cover Characteristics Database (GLCCD 2000). This classification is based on an interpretation of the 30 arc second (approximately 1km) resolution advanced very high resolution radiometer (AVHRR) satellite imagery consolidated into monthly global composites for the period April 1992 to March 1993. Scientists at USGS Eros Data Center (EDC) and the University of Nebraska-Lincoln identified approximately 200 Seasonal Land Cover Regions (SLCRs) per continent by identifying areas that demonstrated similar landcover associations, physiographic characteristics, distinctive patterns of biomass production, such as the onset, peak, and duration of greenness (GLCCD 1998; Loveland et al. 2000). This interpretation captures both spatial and seasonal variations in vegetation cover over the observation period. The most frequently used land cover map is the IGBP Land Cover classification map which is an aggregation of the 961 SLCR classes into 17 land cover classes (IGBP 1998).

For the BNPP/ASB pantropic analysis we hoped to gain an understanding of where agriculture is currently taking place but more importantly where are the areas of mixed forest/agricultural use and the margin areas between agricultural and forest lands. The IGBP classification is less than ideal for these purposes since only 2 of the 17 classes explicitly recognize agriculture, the ‘Cropland’ class in which more than 60 percent of the area is interpreted as containing cropland, and cropland/natural vegetation mosaic areas that contain 40-60 percent cropland. Fortunately the SLCR classification system gave some scope for refining the data interpretation in order to explicitly recognize occurrences of agriculture that occupy less than a dominant share (60 percent) of a SLCR class. This process was first used, as part of the Pilot Analysis of Global Ecosystems (PAGE) Agroecosystems study, to define the global extent of agriculture (Wood et al. 2001) and was further refined for this study to better isolate the agriculture/forest mosaics from the other mosaic classes.

Figure 8

How is the focus area defined based on hydrologically-relevant units? What share of the area and population of the tropics falls within such units?

The second stratification approach was based on hydrology. One of the goals of this analysis was to gain an understanding of the effects that conservation efforts would have on hydrological function and also to determine how many people would be affected by these policy decisions. The hydrological stratification was based on the basins from the Hydro1k Elevation Derivative Database (USGS EDC 2001) and an indicator of terrain type based on roughness derived by A. Nelson (2001) using the methodology from Meybeck et al. (2001).

The Hydro1k basins were overlaid with the ecologically defined Focus Area. Any basin that had any portion within the boundaries of the Focus Area was included in the analyses (see Figure 10). These drainage basins were defined by USGS using the concepts first described by Otto Pfafstetter (Pfafstetter 1989) (Verdin 2001). The system is based on the topography of areas drained on the Earth’s surface and the typology of the resulting hydrographic network. The basins and interbasins are identified by beginning at the mouth of a river and working upstream to determine, using a DEM, the area that is drained by that river. Along the way, it is important to differentiate between the main stems and the tributaries. For the continental basin delineation the four largest tributaries are identified according to the criterion of area drained. These are assigned values 2, 4, 6 and 8 and the interbasins are assigned the odd numbers, with the remaining area receiving a value of 0, thus each continent is comprised of 10 basins (values 0-9: Pfafstetter coding) (Verdin 2001: p.1-3).

From a hydrological perspective we considered it useful to divide basins into their “upper” and “lower” parts. Upper watershed areas are those that occupy the highest elevation ranges, relative to the elevation of the watershed outflow point, and are normally associated with more steeply sloping terrain. Lower watersheds, on the other hand, are those areas whose elevation is low relative to the watershed outflow, and where gradients tend to be gentler. The characterization is useful in distinguishing between areas that tend to erode soil and areas where sedimentation takes place. In upland areas, overland flow energies are high and significantly greater potential for soil erosion exists. In lowland areas, by contrast, overland flow energies are low and erosion is significantly less important. Furthermore, since river channel gradients are low, flow velocities decline and water-borne sediment from upper watershed is often deposited. Though more controversial, there is also evidence that flood frequency and intensity are relative higher in upland areas because of the decreased “time of concentration” of flow from the point of incidence of rainfall to the river channel. Thus, watersheds with higher shares of “upper” watershed terrain will be more erosive and likely more prone to “flash-flooding”.

The upper-lower classification was performed using the terrain typology developed by Meybeck et al (2001), and applied globally for the purposes of the study by Nelson (2001). The global terrain typology identified nine terrain types, based on roughness, ranging from plains to high mountains. These terrain types were aggregated into two classes: lower terrains (all plains, lowlands and low/middle plateaus); and upper terrains (all mountains, and high plateaus) (see Figure 11).

The basins and terrains were combined to create a composite indicator of basin and roughness as shown in Figure 12. Table 5 shows the area and population shares for each region by basin and terrain type[1]. The first thing to notice is how much more land area this stratification approach encompasses in comparison to the first stratification. This should be an important determinant when choosing which stratification variables are best suited for the pantropic analysis in phase II. The basins approach encompasses almost of half of the worlds land surface (excluding Antartica) and includes over two-thirds of the worlds population. This is most likely too broad an area for the overall analysis but it does help determine the broader impact, in terms of area and numbers of people, of policies and changes in land management within the Focus Area particularly from a hydrological perspective.

Figure 13 depicts the share of total basin area by terrain type for the 24 basins within the BNPP/ASB Focus Area. The basins range from those, such as the mouth of the Amazon (SA5), with no area in upper terrains to those, such as the basins encompassing the Andes down to the west coast of Central and South America (MAC9 & SA1), with almost 60 percent of the basin area falling in the upper terrain. As expected, catchments become more rugged, erosive, and prone to flash flooding towards the right-hand side of this figure.

Table 5

Figure 13 Area share of Basins by Upper and Lower Terrain type

Data sources:

Basins: Hydro1k Elevation Derivative Database (USGS EDC 2001).

Terrain: Roughness Indicator at a resolution of 0.04167 dd (Nelson 2001) based on 30 arc second resolution elevation data.

Elevation: GTOPO30: Global 30 Arc Second Elevation Data Set (USGS 1998).

What is the coincidence of the Focus Area and Tropical Forest Biomes? To what extent have these forests been converted?

The third stratification was based on biodiversity. The data used was the World Wildlife Fund’s (WWF) Terrestrial Ecoregions database (WWF-US 2001a). The WWF Ecoregions database was designed with conservation of ecoregion function as the primary goal. Ecoregions are defined as ‘a relatively large area of land or water that contains a geographically distinct assemblage of natural communities …. [which] share a large majority of their species, dynamics, and environmental conditions (WWF-US 2001a).’

The ecoregions database is a hierarchical database consisting of three layers: Realms, Biomes and Ecoregions. The realms are continental-scale biogeographic regions that contain distinct groups of plants and animals. They are differentiated according to geologic and climatic history (see Figure 14).

The Biomes are broad groupings of ecoregions that have similar:

• climatic regimes

• vegetation structure

• spatial patterns of biodiversity

• minimum requirements and thresholds for maintaining certain biodiversity features

• sensitivities to human disturbance.

(WWF-US 2001a).

The terrestrial biomes that have any portion falling within the Focus Area are shown in Figure 15 that also shows the tropical Realms. Including all tropical biomes broadens the focus area extensively to an area that cover 34 percent of the earth’s land surface (excluding Antartica). For the purposes of this study it is more appropriate to focus on the terrestrial forest biomes as shown in Figure 16. It should be noted that this stratification alters the extent of the focus area significantly particularly in Sub-Saharan Africa which has a large share of tropical and subtropical grasslands, savannas and shrublands. It also decreases the total area of interest to 15.3 million square kilometers with a population of approximately 850 million people (from approximately 24 million square kilometers and 1.1 billion people within the Focus Area) (see Table 6).

We identified the proportional area of trees found within the forest biomes as an indicator of the extent of conversion. This proxy was calculated using the Percent Tree Cover database. These data, interpreted by scientists at the University of Maryland, are based on the same detailed satellite data that underpins the Global Land Cover Characteristics Database used here to describe land cover. The satellite data was used to derive additional vegetation characteristics including woody vegetation, defined as mature vegetation whose approximate height is greater than 5 meters (DeFries et al. 2000). The resulting map shows the percent tree cover (ranging from 0 – 80 percent) for each 1km resolution mapping unit (see Figure 17).

Figure 18 shows the extent of conversion by region for the two tropical forest biomes. Mainland Asia shows more conversion in both biomes than any other region and Sub-Saharan Africa shows a high level of conversion in the drier forest biome. Globally, all of the regions show higher levels of conversion in the dry broadleaf forests than in the moist broadleaf forest.

Data used:

WWF Terrestrial Ecoregions (WWF-US 2001a).

UMD Percent Tree Cover Database (DeFries et al. 2000).

Figure 14: Biogeographical Realms

[pic]

Can we improve our representation of the physiographic characteristics of the Focus Area? What is the distribution of physiographic characteristics within the tropical forest biomes?

Table 7 shows the percent area for each biome by region by terrain type. On average 75-82 percent of the forest biomes are in lower terrains (plains, lowlands & low/mid plateaus with 18-25 percent in upper terrains (mountains and high plateaus). The forest areas of South America exhibit the most area in lower terrains for both forest types whereas the drier forest areas of Mesoamerica have a much larger percentage classified as upper terrains.

How well do the WWF Ecoregion boundaries correspond with the boundaries of the Focus Area?

The third level in the hierarchy of the WWF Terrestrial Ecoregions database is the ecoregions. The database contains over 800 ecoregions that form a contiguous coverage of the world. Each ecoregion is identified by name and a number of conservation and biodiversity characteristics including:

Biological Distinctiveness Index (BDI) which is a biogeographic scale-dependent assessment of the biological importance of ecoregions

• Current conservation status

• Future status – based on modifying the current status based on future threat

• Priority status for conservation action based on an integration of biological distinctiveness with future status

• Global status – a 30 year prediction of future conservation status

• Global 200 region – the Global 200 ecoregion of which this ecoregion is a part

This database provides general data on the biological importance of these ecoregions with a particular focus on conservation. WWF-US did extensive research based on the valuation, count and location of species in order to define the boundaries of the ecoregions so that they roughly coincide with the area over which key ecological processes take place (WWF-US 2001a).

With this wealth of data related to biodiversity and conservation as well as the possibility of working with WWF-US in the future to better understand the biodiversity criteria that went into defining these ecoregions a stratification approach based on ecoregions may be more appropriate for the BNPP/ASB study and is certainly something worth exploring in Phase II. Figures 19-22 illustrate how the ecoregions correspond to the climate defined Focus Area.

Section 3

Identifying Critical Areas

What urban populations might be at risk from significantly altered hydrological function as a consequence of upstream land use change?

As mentioned before the use of watersheds or basins as a stratifier is important since it helps us understand the externalities and broader-reaching impacts of policy and land management decisions. By combining available data on urban populations (World Bank 2001) with the Hydro1k watershed data (USGS EDC 2001) we looked at the urban populations that might be at risk from changes in hydrological function as a consequence of upstream land use change. The urban population database contains location and estimated population for approximately 3,200 urban settlements around the world. We used these data to determine 1) the urban population by watershed and 2) how many large cities are located within each watershed (see Tables 8 & 9).

The majority of the tropical urban populations reside in Mainland Asia where 100 percent of the tropical watersheds contain at least one urban center (population greater than 100,000). Meso and South America have the fewest urban centers with only 24 percent of the total urban population in the tropics residing in these regions combined.

In order to define ‘Critical watersheds’ we looked at the watersheds that had a very high urban population to total area ratio. These were defined as ‘critical’ following the assumption that watersheds that were more densely populated with urban areas had greater potential for negative impacts on the flood regime. Figures 23-25 show these plots and highlight the ‘critical’ watersheds on a regional basis.

In Latin America the critical watersheds are in the northern coasts of Colombia and Venezuala and the northeast of Venezuala at the mouth of the Orinoco. In Africa, these critical areas are more spread out covering areas in Nigeria, Cameroon, Gabon, Democratic Republic of Congo and Ghana. Not surprisingly, the critical areas in Asia are in the densely populated coastal areas of southern and eastern India and Bangladesh.

Where, within the Focus Area, are the most important areas of biological diversity?

WWF has identified over 200 ecoregions that are ‘unique expressions of biological diversity, …, ecological processes, and evolutionary phenomena (WWF-US 2001b).’ These Global 200 ecoregions are a collection of the Earth’s most outstanding and diverse areas where loss of biological wealth will be most severely felt, and where conservation efforts must be the strongest (WWF-US 2001b). The WWF Terrestrial Ecoregions database includes information on the Global 200 ecoregions related to their conservation status as a Global 200 region as well as the conservation information included for all of the ecoregions (see question How well do the WWF Ecoregion boundaries correspond with the boundaries of the Focus Area? for a complete list).

We intersected the Focus Area with the ecoregions in the Global 200 to determine the share of the BNPP/ASB mandate area that is considered of biological importance according to WWF (see Figure 26 and Table 10).

Table 10 Global 200 ecoregions within the Focus Area

The map and table include only those Global 200 ecoregions that had at least part of their area within the Focus Area[2]. 45 percent of the Focus Area falls within a Global 200 ecoregion highlighting that much of the Focus Area is considered of biological importance. The area share of the Global 200 ecoregions that coincides with the Focus Area is 74 percent indicating that a large percentage of the tropical areas that are considered to be strong areas for conservation fall within the BNPP/ASB mandate area.

What practical measures of biodiversity ‘value’ are available for these critical areas?

We discovered when doing the overlay of the Global 200 ecoregions with the Focus Area that all of the ASB clusters of sites were located within a Global 200 ecoregion[3]. But, at this scale, can we learn anything about the biodiversity of the ASB sites from looking at characteristics of the Global 200 ecoregions? Table 11 shows the Global 200 characteristics for each of the clusters of ASB sites. There is little value gained from this except for the implication that the ASB sites are in areas of high conservation value and biological significance but there is potential, through working with the researchers at WWF-US, to gain a better understanding of what the biodiversity issues are for each of these sites and to learn more about the spatial patterns of biodiversity within these specific ecoregions, how they correspond to these clusters of sites and thus what conservation efforts are most important at the site level.

Can we identify forest margin areas using existing coarse resolution satellite data?

From a conservation perspective many of the areas that are most critical are those that are on the margins between agriculture and forest. These areas are more vulnerable to further degradation due to poorly managed agricultural practices, clear cutting for agricultural expansion, loss of biodiversity due to fragmentation of habitat and damage to the hydrological function. It is often difficult to gain a systematic understanding of where the forest margin areas are, if they form a pattern across space and if they are a logical place to target for conservation or research efforts.

The Global Land Cover Characteristics Database was used for this study to determine the land cover within the Focus Area (see question What is the distribution of land cover within the Focus Area?). Using these same data, we identified all areas that fell on the margin of forest and agriculture based on a buffer zone of 1 pixel or 1 square kilometer. Figure 27 shows an example of the resulting ‘margin areas’ map. The brown and dark orange areas are those that fall in the areas adjacent to agricultural activity.

This type of analysis can help identify target areas for conservation or improved land management strategies. It can also aid in the identification of patterns of conversion across both narrow and broad areas.

Section 4

Findings and Recommendations

On the basis of this initial interpretation of the data compiled and harmonized for the Phase I Pantropic Assessment, some useful findings and recommendations have emerged.

• It is useful to select characterization variables that are hierarchical (e.g., we used a nested basin/watershed typology, and a nested biome/ecoregion typology) to better support truly multi-scale assessments.

• It appears problematic to define hydrological functions at the pantropic scale.

• If we can define landscapes in terms of physiography (e.g. roughness factor) and land cover then the measures developed here might also be useful in a landscape context

• Biomes appear to hold more promise for defining the BNPP/ASB Focus Area, rather solely climate-based, ecologically-defined areas.

• To test the validity and potential extrapolation power of the type of characterizations done here, we need to test them against

• sub-regional data and analyses (e.g., in Mesoamerica and Montane Mainland SEA).

Section 5

References

Center for International Earth Science Information Network. (CIESIN). 2000. Gridded Population of the World (GPW) v2.0. Data available online at:

DeFries, R.S., M.C. Hansen, J.R.G. Townshend, A.C. Janetos, and T.R. Loveland. 2000. “A New Global 1-Km Data Set of Percentage Tree Cover Derived from Remote Sensing.” Global Change Biology 6: 247-254.

Environmental Systems Research Institute, Inc. (ESRI). 1996. World Countries 1995. Included in ESRI Data and Maps. Volume 1. CD-ROM. Redlands, CA: ESRI. Country names and disputed territories updated at WRI and IFPRI as of 1999.

FAO. 1978-81. Food and Agriculture Organization of the United Nations, Report of the Agro-Ecological Zones Project. World Soil Resources Report no. 48 vol. 1-4, Rome: FAO.

FAO/IIASA. 2000. Food and Agriculture Organization of the United Nations and International Institute for Applied Systems Analysis, Global Agro-ecological Zoning. FAO Land and Water Digital Media Series # 11. CD-ROM version 1.0.

Global Land Cover Characteristics Database (GLCCD), Version 2.0. 2000. Data available online at: .

IGBP. 1998. International Geosphere Biosphere Programme (IGBP) Data and Information System, IGBP-DIS Global 1-km Land Cover Set DISCover.

Loveland, T.R., B.C. Reed, J.F. Brown, D.O. Ohlen, Z. Zhu, L. Yang, and J.W. Merchant. 2000. “Development of a Global Land Cover Characteristics Database and IGBP DISCover from 1 km AVHRR Data.” International Journal of Remote Sensing 21 (6/7): 1303-1330).

Meybeck, M, P. Green, C. Vorosmarty. 2001. “A New Typology for Mountains and Other Relief Classes: An Application to Global Continental Water Resources and Population Distribution”. In Mountain Research and Development. 21(1): 34-45.

Nelson, A. 2001. Global Terrain Surface. Derived from GTOPO30 Elevation Data (USGS 1998) using methodology from Meybeck et al. 2001. Unpublished Data.

Oak Ridge National Laboratory. (ORNL). 2000. LandScan 2000: Global Population Project. . Data available online at:

Pfafstetter, O., 1989, "Classification of hydrographic basins: coding methodology", unpublished manuscript, DNOS, August 18, 1989, Rio de Janeiro; translated by J.P. Verdin, U.S. Bureau of Reclamation, Brasilia, Brazil, September 5, 1991.

USGS EROS Data Center (EDC). 2001. Hydro1k Elevation Derivative Database. Data available online at: .

USGS. 1998. United States Geological Surveys Earth Resources Observation Systems (EROS) Data Center, GTOPO30: Global 30 Arc Second Elevation Data Set. Sioux Falls, SD: USGS EDC.

Verdin, K. 2001. A System for Typologically Coding Global Drainage Basins and Stream Networks. Online documentation: . Downloaded 8/29/01.

Wood, S., K. Sebastian, S.J. Scherr. 2001. Pilot Analysis of Global Ecosystems: Agroecosystems. Washington, D.C.: International Food Policy Research Institute and World Resources Institute.

World Bank. 2001. Urban Settlements. Unpublished Data.

World Wildlife Fund (WWF-US). 2001a. Terrestrial Ecoregions Database. WWF-US. Washington, DC. Unpublished Data and readme file).

World Wildlife Fund (WWF-US) 2001b. Global 200 information page located at: . December 9, 2001.

-----------------------

[1] The Hydro1k Database available on the web did not include the names of the basins thus the table only identifies the basin number. These numbers correspond to the numbers present on the map in Figure 11. We have submitted a request to USGS EDC for the names.

[2] 96 percent of the Global 200 terrestrial ecoregions are included in this ‘tropical’ subset

[3] Note: all of the ASB clusters of sites are located in the Tropical and subtropical moist broadleaf forest biome.

-----------------------

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download