Sunteți pe pagina 1din 21

19

SUPPLEMENTAL ON-LINE MATERIALS


For Land-Use Practices Have Negative, Global Scale Effects on Ecosystem Services and
Human Welfare by Foley et al.

Box S1. Land Use Practices and the Loss of Biodiversity. The current rate of species
extinction on Earth is orders of magnitude faster than expected background (i.e., without
human influence) rates (S4). It is estimated that ~800 plant and animal species are now
extinct in the wild (S5), and approximately half of the global flora is now threatened (S6).
The local extirpation of species, which affect local ecosystem dynamics and services, may
be twice as prevalent as global extinctions (S6-S8).

Land use practices have caused large losses of biodiversity, defined here as the number
and relative abundance of species that occur naturally in that biome (S9). Land use
changes ecosystems via habitat loss, modification and fragmentation, soil and water
degradation, reduction of water supply, exploitation of native species and introduction of
non-native species (S10-11). These changes often favor generalists and fast-growing
species, at the expense of native and rare organisms. Freshwater species are also
particularly vulnerable to land-use impacts (S7). Tropical moist forest conversion to
agriculture has caused the most extensive and direct reductions of biodiversity, by
deliberately removing the species-rich native forests and replacing them with simpler, less
diverse ecosystems.

Future land use will likely cause increasing rates of global species extinctions and local
biodiversity loss during the next century, mostly because agricultural land use is expected
to become both more extensive and more intensive (S12-13). One possible outcome is that
tropical forest biodiversity hotspots may lose 18-40% of their eukaryotic species over the
next few centuries, depending on how well the hotspots are protected between now and
2100 (S14). Using scenarios of future land use and climate, Sala et al. (S9) found that land

20

use will cause the biggest share of biodiversity loss by 2100, especially in tropical,
Mediterranean and grassland regions.

Box S2. What are Ecosystem Goods and Services?

Ecosystem goods and services

refer to the essential products (e.g., food, fiber, freshwater), vital environmental processes
(e.g., pollination, flood control, water purification, climate regulation), cultural and
aesthetic benefits (e.g., recreation and tourism, heritage, serenity and inspiration), and
preservation of options (e.g., genetic and species diversity for future use) provided by the
biosphere. The term has received considerable attention in recent years (S1-S3) and is the
focus of the recent international Millennium Ecosystem Assessment (S2).

Figure S1A-G.

Global Patterns of Agriculture.

Today, nearly 16-18 million km2

(roughly the size of South America) are in some form of cultivation, while another 33-35
million km2 (roughly the size of Africa) are used for pastures and rangeland (S15-16). Here
we illustrate the global patterns of croplands (adapted from S15) across each region of the
planet.

Figure S2A.

Diminishing Returns in U.S. Agriculture? The average annual maize

productivity in the United States has risen from 1.6 T ha-1 in the 1930s to 8.6 T ha-1 in
2001, primarily as a result of large increases in nitrogen fertilizer use. However, the
average rate of annual productivity gains for the Corn Belt region has declined from 3.4%
yr-1 (124.5 kg ha-1 yr-1) in the 1960s to 0.78% yr-1 (49.2 kg ha-1 yr-1) in the 1990s (ref. S17).
It has been suggested that as corn yields approach their upper limit defined by yield
potential, it becomes increasingly difficult for farmers to overcome the complex
interactions of soil nutrients, weather, and disease that inhibit productivity (S18-19).

Figure S2B. Effects of Changing Fertilizer Application on Maize Yield and Nitrate
Leaching in the Upper Mississippi Basin. A process-based agricultural ecosystem model

21

(S20) has estimated changes in maize yield and nitrate leaching for farms in the Upper
Mississippi drainage basin, in response to variations in fertilizer application rates (S21).
These results show how increasing fertilizer applications can result in low to moderate
increases in crop yield but potentially generate moderate to large increases in nitrate
leaching. Balancing the benefits (increased crop yield) and environmental costs (increased
nitrate leaching, among others) of farming practices may depend on recognizing such nonlinear relationships.

Figure S3A. Global Water Withdrawals for Agricultural, Domestic and Industrial
Demands. Geographic patterns of global water withdrawals estimated for the year 2000
(S22) show high levels of water use in large urban areas (including cities in the Eastern
U.S., Western Europe and China) and from irrigation across many major food-producing
regions, such as the Indogangenic Plain, eastern China, central Europe, the Nile River, and
throughout the western United States. (Adapted from ref. S22)

Figure S3B. Global Water Withdrawals Compared to Long-Term Average Renewable


Water Supply. Global water withdrawals for the year 2000 are compared to estimates of
the long-term average renewable water supply (S22), as estimated by a water balance model
and global climate data from 1901 to 1995. The ratio of total water withdrawals to longterm average renewable water supply (an indicator of relative water stress) is presented for
each major river basin of the globe. Values that approach one suggest that basins could
experience persistent water shortages, as a result of annual withdrawal amounts exceeding
the average rate of renewable water supply. (Adapted from ref. S22)

Figure S2C. Global Water Withdrawals Compared to Renewable Water Supply of


~10% Driest Years. Here global water withdrawals (for 2000) are compared to the
renewable water supply estimated for the driest 10 years between 1901 and 1995 (S22). In
this comparison, several regions exhibit increased vulnerability to water shortages,
including:

the southwest U.S., portions of Western Europe, northern Africa, the Rio

Colorado Basin in Argentina, northeast Brazil, southern Africa, and much of Australia.

22

Even though this comparison represents an extreme case, it does illustrate the susceptibility
of many of the regions over the globe to possible water scarcity from drought. (Adapted
from ref. S22)

23

Table Legends

Table S1. Global Cropland Areas, By Continent. Adapted from ref. S15.

Table S2A. Estimated Changes in Temperature and Precipitation That Could Result
from Large-Scale Land Cover Change. A set of global climate model simulations (S23)
have demonstrated how large-scale land cover change removing an entire biome from the
surface of the planet, one at a time could affect the global climate system. Each biomes
potential influence on the annual average temperature (C) and precipitation (mm day-1) are
presented as differences (vegetation removal control simulations).

The results are

summarized over the areas where vegetation was removed (devegetated), over all land
areas (all land) and the entire globe (global).

Only gridcells with a statistically

significant change in temperature or precipitation (using a two-sided Students t-Test, at


95% confidence) are used in this analysis. (Adapted from ref. S23)

Table S2B. Summary of How Large-Scale Land Cover Change Could Affect Climate.
Global climate model simulations (S23) are analyzed to determine the primary mechanisms
by which large-scale land cover change could affect climate. Each simulation, wherein an
entire biome is removed from the planet, is ranked by how changed in albedo or
evapotranspiration (ET) affect the climate. The qualitative ranking is from strong to
moderate to weak and is based on the results from the whole suite of land cover change
simulations. A brief description highlights the important biophysical mechanisms and
climatic effects associated with each land-cover change simulation. (Adapted from ref.
S23)

Supplemental References

24

S1. Daily, G.C. Nature's Services: Societal Dependence on Natural Ecosystems (Island Press,
Washington, D.C., 1997).
S2. Millennium Ecosystem Assessment. Ecosystems and Human Well- being: A Framework for
Assessment (Island Press, Washington, D.C., 2003).
S3. Palmer et al., Science 304, 1251-1252 (2004).
S4. Hanski, I., J. Clobert and W. Reid. in Global Biodiversity Assessment, Section 4, R. Barbault
and S. Sastrapradja, Eds., Cambridge Univ. Press, Cambridge (1995).
S5. IUCN Red List. http://www.redlist.org/ (2003).
S6. Pitman, N.C.A. and P.M. Jrgensen, Science 298, 989 (2002).
S7. Brook, B.W., N.S. Sodhi and P.K.L. Ng, Nature 424 420-423 (2003).
S8. Myers, N., et al., Nature 403, 853-858 (2000).
S9. Sala, O.E. et al., Science 287,1770-1774 (2000).
S10. Sala, O.E. in Global Biodiversity Assessment, Section 5, H.A. Mooney, J. Lubchenco, R.
Dirzo and O.E. Sala, Eds., Cambridge Univ. Press, Cambridge (1995)
S11. IPCC, Climate Change 1995: Impacts, Adaptations and Mitigation of Climate Change:
Scientific-Technical Analysis, edited by R.T. Watson, M.C. Zinyowera, R.H. Moss, and D.J.
Dokken, pp. 878, Cambridge University Press, Cambridge (1995).
S12. Tilman, D., et al., Science 292, 281-284 (2001).
S13. IPCC. IPCC Technical Paper V, H. Gitay, A. Surez, R.T. Watson and D.J. Dokken, Eds.,
Cambridge Univ. Press, Cambridge (2002).
S14. Pimm, S.L. and P. Raven., Nature 403, 843-845 (2000).
S15. Ramankutty, N. & J.A. Foley, Global Biogeochem. Cy. 13, 997-1027 (1999).

25

S16. Asner, G.P., A.J. Elmore, L.P. Olander, R.E. Martin, & A.T. Harris, Ann. Rev. Environ.
Resources, 29 (2004).

S17. Kucharik, C.J. & Ramankutty, N. Earth Interactions, in press.


S18. Mann, C.C., Science 283, 310-314 (1999).
S19. Cassman, K.G., Dobermann, A., Walters, D.T. & Yang, H., Annu. Rev. Environ.
Resour. 28, 315-358 (2003).
S20. Donner, S.D., Kucharik, C.J. & Foley, J.A. Global Biogeochem. Cy. 18, GB1028,
doi:10.1029/2003GB002093 (2004).
S21. Donner, S.D. and C.J. Kucharik. Global Biogeochem. Cy 17:
doi:10.1029/2001GB001808 (2003).
S22. Helkowski, J. M.S. Thesis. (Environmental Monitoring Program, University of Wisconsin
Madison, Madison, WI, 2004).
S23. Snyder, P.K, C. Delire, J.A. Foley, Clim. Dynam. 23, 279-302 (2004).

Region

cropland area

total land area

cropland percentage

North America

2.90 x 106 km2

21.28 x 106 km2

13.6%

South America

2.01 x 106 km2

17.62 x 106 km2

11.4%

Africa

2.35 x 106 km2

29.71 x 106 km2

7.9%

Eurasia

10.00 x 106 km2

50.75 x 106 km2

19.7%

Australia-Pacic

0.74 x 106 km2

10.72 x 106 km2

6.9%

Global

18.00 x 106 km2

130.09 x 106 km2 *

13.8% *

(* not including ice covered regions)

(* not including ice covered regions)

Table S1. Global Cropland Areas, By Continent. Adapted from ref. S4.

Biome Removed
tropical forest
boreal forest
temperate forest
savanna
grassland / steppe
shrubland / tundra

T devegetated region
1.18C
-2.75C
-1.07C
0.87C
0.75C
0.32C

Tall land
0.71C
-1.58C
-0.46C
0.37C
0.20C
0.04C

T global
0.24C
-0.77C
-0.22C
0.12C
0.05C
-0.01C

P devegetated region
-1.34 mm day-1
-0.27 mm day-1
-0.49 mm day-1
-1.10 mm day-1
-0.41 mm day-1
-0.35 mm day-1

Pall land
-0.55 mm day-1
-0.11 mm day-1
-0.18 mm day-1
-0.61 mm day-1
-0.31 mm day-1
-0.31 mm day-1

P global
-0.11 mm day-1
0.00 mm day-1
-0.06 mm day-1
-0.19 mm day-1
-0.12 mm day-1
-0.08 mm day-1

Table S2A. Estimated Changes in Temperature and Precipitation That Could Result from Large-Scale Land Cover Change. A set of global
climate model simulations (S23) have demonstrated how large-scale land cover change removing an entire biome from the surface of
the planet, one at a time could affect the global climate system. Each biomes potential inuence on the annual average temperature
(C) and precipitation (mm day1) are presented as differences (vegetation removal control simulations). The results are summarized
over the areas where vegetation was removed (devegetated), over all land areas (all land) and the entire globe (global). Only
gridcells with a statistically signicant change in temperature or precipitation (using a two-sided Students t-Test, at 95% condence) are
used in this analysis. (Adapted from ref. S23)

Biome Removed

albedo effects

ET effects

description

tropical forest

moderate

strong

Moderate surface albedo increase causes a decrease in surface net radiation. Severe reduction in ET along with reduction in
surface energy limits latent cooling and surface warms considerably. Near-surface specic humidity and precipitation
reduced. Cloud cover decreases.

boreal forest

strong

weak

temperate forest

moderate

moderate

Increase in albedo decreases net radiation during all seasons. Winter and spring increase causes a cooling, while summer and
fall increase causes a warming when combined with moderate reduction in ET. Decrease in ET reduces near-surface
moisture and precipitation (mostly in summer).

savanna

moderate

strong

Moderate surface albedo increase causes a decrease in surface net radiation. Large reduction in ET limits latent cooling and
surface warms. Near-surface humidity and precipitation severely reduced. Cloud cover decreases.

grassland / steppe

moderate

moderate

Moderate surface albedo increase causes a reduction in net radiation. In winter and spring, this cools the surface while in
summer it warms the surface when combined with a moderate reduction in ET. Grassland region of central US most
affected by ET reduction (warming). Steppe regions of Asia mostly affected by albedo increase (cooling).

shrubland / tundra

moderate

moderate /
weak

Moderate surface albedo increase causes a reduction in net radiation over all regions. Albedo increase in tundra regions
causes cooling; albedo increase in shrubland regions of Australia, combined with reduced ET, causes warming during austral
summer and cooling in austral winter. Weak reduction in ET over tundra regions.

Large surface albedo increase causes a large decrease in net radiation and the surface cools. Reduction in ET limits nearsurface moisture and precipitation only during summer growing season. Surface cooling leads to a reduction in planetary
boundary layer height and a large increase in cloud cover.

Table S2B. Summary of How Large-Scale Land Cover Change Could Affect Climate. Global climate model simulations (S23) are
analyzed to determine the primary mechanisms by which large-scale land cover change could affect climate. Each simulation, wherein
an entire biome is removed from the planet, is ranked by how changed in albedo or evapotranspiration (ET) affect the climate. The
qualitative ranking is from strong to moderate to weak and is based on the results from the whole suite of land cover change
simulations.A brief description highlights the important biophysical mechanisms and climatic effects associated with each land-cover
change simulation. (Adapted from ref. S23)

S-ar putea să vă placă și