Land-use change affects water recycling in Brazil's last agricultural frontier
Historically, conservation‐oriented research and policy in Brazil have focused on Amazon deforestation, but a majority of Brazil's deforestation and agricultural expansion has occurred in the neighboring Cerrado biome, a biodiversity hotspot comprised of dry forests, woodland savannas, and gras...
Uloženo v:
| Vydáno v: | Global change biology Ročník 22; číslo 10; s. 3405 - 3413 |
|---|---|
| Hlavní autoři: | , , , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
England
Blackwell Publishing Ltd
01.10.2016
|
| Témata: | |
| ISSN: | 1354-1013, 1365-2486, 1365-2486 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Shrnutí: | Historically, conservation‐oriented research and policy in Brazil have focused on Amazon deforestation, but a majority of Brazil's deforestation and agricultural expansion has occurred in the neighboring Cerrado biome, a biodiversity hotspot comprised of dry forests, woodland savannas, and grasslands. Resilience of rainfed agriculture in both biomes likely depends on water recycling in undisturbed Cerrado vegetation; yet little is known about how changes in land‐use and land‐cover affect regional climate feedbacks in the Cerrado. We used remote sensing techniques to map land‐use change across the Cerrado from 2003 to 2013. During this period, cropland agriculture more than doubled in area from 1.2 to 2.5 million ha, with 74% of new croplands sourced from previously intact Cerrado vegetation. We find that these changes have decreased the amount of water recycled to the atmosphere via evapotranspiration (ET) each year. In 2013 alone, cropland areas recycled 14 km3 less (−3%) water than if the land cover had been native Cerrado vegetation. ET from single‐cropping systems (e.g., soybeans) is less than from natural vegetation in all years, except in the months of January and February, the height of the growing season. In double‐cropping systems (e.g., soybeans followed by corn), ET is similar to or greater than natural vegetation throughout a majority of the wet season (December–May). As intensification and extensification of agricultural production continue in the region, the impacts on the water cycle and opportunities for mitigation warrant consideration. For example, if an environmental goal is to minimize impacts on the water cycle, double cropping (intensification) might be emphasized over extensification to maintain a landscape that behaves more akin to the natural system. |
|---|---|
| Bibliografie: | istex:99648AFEA6694393B2ECFB7500E701222A21494D ArticleID:GCB13298 Figure S1. Map of cropland agriculture in the Cerrado states of Mato Grosso (MT), Goiás (GO), Maranhão (MA), Piauí (PI), Tocantins (TO), and Bahia (BA) in 2003 and 2013. Figure S2. Producer accuracy (left) and user accuracy (right). Figure S3. Comparison of government reported cropland area (soy - green, corn - orange, cotton - purple, total - black) across Matopiba (dotted lines, referred in legend to as GOV) to mapped agricultural area using the modified Spera et al. () algorithm (solid lines, referred in legend to as ARC). Figure S4. Comparison of government reported cropland area (soy - green, corn - orange, cotton - purple, total - black) across Goias (dotted lines, referred in legend to as 'GOV') to mapped agricultural area using the modified Spera et al. () algorithm (solid lines, referred in legend to as ARC'). Figure S5. Comparison of government reported cropland area (soy - green, corn - orange, cotton - purple, total - black) across Mato Grosso (dotted lines, referred in legend to as 'GOV') to mapped agricultural area using the modified Spera et al. () algorithm (solid lines, referred in legend to as 'ARC'). Figure S6. Cropland area in the Matopiba region in 2013 and baseline (6) land cover classes. Figure S7. Average annual rainfall and average annual growing season duration as calculated from TRMM data over 1999-2014. Figure S8. Monthly difference between ETNOLUC, a derived, interpolated surface in which no land use change has occurred after August 2002, and ETLUC, the observed monthly composite ET data. Figure S9. Area of large-scale mechanized agriculture across the Cerrrado. Figure S10. Semivariograms for each variable exhibiting spatial autocorrelation. Figure S11. Climate data aggregated to annual averages across Maptioba. Figure S12. Trendlines from monthly MODIS ET data. Figure S13. Streamflow trends over time. Table S1. Confusion Matrix. Table S2. Confusion Matrix with Crop Types Separated. Table S3. Regression result for each month. Table S4. Regression result for each month. Table S5. Regression result for each month. Table S6. Regression result for each month. Table S7. Regression result for each month. Table S8. Regression result for each month. Table S9. Regression result for each month. Table S10. Regression result for each month. Table S11. Regression result for each month. Table S12. Regression result for each month. Table S13. Regression result for each month. Table S14. Regression result for each month. Gordon and Betty Moore Foundation NASA Terrestrial Ecology - No. NNX12AK11G ark:/67375/WNG-2RQ75GS1-W NASA Land-Cover/Land-Use Change Program - No. NNX11AH91G; No. NNX11AE56G ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23 |
| ISSN: | 1354-1013 1365-2486 1365-2486 |
| DOI: | 10.1111/gcb.13298 |