Regional Environmental Change

, Volume 16, Supplement 1, pp 69–82

Quantifying fluctuations in winter productive cropped area in the Central Indian Highlands

  • Pinki Mondal
  • Meha Jain
  • Mateusz Zukowski
  • Gillian Galford
  • Ruth DeFries
Original Paper

DOI: 10.1007/s10113-016-0946-y

Cite this article as:
Mondal, P., Jain, M., Zukowski, M. et al. Reg Environ Change (2016) 16(Suppl 1): 69. doi:10.1007/s10113-016-0946-y


The Central Indian Highland landscape (CIHL) represents a complex, diverse, and highly human-modified system. Nearly half the landscape is cropland, yet it hosts 21 protected areas surrounded and connected by forests. Changing farming practices with increasing access to irrigation might alter this intensifying landscape in the near future particularly in light of weather variability. We analyzed a decade of remote sensing data for cropping patterns and climatic factors combined with census data for irrigation and demographic factors to understand winter cropping trajectories in the CIHL. We quantified ‘productive cropped area’ (PCA), defined as the area with planted crop that is green at the peak of the winter growing season. We find three primary trajectories in PCA—increasing, fluctuating, and decreasing. The most dominant trend is fluctuating PCA in two-thirds of the districts, ranging from ~2.11 million to ~3.73 million ha between 2001 and 2013, which is associated with village-level access to irrigation and local labor dynamics. In 58 % of all districts, clay soils were associated with winter cropping (p < 0.05). Increasing irrigation is associated with increased winter PCA in most (94 %) districts (p < 0.00001). We find strong negative association between PCA and land surface temperature (LST) in most (66 %) districts (p < 0.01). LST closely corresponds to daytime mean air temperature (p < 0.001) for available meteorological stations. Fine-scale meteorological and socioeconomic data, however, are needed to further disentangle impacts of these factors on PCA in this landscape.


Agriculture Crop Climate Small-holder farmers Central India 

Supplementary material

10113_2016_946_MOESM1_ESM.pptx (155 kb)
Fig. S1Changes in winter irrigation during 2001-13 in the five district categories with – (a) low increasing, (b) high increasing, (c) low fluctuating, (d) high fluctuating, and (e) decreasing cropped area. Irrigation data for 2012 was not available for any district, and was interpolated in these graphs. Irrigation data for 2001-2013 was not available for the districts of Wardha, Nagpur, Bhandara and Gondiya, hence these districts were not included in these graphs (PPTX 154 kb)
10113_2016_946_MOESM2_ESM.tif (114 kb)
Fig. S2Enhanced Vegetation Index (EVI) time-series data for representative shrub pixel and crop pixel during 2000-2010. Shrub phenology indicates a green-up during late monsoon, whereas crop pixel shows EVI peaks during both monsoon and winter, thus representing a double-crop pixel. Our scaling method focuses only on winter season (defined by solid black lines) for each year, hence shrubland is unlikely to influence winter crop signal (TIFF 113 kb)
10113_2016_946_MOESM3_ESM.pdf (103 kb)
Fig. S3Graphs showing: (a) productive cropped area (ha) vs. agricultural census data on winter cropped area (ha) for all districts in Madhya Pradesh for 2001–2013; (b) productive cropped area (ha) vs. agricultural census data on winter + all cropped area (ha) for all districts in Madhya Pradesh for 2001–2013; (c) mean percent productive cropped area at the pixel level for all districts compared to the yield of wheat (million ton/ha) as defined in the census data for all years available during 2001–2013; (d) mean percent productive cropped area at the pixel level for all districts compared to the yield of winter pulse (million ton/ha) as defined in the census data for all years available during 2001–2013. Inset values denote adjusted R2 values for each of these graphs (PDF 103 kb)
10113_2016_946_MOESM4_ESM.tif (66.9 mb)
Fig. S4Maps showing changes in productive cropped area for each year between 2001 and 2013 (TIFF 68487 kb)
10113_2016_946_MOESM5_ESM.pptx (94 kb)
Fig. S5Temporal mean of the productive cropped area percentages for the districts with (a) increasing cropped area, and (b) fluctuating cropped area. The maximum break classification method was used to differentiate between the ‘high’ and ‘low’ groups (divided by solid vertical line) in each of these two categories (PPTX 94 kb)
10113_2016_946_MOESM6_ESM.tif (1010 kb)
Fig. S6Spatial distribution of soil types in the study region. This map has been reclassified from the Food and Agriculture Organization of the United Nations (FAO) digital soil map of the world (FAO 2003) based on the soil types most abundant in each of the study districts. Detailed composition of each of the soil types has been listed in Table 1 (TIFF 1010 kb)
10113_2016_946_MOESM7_ESM.pdf (155 kb)
Fig. S7Association between MODIS land surface temperature (LST) and winter daytime mean air temperature at five meteorological stations located within the study landscape. Air temperature data are from station locations, and LST values are from corresponding pixels within which these stations are located (PDF 155 kb)
10113_2016_946_MOESM8_ESM.docx (18 kb)
Supplementary material 8 (DOCX 18 kb)

Funding information

Funder NameGrant NumberFunding Note
National Aeronautics and Space Administration
  • 522363

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  • Pinki Mondal
    • 1
  • Meha Jain
    • 2
  • Mateusz Zukowski
    • 3
  • Gillian Galford
    • 4
  • Ruth DeFries
    • 1
  1. 1.Department of Ecology, Evolution and Environmental BiologyColumbia UniversityNew YorkUSA
  2. 2.Department of Environmental Earth System ScienceStanford UniversityStanfordUSA
  3. 3.Department of Applied Physics and Applied MathematicsColumbia UniversityNew YorkUSA
  4. 4.Gund Institute for Ecological Economics, Rubenstein School of Environment and Natural ResourcesUniversity of VermontBurlingtonUSA

Personalised recommendations