@article{fdi:010079550,
  title = {{U}sing remote sensing to assess the effect of trees on millet yield in complex parklands of {C}entral {S}enegal},
  author = {{L}eroux, {L}. and {F}alconnier, {G}. {N}. and {D}iouf, {A}. {A}. and {N}dao, {B}. and {G}bodjo, {J}. {E}. and {T}all, {L}. and {B}alde, {A}. {A}. and {C}lermont {D}auphin, {C}athy and {B}egue, {A}. and {A}ffholder, {F}. and {R}oupsard, {O}.},
  editor = {},
  language = {{ENG}},
  abstract = {{A}groforestry is pointed out by the {I}ntergovernmental {P}anel on {C}limate {C}hange report as a key option to respond to climate change and land degradation while simultaneously improving global food security ({IPCC}, 2019). {F}aidherbia albida parklands are widespread in {S}ub-{S}aharan {A}frica and provide several ecosystem services to populations, notably an increase in crop productivity. {W}hile remote sensing has been proven useful for crop yield assessment in smallholder farming system, it has so far ignored the woody component. {W}e propose an original approach combining remote sensing, landscape ecology and statistical modelling to i) improve the accuracy of millet yield prediction in parklands and ii) identify the main drivers of millet yield spatial variation. {T}he parkland of {C}entral {S}enegal was chosen as a case study. {F}irstly, we calibrated a remote sensing-based linear model that accounted for vegetation productivity and tree density to predict millet yield. {I}ntegrating parkland structure improved the accuracy of yield estimation. {T}he best model based on a combination of {G}reen {D}ifference {V}egetation {I}ndex and number of trees in the field explained 70% of observed yield variability (relative {R}oot {M}ean {S}quared {E}rror ({RRMSE}) of 28%). {T}he best model based solely on vegetation productivity (no information on parkland structure) explained only 46% of the observed variability ({RRMSE} = 34%). {S}econdly we investigated the drivers of the spatial variability in estimated yield using {G}radient {B}oosting {M}achine algorithm ({GBM}) and biophysical and management factors derived from geospatial data. {T}he {GBM} model explained 81% of yield spatial variability. {P}redominant drivers were soil nutrient availability (i.e. soil total nitrogen and total phosphorous) and woody cover in the surrounding landscape of fields. {O}ur results show that millet yield increases with woody cover in the surrounding landscape of fields up to a woody cover of 35%. {T}hese findings have to be strengthened by testing the approach in more diversified and/or denser parklands. {O}ur study illustrates that recent advances in earth observations open up new avenues to improve the monitoring of parkland systems in smallholder context.},
  keywords = {{A}groforestry ; {L}andscape ; {C}rop yield ; {S}mallholder agriculture ; {R}emote sensing ; {F}aidherbia albida ; {A}frica ; {SENEGAL}},
  booktitle = {},
  journal = {{A}gricultural {S}ystems},
  volume = {184},
  numero = {},
  pages = {102918 [13 p.]},
  ISSN = {0308-521{X}},
  year = {2020},
  DOI = {10.1016/j.agsy.2020.102918},
  URL = {https://www.documentation.ird.fr/hor/fdi:010079550},
}