• eddy covariance;
  • evapotranspiration;
  • rainfall interception;
  • soil water deficit;
  • sugarcane;
  • yield


The evapotranspiration (E) from a sugarcane plantation in the southeast Brazil was measured by the eddy-covariance method during two consecutive cycles. These represented the second (393 days) and third year (374 days) re-growth (ratoon). The total E in the first cycle was 829 mm, accounting for 69% of rainfall, whereas in the second cycle, it was 690 mm, despite the total rainfall (1353 mm) being 13% greater. The ratio of E to available energy, the evaporative fraction, exhibited a smaller variation between the first and second cycles: 0.58 and 0.51, respectively. The estimated interception losses were 88 and 90 mm, respectively, accounting for approximately 7% of the total rainfall. The sugarcane yield in the second cycle (61.5 ± 4.0 t ha−1) was 26% lower than the first cycle, as well as lower than the regional average for the third ratoon (76 t ha−1). The below average yield was associated with less available soil water at the beginning of the cycle, with the amount of rainfall recorded during the first 120 days of re-growth in the second cycle being 16% of that recorded in the first (203 mm).