How agrometeorological and water deficit variations influence the growth and yield of sugarcane

Abstract

Edaphoclimatic variables play a crucial role in shaping both the growth and yield of sugarcane. This study aimed to evaluate the intricate relationships among plant variables, agrometeorological factors, and water stress conditions in three successive crop cycles of sugarcane (cane plant, ratoon 1, and ratoon 2). The investigated plant variables included stem height, stalk diameter, number of tillers, leaf area, leaf area index, and stalk fresh yield, alongside growth rates (growth/time). Concurrently, climate variables, such as air temperature, humidity, rainfall, wind speed, solar and net radiation, reference evapotranspiration, and degree-days, were monitored. Additionally, water stress parameters, including water deficit and water surplus, were quantified. Statistical models were skillfully fitted (r2 > 0.90) to the biometric data, employing thermal time as a critical determinant. Surprisingly, no adverse agrometeorological or soil moisture conditions, as indicated in the literature, were detected when growth rates started to decline. This suggests that other unmeasured stress factors may have influenced the plants during that period. The analysis of the ratio between actual and maximum crop evapotranspiration (ETa/ETc) revealed the most pronounced sensitivity to water deficit during the vegetative growth phase (phase III). Moreover, the study identified that growth achieved satisfactory levels when at least 60% of the maximum crop evapotranspiration was met during the initial phases. A comprehensive cluster analysis encompassing height, leaf area, leaf area index, relative air humidity, soil moisture, and actual evapotranspiration rates provided valuable insights into the interrelated dynamics of these variables. Furthermore, a significant exponential reduction in yield was observed as the number of harvests increased. This decline in yield was attributed to the combined effects of 50% of biometric variables, 63% of agrometeorological variables, and 50% of water stress variables, all of which exhibited negative correlations with yield. Approximately 82%, 63%, and 71% of the correlations among biometric, agrometeorological, and water stress variables, respectively, were strong or very strong (r ≥ 0.70). As a result, this study highlights that: i) estimating sugarcane growth and tracking its developmental stages can be accomplished by employing appropriate models based on thermal time; ii) most biometric measurements exhibit correlations of r < 0.70 with agrometeorological variables; iii) an integrated understanding of biometric, agrometeorological, and water stress variables can effectively explain the observed reductions in sugarcane yield