Effect of climate change-induced water-deficit stress on long-term rice yield

correlated with each other in the same groups (Fig 3). The correlation coefficients of the yearly yields among the rice cultivar pairs with overlapping cultivation periods averaged 0.838, 0.605, and 0.665 in the cool cropping season during 1925–1944, 1945–1983, 1996–2019, respectively, and 0.716, 0.566, and 0.735 in the warm cropping season during 1925–1944, 1945–1983, 1996– 2019, respectively. The correlations between the changes in crop water status under the four growth stages may make it difficult to separate the effects of different growth stages due to the co-linearity [48]. Although these problems have been discussed, the observations at our station showed low to little correlation among the values at different growth stages during the crop- ping seasons. The cultivars of japonica type rice in Taiwan are the only japonica type rice that can grow under relatively high temperatures and produce good-quality rice with a high yield [49, 50]. Crops in tropical regions have been reported to be more sensitive to warming because their temperature is already close to their optimum temperature during the growing period [3]. In many regions, a slight increase in temperature with sufficient rainfall may have a positive effect on crops [51]. The lowland rice varieties were reported to be highly sensitive to soil drying, and their yields decline when the soil dries below saturation [52]. Data were collected from the same research station during the long-term experimental period. To obtain general results, the analysis of the crop yield response to global or national

water-deficit stress should be extended. Data collected in different areas or on different tempo- ral and spatial scales may result in different conclusions [53, 54]. For example, up to 45% yield reductions of rice are expected by the end of this century due to climate change, including water deficit, in the countries in eastern Africa [55]. In Iran, it was reported that water deficit during vegetative, flowering and grain filling stages reduced mean grain yield by 21, 50 and 21% on average in comparison to control, respectively [56]. In this study, long-term temporal variation in the rice yield response to water-deficit stress was revealed, even though the rice cultivars varied throughout the study period. During an experimental period of over 90 years since 1925, it is impossible to maintain the crop yield experiments using the same cultivar and maintaining the same environmental and cultivational conditions consistently. It is also diffi- cult to consider the factors that may affect the growth and production of crops, such as insects, diseases, and soil fertility [57–60], as well as human-induced effects, such as modern manage- ment, improving technology, and cultivator practices [26, 58] for long-term observations. Crop evapotranspiration could be influenced by other factors, such as soil condition, canopy cover, and the fraction of leaf senescence; thus, the information of these coefficients may be considered for the calculation of crop evapotranspiration, if possible [19, 22]. In addition,

extreme climatic events, such as floods and heatwaves, may pose additional risks to crop pro- duction [61]. This study revealed the effect of water-deficit stress on rice yield in both cool and warm cropping seasons. The results provide long-term evidence of declining crop water status dur- ing the rice-growing seasons. The average values of ET0 were estimated as 3.3–4.4 mm day-1, and 2.8–4.6 mm day-1 in cool and warm cropping seasons, respectively, under the rice growth stages. The crop water status has decreased by 24.7–198.6 mm in the cool cropping season and 3.8–63.7 mm in the warm cropping season under the rice growth stages since 1925 and during the 95 years. Compared with the cool cropping season, the decreasing trend in crop water sta- tus in the warm cropping season was relatively slight under the four growth stages. The total water-deficit stress related yield change in the cultivars in the cool cropping season during 1925–1944, 1945–1983, and 1996–2019 were -56.1 to 37.0, -77.5 to -12.3, 11.2 to 19.8, and -146.4 to 39.1 kg ha-1 under the initial, crop development, reproductive, and maturity stages, respectively. The total yield change related to the CWS on the cultivars in the warm cropping season during 1925–1944, 1945–1983, and 1996–2019 are -16.5 to 8.2, -12.9 to 8.1, -2.3 to 9.0, PLOS ONE Effect of climate change-induced water-deficit stress on long-term rice yield PLOS ONE | https://doi.org/10.1371/journal.pone.0284290 April 17, 2023 10 / 14