In theory, optimal plant water balance may be realized through harmony between root water absorption and leaf transpiration (Ibrahim et al., 2018). Adequate water supply is essential for healthy plant growth, but when the soil becomes saturated, potentially resulting in plant waterlogging stress (Singhal et al., 2022). Soil waterlogging can manifest in agricultural fields for various reasons, such as excessive rainfall or irrigation, inadequate soil drainage, rising or perched water tables and lateral surface or subsurface flows (Liu et al., 2021). This can deplete oxygen levels within soil pores, leading to diminished growth, senescence and, in severe cases, crop mortality (Githui et al., 2022). Waterlogging can also indirectly impact on growth via soil mineral nitrogen, organic matter and carbon, as well as soil microbiota (Phelan et al., 2015; Rawnsley et al., 2019; Sándor et al., 2020).

The middle and lower reaches of the Yangtze River region are among China’s primary wheat production zones, but they are also severely afflicted by waterlogging (Wan et al., 2022). Around 41% of total arable land area in this region grapples with waterlogging, significantly hampering the consistency and stability of crop yields (Hu, 2023). Investigating the impact of waterlogging, particularly hypoxia stress, on wheat growth and yield is crucial, together with how waterlogging may be amplified or interplay with other crop stresses, such as extreme heat, that may occur later in the growing season (Langworthy et al., 2018; Harrison, 2021).

Two approaches are typically employed to assess the extent of crop waterlogging. The first relies on soil groundwater depth, often computed in crop models using algorithms pertaining to cumulative excess groundwater depth (SEW30) (Kanwar et al., 1988; Qian et al., 2015) and consecutive suppression days of waterlogging (CSDI) (Evans and Skaggs, 1993). This method primarily elucidates the impact of groundwater and surface water on crop waterlogging stress, yet does not allow other waterlogging pathways to be captured, such as waterlogging triggered by excessive irrigation (Liu et al., 2020). Conversely, the second category is grounded upon occurrence of climatic elements, which then allows computation of indices such as the waterlogged day index (Guo et al., 2016) and wetland day index (Zhang et al., 2019). Such parameters ignore the multifaceted and integrated effects of soil type, topography and hydrology. Although remote sensing may potentially furnish insights into such factors at the regional scale (Shahpari et al., 2021), their availability to practitioners at present is limited.

The period from March to April in the middle and lower reaches of the Yangtze River region typically witnesses total precipitation from 300 to 400 mm, constituting 30% to 40% of the annual rainfall. This timeframe coincides with critical nutritional and reproductive growth stages of wheat crops, and is often punctuated by frequent flood events (Gao et al., 2020). Although wheat has three distinct waterlogging stress responses (tolerant, inhibitory, and adaptive phases) (Shaw and Meyer, 2015), recurring episodes of waterlogging stress can significantly inhibit growth and yields, especially when high-intensity waterlogging coincides with extreme weather fluctuations. This explains why many models perform well with waterlogging stress under experimental conditions, but fall short of accurately predicting outcomes under field conditions (Shaw et al., 2013; Harrison et al., 2019).

Broadly, two approaches are employed to simulate waterlogging in crop models. The first calculates the Stress Day Index (SDI) based on groundwater table depth, as exemplified by DRAINMOD (Skaggs, 2018). However, this approach overlooks genotypic differences in crops, considering only groundwater and surface water, rendering it unsuitable for addressing waterlogging stemming from overirrigation. The second method computes hypoxic stress as a function of soil moisture. Notable examples include CROPR (Qian et al., 2017), SWAGMAN Destiny (Meyer et al., 1996), and the Agricultural Production Systems Simulator (APSIM) (Asseng et al., 1998; Liu et al., 2021). Shaw and Meyer (2015) showed that yield reductions were 6% higher at 65% air-filled pore space compared to 10%, leading to their proposal of three stages for simulating plant responses to waterlogging damage. Their hypoxic stress factor accounts for waterlogging duration and crop tolerance, making it applicable to damage caused by excessive irrigation and facilitating more precise quantitative damage analysis (Shaw et al., 2013). The Crop Estimation through Resource and Environment Synthesis (CERES) model, commonly used for predicting crop growth and yield, has been enhanced by Lizaso et al., who successfully integrated low oxygen (anaerobic) stress factors into the CERES-Wheat model (Lizaso and Ritchie, 1997).

The objectives of this study were thus to quantify waterlogging by integrating computations from both the SWAGMAN Destiny and CERES models. Specifically, we develop five wheat waterlogging injured indices: WI_ASD2 and WI_ASD1 were conceptualized as adaptations of the CERES model (Lizaso and Ritchie, 1997), WI _Destiny was derived from the SWAGMAN Destiny model, WI _Growth introduces differences in waterlogging tolerance across crop growth stages (following WI _Destiny), while WI _Plus considers the influence of soil temperature in addition to WI _Growth. To assess the accuracy of these indicators, we conducted pot experiments with winter wheat (involving single or double flooding) using local genotypes (“Yang mai 11” and “Zheng mai 7698”). By analyzing relationships between our WIs and yield, we provide insight into how integrated physiological determinants impact on crop yield.

The middle and lower reaches of the Yangtze River region are the main wheat-growing areas in China. They are greatly influenced by the subtropical monsoon climate. In the months of March and April each year, extensive waterlogging occurs, and water stress can lead to a decrease in wheat yield or even complete crop failure (Ghobadi et al., 2017). This study has established five WI indices and analyzed them through pot experiments on two wheat varieties, “Yang mai 11” and “Zheng mai 7698,” under different waterlogging treatments.

Among the four waterlogging treatments, there is a strong negative correlation between the duration of waterlogging and the wheat yield increment/decrement rate. In other words, the longer the waterlogging period, the lower the wheat yield increment/decrement rate. When comparing the effects of waterlogging during the jointing stage and the booting stage on wheat yield increment/decrement rate, it was found that waterlogging during the booting stage had a greater impact on wheat yield compared to the jointing stage, which is consistent with the findings of Ghobadi et al. Additionally, the study discovered that, under the same duration of waterlogging, the slope of the yield increment/decrement curve for a single waterlogging treatment was greater than that of the curve for two waterlogging treatments. This could be attributed to the adaptive phase that occurs when the waterlogging duration reaches a critical point (5 days) in wheat growth, where the crop gradually adapts and recovers normal growth. This includes processes such as stomatal regulation, physiological metabolism adjustment, root adaptation, and nutrient regulation to cope with the impact of waterlogged conditions. During this phase, wheat relies on self-regulation and adaptive capacity to withstand the stress caused by waterlogging and maintain growth and yield (Shaw and Meyer, 2015). Therefore, although there is a strong negative correlation between the duration of waterlogging and the wheat yield increment/decrement rate, the linear simulation curve parameters of the waterlogging duration and wheat yield increment/decrement rate are significantly influenced by factors such as the waterlogging period, frequency, and wheat variety. Using only the duration of waterlogging as a universal characteristic to represent the degree of waterlogging in wheat is not highly applicable.

It has been found through research that there is a correlation between the five WI indices and wheat yield. Before reaching the threshold, the fitting curve of the five WI indices and wheat yield shows an upward trend, indicating that a higher WI value corresponds to a higher wheat yield. However, once the threshold is exceeded, the curve starts to decline, and an increase in WI value will lead to a decrease in wheat yield. In the second stage of the fitting curve, the five WI indices show the following characteristics in relation to wheat yield: for the “Zheng mai 7698” variety, the curve’s slope relating to the WI index and yield is greater than that of the “Yang mai 11” variety. This indicates that the “Yang mai 11” variety has stronger tolerance to waterlogging, while the “Zheng mai 7698” variety is less capable of withstanding waterlogged conditions, which is consistent with the findings of Yu J J and others (Yu et al., 2014). Therefore, it can be seen that the WI indices are influenced by the tolerance of the variety to waterlogging, although the present study did not consider the tolerance of different varieties. However, the WI indices can reflect the level of waterlogging tolerance among different wheat varieties.

Many studies have shown a positive correlation between soil respiration and temperature (Lloyd and Taylor, 1994; Das and Paul, 2015). Considering the influence of temperature on soil oxygen content under waterlogged conditions where oxygen diffusion is restricted, the commonly used fixed oxygen threshold may not be suitable. Experimental results on wheat also indicate that the improved WI _Plus is more suitable for quantitatively analyzing the impact of low oxygen stress on wheat yield. The threshold for WI _Plus is 14.9, where WI _Plus ≤ 14.9 (“Zheng mai 7698”) and ≤ 10.8 (“Yang mai 11”) show a positive correlation between WI _Plus and wheat yield, indicating that the wheat is not waterlogged. When WI _Plus > 14.9 (“Zheng mai 7698”) and > 10.8 (“Yang mai 11”), there is a negative correlation between WI _Plus and wheat yield, indicating that wheat is affected by waterlogging, and a higher WI value indicates more severe waterlogging.

This study calculates the soil aeration index based on soil volumetric water content and constructs the WI index using characteristic values of wheat root responses to low oxygen (anoxic) stress, making it applicable in general. Additionally, advancements in remote sensing technology, particularly satellite remote sensing, have shown that soil moisture can be measured using various remote sensing techniques (Wang et al., 2022). Firstly, spectral pictures of wheat planting regions may be collected using high-resolution remote sensing equipment carried by satellites or drones. Secondly, we can preprocess the collected images, including radiation calibration, atmospheric correction, geocoding, and other steps, to eliminate noise and distortion in the images and ensure the accuracy of the data. Thirdly, using image processing technology, extract features related to wheat waterlogging, such as soil temperature and soil moisture content. Finally, the relevant feature quantities were brought into the exponential model of this study, and evaluated and verified through yield data. This can achieve a wider range of wheat stain monitoring. However, its practical application is limited by factors such as insufficient temporal resolution, difficulty in extracting characteristic quantities of waterlogging, and large computational load. Therefore, in the application process, these factors need to be fully considered, and these effects can be reduced through optimization methods.

Currently, there are three remote sensing methods commonly used for soil moisture retrieval: optical remote sensing, thermal infrared remote sensing, and microwave remote sensing. Among them, microwave remote sensing is unaffected by various meteorological conditions and has strong cloud penetration capability. Radar backscatter coefficients are sensitive to changes in soil moisture (Li et al., 2020; Wang et al., 2022). For example, Xiong Q X., et al. used Sentinel-1A data to extract soil surface relative moisture content data. Then, combined with the precipitation index in the previous period, they used the Kalman filter interpolation method to obtain daily soil moisture information in the observation area. In addition, remote sensing technology can also obtain soil temperature information (Xiong et al., 2021). This provides more opportunities for the widespread application of the WI index in monitoring waterlogging. However, since this experiment was conducted in pots, there is minimal vertical variation in soil volumetric water content, which significantly differs from field moisture characteristics. If the results are further applied in field conditions, it would be an area for future improvement.

Using only the duration of waterlogging cannot quantitatively reflect the severity of wheat waterlogging. It is also related to factors such as the tolerance of wheat varieties and the frequency of waterlogging events. Furthermore, the WI index can be applied to assess waterlogging damage under excessive irrigation conditions and can also reflect the impact of multiple waterlogging events on wheat yield. Comparing WI _Destiny, WI _Growth, WI _ASD1 and WI _ASD2, the improved WI _plus shows better performance. When WI _plus is ≤14.9 (“Zheng mai 7698”) or ≤10.8 (“Yang mai 11”), wheat does not experience waterlogging. When WI _plus is >14.9 (“Zheng mai 7698”) or >10.8 (“Yang mai 11”), wheat experiences waterlogging.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

F-LX: Writing-original draft, Formal analysis, Methodology. P-MH: Writing – review & editing, Funding acquisition. XW: Writing – review & editing, Date curation, Investigation. MH: Writing – review & editing, Revise the paper. KL: Writing – review & editing, Revise the paper. Q-XX: Writing – review & editing, Conceptualization, Methodology. All authors contributed to the article and approved the submitted version.