One-third of the population of the world lives in regions stressed by water. With the changes in climatic patterns and frequent global warming reporting in recent decades, the hydrological regimes in which crops grow are getting altered and crop growth is affected (McCarthy et al., 2005). Water availability is a significant cause of variability in crop yields in many conditions (Bal et al., 2022; Waseem et al., 2022). Changes in seasonal precipitation, in-season precipitation, and inter-annual variation of precipitation affect the soil water regime. Precipitation is recognized globally as a leading factor affecting, in particular, the yield of rainfed crops (Izaurralde et al., 2003). The Indian sub-continent experiences large-scale drought in some part or the other, almost every year (Zhang et al., 2016). Drought occurs in nearly all climatic zones of the world at one time or another, but this creeping phenomenon mostly affects the tropics and adjoining regions (Khanday and Javed, 2008). Even though, the relationship between soil moisture and rainfall is consistent but non-linear (Douville et al., 2001). In the preview of such variability, the seasonal rainfall variability affects the variability in soil moisture, which influences the wetness/dryness of monsoon season needs to be understood (Douville et al., 2001; Bal et al., 2022). Many past researches on soil moisture as a boundary forcing to the atmospheric precipitation suggested that reduced soil moisture during the monsoonal periods could induce a dry summer condition over a particular region (Namias, 1960). Therefore, the study of soil moisture is imperative to understanding the behavior of the Indian monsoon system. “Drought Declaration” means the beginning of the government’s response to circumstances that reflect a situation of drought. The decision to declare the drought over a designated administrative unit level (District/Taluka/Tehsil/Block/Gram Panchayat) should be directed and made without undue delay by objective criteria so that relief assistance and concessions can be given in time to the citizens affected by drought (Neenu et al., 2013). Five categories of indices are recommended to construct a drought matrix by the Government of India (Anon, 2016). They are rainfall-related indices (rainfall deviation, dry spell, standardized precipitation index), remote sensing-based vegetation indices, crop situation-related indices (area under sowing), soil moisture-based indices (percent available soil moisture, moisture adequacy index) and hydrological indices (reservoir storage index, groundwater drought index, stream-flow drought index).

These drought declaration criteria are general, i.e., independent of crops. But, in many environments, water supply is a major source of variability in crop yields (Ritchie, 1983). Deficit precipitation affects crop yields, especially if rainfall doesn’t occur in time or dry periods occur during critical development stages. Moisture stress during the flowering, pollination, and grain-filling stages is especially harmful to maize (Denmead and Shaw, 1960 ), soybean ( Bharat et al., 2019), wheat (Decker et al., 1986 ) and sorghum (Machado et al., 2002). In such a situation, available soil moisture is a very relevant indicator of drought, especially in regions fed by rain (Stephens and Lyons, 1998).

Soil-moisture balance technique may be used to measure the soil moisture-based indices, which includes gathering some of the baseline data related to soil properties, climate parameters, and pattern of crop growth (Doorenbos and Pruitt, 1977). Soil moisture balance measures the amount of rainfall accessible to crops depending upon crop water requirement, climatic evaporative demand, and soil water holding capacity (Betts, 2004; Shewale and Shravankumar, 2005). In the latest revision of the Drought Manual (2016), brought out by the Government of India, Percent Available Soil Moisture (PASM) was included as one of the parameters for the drought declaration protocol.

Though the inclusion of PASM has improved the drought declaration protocol, sometimes farmers could not get assistance because of non-declaration of drought even if there was a significant loss in crop yield. Hence, the present study was conducted to fine-tune the relationships between soil moisture availability and with yield of major rainfed crops in various locations of peninsular (central and southern) India to refine the drought categories using PASM. The following research questions were tried to address in the present study 1) Does the existing PASM criteria can precisely capture severe and moderate agricultural drought? 2) If no, how to re-evaluate the criteria and make suitable revision?

Percent available soil moisture is a useful indicator for monitoring crop failure because it reflects the amount of water present in the soil that is available for plant uptake. The PASM includes the component of moisture content at field capacity (FC) and permanent wilting point (PWP). The range between field capacity and the permanent wilting point defines the plant-available water, representing the portion of soil moisture that plants can utilize for growth. When the percent available soil moisture drops below a certain threshold, it can indicate that the plants are experiencing water stress, which can ultimately lead to crop failure. To use percent available soil moisture for monitoring crop failure baseline values for each crop type and soil type under normal growing conditions need to be established. This can be done by measuring the percent available soil moisture throughout the growing season when the crop is healthy and growing normally. Once the baseline values are established, the percent available soil moisture can be used to monitor crop failure by comparing the current moisture levels to the established baseline values. If the current moisture levels fall below the baseline values, it can indicate that the plants are experiencing water stress, which can lead to reduced growth, yield, and eventually crop failure.

The PASM is used as one of the impact indicators (triggers) in the declaration of agricultural drought (Anonymous, 2016). As per the drought manual published by the Ministry of Agriculture and Farmers’ Welfare, Government of India, weekly/crop stage-wise PASM was calculated based on observed soil moisture data using the following formula, PASM  % = SMW − PWP FC − PWP ∗ 100

Where, SMW = Weekly calculated volumetric soil moisture (vol./vol.) for the current week/crop stage; PWP = Permanent wilting point of soil (vol./vol.); FC = Field capacity of the soil (vol./vol.)

Several historical attempts have been made to relate the influence of weather parameters on soil moisture and its impact on crop yield (Table 2). These include aspects of relating climatic data to the occurrence or non-occurrence of rainfall, onset and duration of dry spells, etc. However, in recent years, due to technological advances in data analytics progress, studies using high-resolution time series data are being carried out in the area of rainfall-induced soil moisture variability (Chen and Zhou, 2021). Based on reviewing previous works, the present study was conducted through measurement of week-wise soil moisture, classifying it into existing drought norms, calculating weighted yield based on the drought severity during the study period, checking for a relationship using correlation, drawing a regression equation and yield estimation using drawn equations.

In the present study, crops were grown during the Southwest monsoon period and sown on different dates of sowing. On each date of sowing, weekly soil moisture observations were taken and PASM was calculated based on the existing classification (Anon, 2016).

Since the crops were grown under different dates of sowings during the study period, early sown crops may experience drought at later reproductive or maturity, late sown crops may experience mid-season drought due to erratic onset and cessation of monsoon in the study locations (which is usually June to September). Such data on long-term experiments provide a base to understand crop yield under various drought or moisture stress scenarios at different growth stages of a particular crop. The procedure of arranging weekly/crop stage-wise root zone soil moisture, PASM, and grouping into different drought classes is presented in Table 3.

The methodology to calculate the weighted yield as per the PASM drought classes is provided in Figure 3. In general, the crop yield would decline, if the crop has passed through several severe moisture stress weeks, and vice versa. Several of such severe/moderate/mild and no drought situations along with their yield levels during the study period were averaged to obtain a comprehensive summary of yield and PASM for producing that particular yield. And, this summary provides a way for further correlation and regression analysis.

The average yield and PASM for each group were calculated and tabulated for regression analysis. The PASM versus yield in each group was correlated to check for the relationship if found positively significant, regression equations were developed. Best fitting (estimated by checking R ² value) equations were employed to find the yield of the crop at a particular PASM and PASM required for obtaining 50% of the optimal yield (optimal yield is an average yield of the crop under farmers’ practice and yield of the crop obtained under unlimited resources of experimental fields of a university or institute).

Further, the estimation of the PASM requirement to attain at least 50% of optimal yield using the earlier developed regression equation was carried out to indicate meaningful relief assistance to the farmer under drought situations. i.e., if a farmer gets at least 50% of yield, he will be able to meet the expenses of the cost of cultivation. Hence, the PASM-yield relationship establishment in individual crops lays a scientific base for the fixation of PASM criteria for individual crops for timely and appropriate drought declaration and thereby meaningful drought relief assistance.

Experiments carried out at Kovilpatti indicated that the PASM and yields were positively correlated (0.98) (Table 3). The effect of low soil moisture during the reproductive stage was more as compared to vegetative and maturity stages, i.e., severe drought during the reproductive stage has produced 3209 kg/ha yield as compared to that produced due to drought in vegetative and maturity stages (3385 and 3216 kg/ha, respectively) (Figure 4; Supplementary Table S1).

The finger millet crop is hardy and resistant to flood, drought, pests, and disease and is mainly grown in drylands of south India, whenever rainfall is not sufficient enough to cultivate other major crops. It is raised as a sequence crop of paddy, based on residual moisture. Even though it is a drought resistant crop, soil moisture variability during different growth stages majorly decides its yield levels. The results of this study showed that finger millet yield had a significant positive correlation with soil moisture (0.92) (Table 3). The yield of the crop was as low as 0 kg/ha under severe drought (20% PASM) to 4236 kg/ha under no drought situations (100% PASM) (Figure 5).

The soil moisture availability at different growth stages has influenced the yield of soybean at Parbhani (Supplementary Table S2). The yield variation due to available soil moisture at different stages ranged from 135 to 1803 kg/ha. The effect of low moisture availability at the vegetative and reproductive stages was greater than the effect in the maturity stage (Figure 6). For example, a severe drought at the vegetative and reproductive stages of the crop produced only 135 and 178 kg/ha yield, respectively. However, the yield was 405 kg/ha even though the crop suffered from a severe drought during maturity.

Correlation between PASM and yield under different sowing windows at Akola showed yield increases with delay in sowing, i.e., the magnitude of relation was as low as 0.62 for the crop sown in the first week of July and 1.0 in the last week of July (Table 3). As the crop is sensitive to moisture stress, a dry spell phase of 2 weeks during the mid-vegetative stage caused a 56 percent yield reduction while 39 per cent yield reduction was observed at any of the crop stages.

The groundnut yield variability due to available soil moisture at different stages was in the range of 399–1999 kg/ha at Ananthapuramu. The effect of low moisture availability at the emergence, vegetative, pegging, and pod development stages is higher as compared to that at flowering and pod maturity stages (Figure 7; Supplementary Table S3). At Ananthapuramu, the relationship between soil moisture and PASM was polynomial, and due to severe moisture deficit at emergence, vegetative, pegging, and pod development stages, groundnut productivity was 399, 766, 867, and 661 kg/ha yield, respectively. However, even though the crop suffered severe drought during the flowering and maturity stages, it could achieve higher yields of 1147 and 970 kg/ha, respectively.

While, in Bengaluru, a correlation up to the extent of 93 percent was observed (Table 3). This indicates the impact of higher available soil moisture on pod yield, as evidenced by the significant positive correlation between soil moisture content at different stages with pod yield. The correlation was more significant during the later stages of the crop.

Cotton being one of the major commercially cultivated crops in the drylands of India gets affected by both intra and inter-seasonal droughts. The most discernible response due to soil moisture deficit across different stages of the crop is a reduction in plant stature and other growth parameters. Correlation between PASM and cotton lint yield was positive, i.e. 0.96 and 0.93 (Table 3) in early (June 2FN) and late sown (August 1FN) conditions, respectively indicating a strong relationship between them (Figure 8).

Drought in different crop stages showed varied impacts on cotton yield sown under different sowing windows (Supplementary Table S4). In each sowing window (June 2FN, July 1FN, and July 2FN), a severe drought at the vegetative stage produced lower lint yields (277, 315, and 267 kg/ha lint yield) as compared to other two growth stages. Further, there was a non-significant improvement in lint yield under severe drought both at reproductive and maturity stages since these stages coincide with the high rainfall period (July) of the region.

In the case of late sown conditions (August), the lowest lint yield was recorded due to severe drought at the maturity stage (428 kg/ha) as compared to vegetative (514 kg/ha) and reproductive stages (456 kg/ha).

Correlation between PASM and crop yields under different sowing dates was found to vary with crops and locations. This is because the sensitivity of crops to soil moisture deficit varies from crop to crop and also with stages. Further, the soil water regime is affected by changes in seasonal soil moisture availability, and inter-annual variation of the same due to respective variability in the precipitation (Izaurralde et al., 2003). The higher correlation between crop yield and PASM was due to the dependence of the crop entirely on rainfall to fulfill its water requirements (Rossato et al., 2017). This higher correlation was associated with periods of drought in peninsular India, caused by high seasonal and inter-annual variability of rainfall during the study period. Previous studies on the relationship between vegetation indices (NDVI) and root zone soil moisture (RZSM) in African regions indicated the existence of positive relationships and the association varied with geographic location and lag, which implies that an increase in soil moisture due to the ENSO could promote vegetation greenness (Sazib and Bolten, 2020).

Maize is highly susceptible to atmospheric stresses during the initial stages of reproduction viz., silking, and blistering. Water and heat stress can significantly limit the ability of maize to properly pollinate (Robel et al., 2019). During subsequent reproductive stages, the maize kernels are still small, but water constitutes about 85% of the kernel mass, which is the highest over the growing cycle. During milk, dough, and dent stages, the kernel size significantly increases; however, the amount of water decreases to about 55% of the kernel mass, further dropping to 30% by maturity. Stress during the late reproductive stages, affects dry matter accumulation (Iliana et al., 2017). Soil moisture stress during the vegetative stage will register a lower impact on maize yield compared to stress during the early reproductive stages. Rohit et al. (2020) reported a 13.5% average reduction in maize yield when the stress occurred during the vegetative stage while a 53% reduction in yield occurred with the stress during the reproductive stage. However, prolonged drought during the vegetative stage can also affect the yield significantly as reported by Eck (1986). He observed that yield reduction in maize was 23% for a two-week-long water stress period during the reproductive stage compared to 46% when the stress continued for 4 weeks during the vegetative period.

The drastic yield reduction due to severe drought was evident in the present study. Similar results of soil moisture effect on yield variability were recorded by Muhammad and Azam (2007) indicating a significant effect (p < 0.028) of drought on finger millet yield.

The yield was not drastically reduced even under a severe drought during the maturity stage. This may be due to the well-established root system (Comas et al., 2013) as well as lower water requirement by the crop at the maturity stage compared to the previous stages. Differences in correlation coefficients were mainly due to the soil properties of this region. In typical black cotton soils at Akola, most of the first rainfall received was infiltrated into deeper layers of the soils through widely opened cracks in the previous summer. The rainfall analysis also indicated that there was higher rainfall in July (223.3 mm with 40.6% CV) facilitating better soil moisture availability (Staton, 2020). In regions like central India, variabilities in soybean yield to a significant extent were recorded by Mohanty et al. (2015), and a decrease in yield up to 96 per cent was recorded due to heavy rainfall during flower initiation to maximal pod stage.

In Anantapur, the yield levels after 75 percent of PASM declined mainly due to heavy soil type and excessive rainfall during the later stages of the crop, making poor pod development (Kumar et al., 2012). However, the overall soil moisture content during the total crop period was also significantly and positively correlated with pod yield (Guled et al., 2013 ) in Bangalore.

In Parbhani, both at reproductive and maturity stages coincide with the high rainfall period (July), resulting in improved lint yield even under severe drought. Another reason may be the deeper roots of cotton could have accessed soil moisture from the deeper layers of typical vertisols of the region under moisture stress situations.

In the case of late sown conditions (August), the lowest lint yield was recorded due to severe drought at the maturity stage (428 kg/ha) as compared to vegetative (514 kg/ha) and reproductive stages (456 kg/ha). Under late sown conditions, the higher utilization of available soil moisture during the early stages of crop growth led to lesser soil moisture availability during the reproductive stage. This along with severe drought at the maturity stage, caused lesser lint yield. Pettigrew (2004) also reported that in cotton, no severe impact of moisture deficit on any of the growth parameters was observed in early sown cotton. However, in the late-sown crops, a severe moisture deficit stress developed and impacted most of the growth and yield parameters.

As optimum moisture at the flowering and pod filling stage is crucial for achieving the potential yield, moisture stress at these stages affected its yield. Pigeon pea seed yield was reduced to 37% due to moisture stress at the flowering and pod-filling stages and not significantly affected by stress at the pod-filling stage alone ( Lopez et al., 1996 ). Earlier research also indicated that the productivity of pigeon pea was more influenced by soil depth vis-à-vis moisture availability, i.e., the seed yield of pigeon pea was higher by 46–66 percent due to abundant soil moisture in 20–40 cm soil depths over other lower soil depths (Subbareddy et al., 2001; Kovac et al., 2005).

The study was conducted in 5 locations representing two broader soil types involving 6 major rainfed crops. It is a well-known fact that the percent available soil moisture varies depending on soil type, climate, and other environmental factors. However, Percent available soil moisture and soil type are two important factors that are closely related and can have a significant impact on plant growth and crop yield. Soil type refers to the physical and chemical properties of the soil, including its texture, structure, and nutrient content. Different soil types have varying water-holding capacities and drainage properties, which can affect the percent available soil moisture (Wang et al., 2019). A study assessed the soil moisture regimes in sandy soils of Nigeria and their impact on crop production. The results showed that sandy soils had a low water holding capacity, resulting in low available soil moisture, which negatively impacted crop growth and yield (Fashae and Akintoye, 2019). In his study, a comparative analysis has not been done to see how the yield level of a single crop under different soils under different available soil moisture impacts the crop yield.

The study revealed that the yield was significantly influenced by the per cent Available Soil Moisture (PASM) for six different crops. The PASM for obtaining 50% of optimum yield in cereals (maize: 26% and finger millet: 52.9%), pulses (pigeon pea: 37.2%), oilseeds (soybean: 26.8%–30.5%, groundnut: 53.8%–61.7%) and commercial crops (cotton: 26.3%) ranged from 26 to 61 per cent. The criteria followed for considering <25 PASM as severe drought as per the Drought Manual 2016 of the Govt. of India needs to be modified given the findings of the present study.

The PASM, though an important criterion for drought declaration, its applicability was overlooked due to a lack of understanding of its effect on individual crops at many locations led to unscientific drought declaration vis-a-vis its efficient management so far in the country. The present study highlighted the crop performance especially yield variation at different soil moisture levels and crop stages. The following amendments are put forth based on the findings of this study and they are; 0–50 PASM as severe, 51–75 as mild, and 76–100 as no drought situation. This study indicates the need to re-look into the drought declaration norms presently followed for these crops. The same criteria which are closer to the actual field conditions may be followed by planners of the country to assess and declare the agricultural drought. This study has only considered the relation between weekly/crop stage-wise PASM and crop yield and there are other factors viz., biotic and abiotic factors also which affect crop performance. But, in dryland ecosystems, soil moisture availability is the most dominant factor affecting crop performance. Therefore, the outcome of this study signifies the inclusion of PASM as one of the factors for agricultural drought assessment in India.