Research Article

Assessing the Maize (Zea mays L.) Crop Performance Using Spectral Indices under Different Sowing Dates and Irrigations Schedules

Ch. Pallavi, G. Sreenivas, M. Yakadri, Anima Biswal, A. Madhavi, P. D. Sreekanth and B. Laxman

  • Page No:  319 - 331
  • Published online: 27 Aug 2021
  • DOI : HTTPS://DOI.ORG/10.23910/1.2021.2411a

  • Abstract
  •  pallavireddy8990@gmail.com

Field experiment was conducted at ARI, PJTSAU, Hyderabad with four dates of sowing (18th June, 04th July, 19th July and 03rd August in Kharif, 2016 and 01st November, 18th November, 01st December and 17th December in rabi, 2016–17) as main plots and four irrigations regimes (Control, 0.4 IW/CPE, 0.6 IW/CPE and 0.8 IW/CPE in kharif and 0.4 IW/CPE, 0.6 IW/CPE, 0.8 IW/CPE and 1.0 IW/CPE in rabi) as sub-plots in split plot design replicated thrice. Scheduling irrigation at unstressed conditions, 0.8 IW/CPE (I3) and 1.0 IW/CPE (I4) of maize resulted in low reflectance in visible region (400 to 700 nm) and mid infrared (MIR) region (1350–2500 nm) and high in near infrared (NIR) region (700 to 1350 nm) during kharif and rabi respectively. While, under stressed condition, the reflectance was high in visible and MIR region and low in NIR region in rainfed and 0.4 IW/CPE (I0) respectively in kharif and rabi. Significantly higher drymatter, LAI and grain yield was observed in 04th July (D2) and 01st November (D1) sown crop in kharif and rabi respectively. However, spectral indices (SR, PRI, NDWI at 1240, 1640 and 2130 nm, NDII, NMDI, WBI and SWRI) was attained by 18th June (D1) and 01st November (D1) during kharif and rabi respectively. Higher drymatter, LAI, grain yield and spectral indices was recorded with I3 (0.8 IW/CPE) and I4 (1.0 IW/CPE) in kharif and rabi respectively. All the spectral vegetation indices correlated positively with LAI, drymatter and grain yield.

Keywords :   Maize, remote sensing, spectral indices, stress, water, yield

  • Introduction

    Cereal grains are important staple foods providing substantial amounts of energy, protein and micronutrients for much of the world’s population (Murdia et al., 2016). In cereals, maize (Zea mays L.) is grown throughout the year mainly due to photo-thermo-insensitive character, hence called ‘queen of cereal’ (Verma et al., 2012). Maize is one of the most versatile cereal crops having wider adaptability under varied agro-climatic conditions (Ramachandiran et al., 2016). Maize is the third most important cereal crop in India after rice and wheat (Singh et al., 2012). Its importance lies in the fact that it is not only used for human food and animal feed but at the same time it is also widely used for corn starch industry, corn oil production, baby corns etc (Parihar et al., 2011). It accounts for 9% of total food grain production in the country (Kumar et al., 2013). In India, maize is cultivated in an area of 9.02 mha with a production of 27.7 mt and productivity of 3.07 t ha-1 in 2019 (Anonymous, 2020). Maize crop growth is affected by different stresses viz., water, pest, weed, nutrients, etc., which reduce the productivity (Szeles et al., 2012). Stresses come from a range of factors that limit the potential growth of canopies (Prakash and Singh, 2020). Evaluation of the stress level to which plants are subjected to is therefore vital information required both for the quantification of consequences on production and for taking action for their mitigation (Ma et al., 2020). Plant physiology is significantly affected by abiotic/climatic stresses. It is well known that climate change and environmental extremes induce and enhance the impact of abiotic stresses (particularly drought) on plant fitness and performance (Kaushal and Wani, 2016). 

    Remote sensing is a better method to detect and quantify the impact of plant stress compared to visual techniques because a vegetative unit can be repeatedly, objectively, and nondestructively examined in a fast, robust, accurate, and inexpensive way (Mirik et al., 2012). Earlier research on applications of remote sensing in agriculture focused on the visible and near-infrared regions of the spectrum. The launch of new hyperspectral remote sensing satellites, the focus of current research has been shifted to investigate the possibilities of hyper spectral remote sensing in crop monitoring. Hyper spectral remote sensing techniques allow the early detection of vegetation stress, before the appearance of visible symptoms (Panigada et al., 2010). Hyperspectral technology can sometimes be considered as a part of spectroscopy. The electromagnetic spectrum ranges of hyperspectral sensors mainly concentrate on VIS–NIR (400–1000 nm) and sometimes contain a short wave infrared range (SWIR, 1000–2500 nm). These sensors could acquire spectral information from hundreds of narrow spectral bands (Ghamisi et al., 2017). Using that absorption and spectral reflection characteristics of vegetation researchers has defined many vegetation indices for monitoring vegetation parameters. Thus, imaging-based and RS techniques have attracted considerable attention because they can provide accurate geographic information (Ali et al., 2019).

    Spectral vegetation indices were designed to evaluate vegetation condition, foliage, cover, phenology and processes in addition to be used for land cover classification, climate and land use detection, drought monitoring and habitat loss (Padilla et al., 2011). More recently, spectral vegetation indices are mathematical expressions involving reflectance values from different part of the electromagnetic spectrum, aimed to optimize information and normalize measurements made across varied environmental conditions (Mirik et al., 2012). The advent of spectral mapping and identification of vegetation by remote sensing methods, it assumes importance to study crop stress using remote sensing techniques in maize under different water stress conditions. Keeping these in view an experiment was planned for determination of water stress with hyper spectral reflectance and vegetation indices on maize.


  • Materials and Methods

    2.1.  Site description

    The field experiment was conducted at Agricultural Research Institute, Professor Jayashankar Telangana State Agricultural University, Rajendranagar, Hyderabad having 17.019’ N Latitude, 78.023’ E Longitude and 542.3 m above mean sea level. The soil of the experimental site was sandy loam in texture, neutral in reaction, low in available nitrogen, phosphorus and high in available potassium, during Kharif, 2016 and rabi, 2016–17 with four dates of sowing (18th June, 04th July, 19th July and 03rd August in Kharif, 2016 and 01st November, 18th November, 01st December and 17th December in rabi 2016–17) as main plots and four Irrigations regimes (Control (Rainfed),  0.4 IW/CPE, 0.6 IW/CPE and 0.8 IW/CPE in Kharif, 2016 and 0.4 IW/CPE, 0.6 IW/CPE and 0.8 IW/CPE and 1.0 IW/CPE in rabi, 2016–17) as sub-plots in split plot design replicated thrice. During the crop period rainfall of 834.9 mm was received in 45 rainy days in kharif, 2016 and 11.8 mm in 2 rainy days in rabi, 2016–17, respectively. A uniform dose of 60 kg ha-1 P2O5 as single super phosphate, potassium @ 60 kg ha-1 as muriate of potash and ZnSO4 @ 50 kg ha-1 was applied to all the treatments. The entire P2O5,ZnSO4 and half of K2O were applied at sowing. Nitrogen was applied as per the treatments (wherever it was required) in the form of urea (46% N) in three equal splits (1/3rd each at basal, at knee-high and tasseling). Similarly, the remaining potassium was applied along with urea during second top dressing at tasseling.

    Leaf area was estimated on two plants in each plot at silking and the area of total leaves was measured with digital leaf area meter (LI-3100) and expressed in cm. Leaf area index was calculated by using the formula as proposed by Watson (1952). Two successive plants from the sampling row were uprooted at silking and were shade dried followed by oven dried at 65°C±5°C till constant weight obtained. The kernels from the air-dried cobs from each net plot were separated, cleaned and dried to obtain at least 13% moisture. Weight of grains of each plot was recorded separately and expressed as grain yield in kg ha-1.

    2.2.  Description of spectroradiometer

    Hyper spectral data were collected by using FieldSpec® 3 Spectroradiometer. The Field Spec® 3 Spectrometer is a general purpose spectrometer useful in application area requiring the measurement of reflectance, transmittance, radiance or irradiance. It is specifically designed for field environment remote sensing to acquire visible near infrared (VNIR) and short-wave infrared (SWIR) spectra. The Analytical Spectral Device (ASD) covers the spectral range between 350 to 2500 nm. The sampling interval over the 350–1000 nm range is 1.4 nm with a resolution of 3 nm (band width of half maximum). Over the 1000 -2500 nm range, the sampling interval is about 2 nm and the spectral resolution is between 10 and 12 nm. The results are then interpolated by the ASD software to produce readings at every 1 nm. A 1.2 m long fiber optic cable with a 250 field of view was used for the measurements. Spectral reflectance was derived as the ratio of reflected radiance to incident radiance estimated by a calibrated white reference (Spectralon).

    2.3.  Measurement of canopy reflectance

    The canopy reflectance were measured in the spectral range of 350-2500 nm at sampling interval 1.4 nm @ 350-1000 nm and 2 nm @ 1000-2500 nm with the help of ASD FieldSpec® 3 Spectroradiometer (Analytical Spectral Devices Inc., Boulder, CO, USA). The reflectance measurements were made on sunny days between 11.00 and 13.00 hours in all plots. The FOV was 250 and the distance between the optical head of the spectroradiometer and the top of the plant was kept at 1 m for all observations. For optimization of ASD instrument, a Spectralon (Labsphere, Inc., Sutton, NH, USA) white reference panel was used to obtain reference signal prior to each canopy reflectance measurement.

    The canopy reflectance was obtained after calibration with the white reference panel. Vegetation radiance measurement was taken by averaging 10 scans at an optimized integration time, with a dark current correction at every spectral measurement. The spectral readings were taken from each plot at silking of maize crop during both kharif, 2016 and rabi, 2016–17.

    2.4.  Processing of spectral responses

     The spectral data was processed and exported to MS excel through ASD View Spec Pro (Version 6.2) software. Therefore, following narrow band and hyperspectral growth (i-iii) and canopy water status (iv-xi) indices were calculated using the following formulae (Table 1).


    2.5.  Statistical analysis

    The data were analyzed statistically applying analysis of variance technique for split plot design. The significance was tested by ‘F’ test (Snedecor and Cochron, 1967). Critical difference for examining treatment means for their significance was calculated at 5% level of probability (p=0.05).  The influence of spectral indices was studied by comparing each date of sowing at optimum irrigation regimes (1.0 IW/CPE), thus constituting four treatment combinations. Hence, maize crop growth period was divided into five phenophases viz., P1=Emergence to kneehigh stage; P2=Kneehigh to tasseling stage; P3=Knee high to silking stage; P4 =Silking to dough stage and P5=Dough to physiological maturity stage. Means of all spectral indices that prevailed over different phenophases were calculated. The correlation coefficients were also worked out between spectral indices during different phenophases, dry matter, RWC, CTD, LAI, Yield attributes and grain yield of maize. Step wise regression (Draper and Smith, 1996) was carried out considering those spectral indices, which had significant influence on crop growth, yield and yield attributes.


  • Results and Discussion

    3.1.  Spectral signature in maize

    In visible region (350-700 nm), the reflectance was higher (0.109 and 0.105) in19th July (D3) and 10 August (D4) sown crop at silking stage compared to 18th June (D1) and 04th July (D2) sowings (0.072 and 0.069) during kharif, 2016. Reflectance was recorded higher (0.514) in 18th June (D1) at silking during kharif, 2016 followed by 04th July (D2) and 19th July (D3) while, lowest reflectance (0.371) was observed with 10 August (D4) sown crop (Figure 1) at NIR region (700 to 1350 nm). While, mid infrared (MIR) region (1350-2500 nm), the reflectance was higher (0.245) in 10 August (D4) sown crop at silking stage followed by 18th June (D1) (0.225), 19th July (D3) (0.219) and 04th July (D2) (0.195) sown crop.


    During rabi, 2016–17 in visible region (350–700 nm), the reflectance was higher in 01st December (D3) (0.133) sown crop at silking stage followed by 17th November (D2) (0.115), 01st November (D1) (0.081 and 0.081) and 18th December (D4) (0.080 and 0.052) sown crop. Higher reflectance (0.568) was observed in 17th November (D2) followed by 01st November (D1), 01st December (D3) and 18th December (D4) sown crop at silking at NIR region (700 to 1350 nm). While, mid infrared (MIR) region (1350-2500 nm), the reflectance was higher (0.328) in 17th November (D2) sown crop at silking stage followed by 01st December (D3) (0.281), 18th December (D4) (0.215) and 01st November (D1) (0.212) sown crop. As crop advanced from knee high to silking, the leaf area and crop drymatter increased. The increased chlorophyll with the advancement of crop stage caused decreased reflectance in visible regions due to strong absorption by pigments whereas increase in canopy coverage resulted in increased reflectance in NIR region. As plant senescence from dough to maturity stage, reflectance increased in visible region and decreased in NIR region probably due to decrease in pigments and mobilization of N from leaves to grains (Rani et al., 2016).

    It has been reported in literature that five absorption bands centered 950–970, 1150–1260, 1450, 1950 and 2250 are related to water absorption. Reflectance at these wavelengths show sensitivity to absorption to water and therefore has promising potential in estimating canopy water content. Since, the reflectance data at 2500 nm show a lot of noise due to presence of atmospheric water vapour content, 2250 nm wavelength was selected instead of 2500 nm. In visible region (350–700 nm), the reflectance was higher (0.118) inrainfed (I0) at silking compared to 0.4 IW/CPE (I1), 0.6 IW/CPE (I2) and 0.8 IW/CPE (I3) during kharif, 2016. While, in rabi, 2016-17 the reflectance was higher (0.109) in 0.4 IW/CPE (I1) at silking respectively followed by 0.8 IW/CPE (I3), 0.6 IW/CPE (I2) and 0.4 IW/CPE (I1). Higher reflectance (0.493) was observed in 0.8 IW/CPE (I3) followed by 0.6 IW/CPE (I2), 0.4 IW/CPE (I1) at silking during kharif, 2016 (Figure 1) at NIR region. During rabi, 2016-17 higher reflectance (0.514) was observed in 1.0 IW/CPE (I4) followed by 0.8 IW/CPE (I3), 0.6 IW/CPE (I2) at silking (Figure 2).


    Whereas, lowest reflectance was  observed with Rainfed (I0) and 0.4 IW/CPE (I1) during kharif, 2016 and rabi, 2016-17 respectively at NIR regions (740-1350 nm). Maize reflectance at mid infrared (MIR) region (1350-2500 nm), the reflectance was higher (0.255) was observed in rainfed (I0) at silking and dough stage respectively over 0.4 IW/CPE (I1), 0.6 IW/CPE (I2) and 0.8 IW/CPE (I3) during kharif, 2016.  While, in rabi, 2016-17 the reflectance was higher (0.289) in 0.4 IW/CPE (I1) at silking followed by 0.8 IW/CPE (I3), 0.6 IW/CPE (I2) and 0.4 IW/CPE (I1).

    Scheduling irrigation at unstressed conditions, 0.8 IW/CPE (I3) and 1.0 IW/CPE (I4) of maize resulted in low reflectance in visible region (400 to 700 nm), high in near infrared (NIR) region (700 to 1350 nm) and low in mid infrared (MIR) region (1350–2500 nm) during kharif, 2016 and rabi, 2016-17 respectively. Under stressed condition, the reflectance was high in visible region and low in NIR region and high in MIR in rainfed and 0.4 IW/CPE (I0) respectively in kharif, 2016 and rabi, 2016–17 compare to unstressed maize. The spectral data indicated that in the ultra-violet visible region (350-710 nm), the spectral reflectances were almost same and did not change with increasing water stress whereas, in NIR regions (740-1350 nm) the reflectance tended to increase with decreasing water stress, resulting in higher differential response at NIR than in visible regions. At higher wavelengths, in MIR (1350–2500 nm) is dominated by soil and leaf water absorption, particularly at 1400 and 1900 nm with reflectance increasing when leaf liquid water content decreases but their response to varying water stress was low compared to that in NIR region (Huete, 2004; Genc et al., 2013). This scattering occurred deep within the leaf tissue, and hence percentage light reflected in the NIR region provided details on the physiological condition of plant under stress (Nilsson, 1995; Hatfield and Pinter, 1993).  The decline in reflectance was due to the reduced turgidity of spongy - mesophyll layer in water stressed compared to the turgidity levels of the crop received more irrigation (Jayasree et al., 2013).

    3.2.  Spectral vegetation indices

    Spectral vegetation indices are mathematical expressions involving reflectance values from different part of the electromagnetic spectrum, aimed to optimize information and normalize measurements made across varied environmental conditions. The reflectance data were transformed in to vegetation indices viz., Simple Ratio (SR), Photochemical Reflectance Index (PRI), Hyperspectral Normalized Difference Vegetation Index (hNDVI), Normalized Difference Water Index (NDWI1240), (NDWI1640), (NDWI2130), Normalized Difference Infrared Index (NDII), Normalized Multi-band Drought Index (NMDI), Moisture Stress Index (MSI), Water Band Index (WBI), Simple Ratio Water Index (SRWI) were used to distinguish water stress severity in maize at silking during kharif, 2016 and rabi, 2016–17 (Table 2).


    Simple ratio is one of the most important ratio indices to represent plant health and dry matter production. Highest value of SR, PRI, NDWI at 1240, 1640 and 2130 nm, NDII, NMDI, WBI and SWRI was attained by 18th June (D1) sowing followed by 04th July (D2), 19th July (D3) and 10th August (D4) while, 01st November (D1) followed by 17th November (D2), 01st December (D3) and 18 December (D4) during kharif, 2016 and rabi, 2016-17 respectively at all the growth stages. The data revealed that highest SR, PRI, NDWI at 1240, 1640 and 2130 nm, NDII, NMDI, WBI and SWRI values were recorded with 0.8 IW/CPE (I3), 0.6 IW/CPE (I2), 0.4 IW/CPE (I1) and rainfed (I0) during kharif, 2016 however, 1.0 IW/CPE (I4), 0.8 IW/CPE (I3), 0.6 IW/CPE (I2) and 0.4 IW/CPE (I1) during rabi, 2016-17 respectively.

    Similar results in wheat for different spectral indices (Prasad et al., 2007). PRI index as a measure of light use efficiency by wheat crop (Wu et al., 2008). Combination of SWIR and NIR is necessary to improve the accuracy in estimating vegetation water content at the leaf level from optical observations (Ceccato et al., 2002). The response of the NDWI1240 index was ten times as strong for the stress group as for the control in the early stages of the stress (Bayat et al., 2016). This index is a measure of liquid water molecules in vegetation canopies that interacted with the incoming solar radiation. The NDWI was proposed by Gao (1996) as a tool to assess water stress and was used to estimate vegetative water content (Jackson et al., 2004). NDWI1940 water indices may have special sensitivity to severe drought conditions given its high sensitivity to low EWT (Ranjan et al., 2015). It was observed that at booting stage, EWT and RLWC were significantly correlated with most of the hyperspectral indices whereas correlation of VWCd was significant with only WBI (Ranjan et al., 2015). The hyperspectral indices which have been identified in this study as good estimators of crop water status include reflectance at 860 nm wavelength in their formulae (Jacquemoud and Baret, 1990).

    NDVI values increased with advancement of vegetative phase and reach highest level silking stage which also coincide with maximum leaf area index. Thereafter, NDVI values decreased as the crop progressed towards physiological maturity. Among the sowing environments, NDVI was higher under early sown crop at all growth stages followed by timely and late sown crop in both the seasons. Under incremental irrigation scheduling, hNDVI values were highest when irrigation scheduled at 0.8 IW/CPE (I3) and 1.0 IW/CPE (I4) during kharif, 2016 and rabi, 2016-17 respectively at all growth stages. NDVI values increased till production of maximum LAI by wheat crop and thereafter decreased towards maturity due to decrease in LAI. This index behaved closely with leaf area production by the crop. So, early sown crop showed greater NDVI values as compared to other sowing dates in 2010–12 (Verma et al., 2010).Unlike other ratio indices, MSI followed just the opposite trend. First sowing date recorded the lowest value of MSI, followed by remaining sowing dates during both the seasons. The lowest MSI value was noticed when irrigation scheduled at 0.8 IW/CPE (I3) and 1.0 IW/CPE (I4) during and kharif, 2016 and rabi, 2016–17 respectively at all growth stages.

    3.3.  Growth parameters and grain yield of maize

    3.3.1.  Growth parameters

    In kharif, 2016 significantly higher DM accumulation and LAI was observed at silking (93.1 g plant-1, 2.993), in 04th July (D2) sown crop and significantly superior to 19th July (D3), 18th June (D1) and 10th August (D4) sown crops respectively. However, 19th July (D3) was comparable with 18th June (D1) (Table 3).


    Lowest DM accumulation and LAI at silking (79.9 g plant-1, 2.457) was observed in 10th August (D4) sown crop. In rabi, 2016-17 01st November (D1) sown crop showed maximum DM accumulation and LAI at silking (109.7 g plant-1, 2.769) and significantly superior to 17th November (D2), 01st December (D3) and 18th December sown crops respectively while, 17th November (D2) dry matter production was comparable with 01st December (D3) sown crop. The lowest DM accumulation and LAI was observed at silking (130.4 g plant-1, 2.170) with 18th December (D4) sown crops respectively. Crop sown on 10th August (D4) and 18th December (D4) recorded low DM production by 33% and 19% over 04th July (D2) and 01st November (D1) sown crops in kharif, 2016 and rabi, 2016–17 respectively. The reduction in maximum LAI in 10th August and 18th December, at silking stage was by 21% and 23% over 04th July and 01st November sown crop in kharif, 2016 and rabi, 2016–17 respectively.

    The observed difference in DM accumulation during development may result from differences in climate, differential absorption of PAR due to variation in plant population and LAI, or differences in the efficiency of converting absorbed photosynthetic active radiation (APAR) into DM (Tollenaar and Agui-lera, 1992). Higher temperature during vegetative growth stage hastened growth rate in early sown crop (Van Dobben, 1962), whereas, in delayed sowings, minimum temperatures reduced growth rate resulting in lower DM production. Early sowing date of in spring and fall resulted in high biomass because of greater solar radiation interception and thus resulted in higher RUE. Because, total biomass is depended on light interception and RUE (Abbas et al., 2020). Warmer temperatures speed up crop development rate, consequential shorter phases and less affective grain filling phase resulted in lesser partitioning of biomass toward grains in rabi (Rani et al., 2013).

    Relatively higher DM accumulation was recorded with I3 (0.8 IW/CPE) and I4 (1.0 IW/CPE) at silking (99.4 and 111.4 g plant-1) stage in kharif, 2016 and rabi, 2016–17 respectively. Which was however, significantly superior to I2 (0.6 IW/CPE), I1 (0.4 IW/CPE) and I0 (rainfed) during kharif, 2016 and I3 (0.8 IW/CPE), I2 (0.6 IW/CPE) and I1 (0.4 IW/CPE) during rabi, 2016–17 (Table 3). DM production during kharif, I2 (0.6 IW/CPE) was comparable with I1 (0.4 IW/CPE). LAI reached to a maximum value (3.149) at silking in the 0.8 IW/CPE (I3) which was significantly superior to I2 (0.6 IW/CPE), I1 (0.4 IW/CPE) and rainfed (I0) in kharif, 2016. However, I2 (0.6 IW/CPE) was comparable with I1 (0.4 IW/CPE) during kharif, 2016. Higher LAI (2.969) was observed with I4 (1.0 IW/CPE) and significantly superior over I3 (0.8 IW/CPE), I2 (0.6 IW/CPE) and I1 (0.4 IW/CPE) during rabi, 2016–17. The control rainfed plot in kharif, 2016 and 0.4 IW/CPE in rabi, 2016–17 recorded the significantly lower LAI (2.451 and 1.921) over other treatments at silking. The rate of increase in LAI due to increase of number of irrigations (0 to 3 and 3 to 8) at silking varied from 22 to 30% and 28 to 35% during kharif, 2016 and rabi, 2016–17 respectively.

    The dry matter accumulation in maize is associated with plant height, number of leaves and leaf growth and its persistence during vegetative phase and a combination of cobs, grains cob-1 and grain weight with concurrent shifts in leaf and stem mass during reproductive period (Payero et al., 2009). Higher leaf number, expansion and duration of leaf area coupled with higher vegetative growth period caused 28–32% loss of final dry matter weight (Cakir, 2004).

    3.3.2.  Grain yield

    During kharif, 2016–17 grain yield of maize was highest (6963 kg ha-1) in 04th July (D2) sown crop, which was significantly superior to 19th July (D3), 18th June (D1) and 10th August (D4) sown crops, however, 19th July (D3) was on par with 18th June (D1). Lowest grain yield (3011 kg ha-1) was recorded with 10th August (D4). In rabi, 01st November (D1) sown crop recorded the highest (5396 kg ha-1) grain yield, which was significantly superior over 17th November (D2), 01st December (D3) and 18 December (D4) however, 17th November (D2) sown crop, in turn comparable with 01st December (D3) sown crop. Significantly the lowest grain yield was observed in 18th December (D4) sown crop (4047 kg ha-1) across all other sowing dates. Sowing on 10th August (D4) and 18th December (D4) reduced the grain yield by 57% and 25% over 04th July (D2) and 01st November (D1) in kharif, 2016 and rabi, 2016–17 respectively. Late sowing affected grain yield by decreasing kernel number per unit area (Table 3). Yield increased with optimum planting date probably due to improved physiological conditions during the silking period for optimum kernel set (Barbieri et al., 2000). Late sowing of maize manifested significant reduction in grain yield. Low growth rate in the late sown crop could be mainly due to unfavourable environmental effects encountered during the reproductive phase and due to the low net assimilation rate (Rani et al., 2016). Optimum planting date for maize was 15th July in the Peshawar region. Variation in biomass yield across sowing dates was associated more with differences in the amount of radiation intercepted than differences in radiation use efficiency (Ahmad et al., 2001).

    During kharif, 2016 grain yield of maize was highest (6369 kg ha-1) in 0.8 IW/CPE (I3), which was significantly superior to I2 (0.6 IW/CPE), I1 (0.4 IW/CPE) and I0 (rainfed). However, I2 (0.6 IW/CPE) was comparable with I1 (0.4 IW/CPE) and lowest grain yield (4108 kg ha-1) was recorded with control (I0). During rabi, 2016-17 significantly highest grain yield (7105 kg ha-1) was obtained with I4 (1.0 IW/CPE) over I3 (0.8 IW/CPE), I2 (0.6 IW/CPE) and I1 (0.4 IW/CPE). The lowest grain yield was obtained (2464 kg ha-1) with irrigation scheduling of 0.4 IW/CPE (I1). The increase in grain yield with irrigation scheduling of 0.8 IW/CPE (I3) over control (I0) and 1.0 IW/CPE (I4) over 0.4 IW/CPE (I1) was 36% and 65%, respectively in kharif, 2016 and rabi, 2016-17 respectively.Results of the study showed that the yield and yield components were limited by soil moisture content and low temperature stress in the delayed sowing date. This reduction in water use is accompanied by decrease in plant water content affecting the crop growth and development of the plants resulting in reduced crop yields (Kumar et al., 2013).

    3.4. Relationship between spectral vegetation indices and growth parameters

    Correlation coefficients between spectral vegetation indices (SR, PRI, hNDVI, NDWI1240, NDWI1640, NDWI2130, NDII, NMDI, MSI, WBI, SRWI) and biophysical and growth parameters of maize crop at different phenophases during kharif, 2016 and rabi, 2016−17 are presented in Table 4.


    The results indicate that at knee high stage PRI was positively associated with drymatter (0.56). SR (-0.37), NDWI1240 (-0.41), NDWI1640 (-0.37), NDII (-0.33) and NDMI (-0.44) and SRWI (-0.40) were significantly and negatively associated with LAI except with PRI (0.53) which is positively correlated. At silking stage, superior and positive correlation was observed with drymatter and SR (0.72), PRI (0.54), hNDVI (0.51), NDWI1640 (0.41), NMDI (0.43) and WBI (0.38). LAI was positively associated with SR (0.50), PRI (0.37), NDWI1240 (0.57), NDWI2130 (0.67), NDII (0.71) and WBI (0.65) while negatively correlated with MSI (-0.68). At dough stage, all the spectral indices were positively correlated with drymatter and LAI except PRI and MSI which are negatively associated. At physiological maturity, drymatter was positively correlated with SR (0.62), NDWI1640 (0.55), NDWI2130 (0.74), NMDI (0.36), SRWI (0.66) while, PRI (-0.37) and MSI (-0.58) were negatively correlated. LAI was positively correlated with SR (0.47), NDWI2130 (0.74), NMDI (0.50), WBI (0.36) and SRWI (0.50) while, negatively associated with MSI (-0.53).

    However, the relationship between NDVI and LAI for pooled data over the two years, the overall best fit was exponential with a coefficient of determination (R2 =0.85) (Jayasree et al., 2013). An exponential relationship between LAI and NDVI was observed for soybean, beans, peas, corn and wheat (Haboudane et al., 2004). hyperspectral indices hNDVI and PRI for better ABM prediction. PRI is the best index for chlorophyll content prediction by an exponential model in wheat crop (Zhao et al., 2007). All the hyperspectral indices based on SWIR reflectances which are identified as the best estimator of crop water stress include 1640 nm in their formulae. Similarly, the exponential relation of broadband NDVI with LAI (Vina et al., 2011) with RMSE 1.18 was improved to an exponential 44 relationship when hNDVI was related with LAI, which showed much lower RMSE of 0.30. Linear relationship between SR and LAI (R2=0.87), exponential relationship NDVI and PRI with LAI (R2=0.91 and 0.87, respectively) of wheat (Aparicio et al., 2000). This study also supported the similar type of relationship when hNDVI was used as an estimator for LAI.

    The results indicate that grain yield were significantly and positively associated with SR, hNDVI, NDWI1240, NDWI1640, NDWI2130, NDII, NMDI, WBI and SRWI at knee high, silking, dough and physiological maturity. Yield was negatively and significantly correlated with MSI and PRI at all phenophases. Superior correlation was observed at dough and silking stages of maize crop as compared to knee high and physiological maturity stage. Highest values at the booting stage may be due to the highest amount of green leaf area. Decreasing trend from booting to maturity can be attributed to reduced reflectance in the NIR region and increased reflectance in visible region. This is due to damage of green tissue from booting to harvesting stage of crop growth in wheat (Prasad et al., 2007; Pradhan et al., 2012; Ramachandiran et al., 2016).

    3.5.  Regression analysis

    Stepwise regression analysis was performed to find out the critical spectral indices responsible for higher maize drymatter, grain yield at various phenophases. The regression models obtained were presented in the Table 5.


    3.5.1.  Regression models for prediction of drymatter

    Model I: Drymatter at silking

    The SR prevailed during P3 stage (knee high to silking stage) of the crop growth, accounted for 52% variation in DMP at silking stage.

    Model  II: Drymatter at dough

    During the growing season, DMP at dough was influenced by SR at silking, SRWI and WBI at dough during crop growth. These spectral indices accounted for 57% of variation in DMP at dough stage.

    Model III: Drymatter at physiological maturity

    The PRI prevailed during emergence to knee high stage, NDWI2130 prevailed during P3 stage (knee high to silking stage) of the crop growth, NDWI2130, SRWI and hNDVI during physiological maturity accounted for 84% variation in DMP at physiological maturity.

    3.5.2.  Regression models for prediction of leaf area index

    Model IV: Leaf area index at silking

    The leaf area index (LAI) at P3 (Silking) was influenced by NDII, MSI and NDWI1640 at silking and which accounted for 67% of variation in leaf area index in maize.

    Model V: Leaf area index at dough

    During the growing season, leaf area index at dough stage was influenced by NDWI2130 at silking and NMDI at dough stage. These spectral indices prevailed at P3 and P4 stage (Silking and dough stage) accounted for 60% of variation in LAI at dough stage.

    Model VI: Leaf area index at physiological maturity

    Hyperspectral normalized difference vegetation index prevailed at dough and physiological maturity, photochemical reflectance index at dough and normalized difference water index at 2130 nm at PM influenced the LAI at physiological maturity stage and which accounted for 77% of total variation in LAI at physiological maturity.

    In rainfed conditions, integration of either NDVI, SR, or PRI from heading to maturity explained 52, 59, and 39% of the variability in yield within genotypes in rainfed conditions and 39, 28, and 26% under irrigation. The usefulness of the SR and NDVI for calculating green area and grain yield was limited to LAI values <3 (Aparicio et al., 2000; Babar et al., 2006).

    3.5.3.  Regression models for prediction of grain yield

    Model VII: Grain yield

    During the growing season, grain yield was influenced by SR, WBI and NDWI1240 at silking, NDWI1240 and SRWI at dough, NDWI1640 and hNDVI at physiological maturity which accounted for 95% variation in grain yield of maize crop. Prediction of crop yield improved by individually combining NDVI and CV (Arnall et al., 2006). NDVI obtained from midseason spectral reflectance data of crop might prove a better indicator and could be used to predict yield (Raun et al., 2001). The relationship of NDVI with grain and biomass yields was stronger at the heading stage (r=0.98 and R2=0.93) in irrigated wheat. Similar relationship of grain and biomass yields was also observed for NDWI (Chandel et al., 2019).


  • Conclusion

    To assess the combined effect of dates of sowing and water stress in maize, hyperspectral reflectance value were transformed into spectral vegetation indices. All the spectral vegetation indices viz., SR, PRI, hNDVI, NDWI1240, NDWI1640, NDWI2130, NDII, NMDI, MSI, WBI and SRWI were positively correlated with LAI, dry matter and grain yield. This result indicates the ability of spectral vegetation indices to quantify effect of water stress on maize through remote sensing.


  • Acknowledgement

    We thank Dr. M.V.R. Seshasai, Deputy Director, Earth & Climate Sciences Area (ECSA), Dr. C.S.K. Murthy, Group Director (AS&AG), National Remote Sensing Centre, Balanagar, Hyderabad for their inspiring guidance and encouragement. Authors are grateful to University for financial assistance to carry out the experiment and analysis of the related samples.


  • Reference
  • Abbas, G., Fatima, F., Hussain, M., Hussain, S., Atique-ur-Rehman, Sarwar, N., Ahmed, M., Ahmad, S. 2020. Nitrogen rate and hybrid selection matters productivity of maize–maize cropping system under irrigated arid environment of southern Punjab, Pakistan. International Journal of Plant Production 14, 309–320.

    Ahmad, N., Waheed, A., Hamid, F.S., 2001. Different sowing dates of maize cultivars and their performance. Pakistan Journal of Biological Sciences 1, 106–108.

    Ali, M.M., Bachik, N.A., Muhadi, N.A., Yusof, T.N.T., Gomes, C. 2019. Non-destructive techniques of detecting plant diseases: A review. Physiological and Molecular Plant Pathology 108, 101426–101437.

    Anonymous, 2020. FAOSTAT Crop Statistics 2019. FAO of UN, Available from http://www.fao.org/faostat/en/#data/QC. Accessed on 22nd  December, 2020.

    Aparicio, N., Villegas, D., Casadesu, J., Araus, J.L., Royo, C. 2000. Spectral vegetation indices as non-destructive tools for determining durum wheat yield. Agronomy Journal 92, 83–91.

    Arnall, D.B., Raun, W.R., Solie, J.B., Stone, M.L., Johnson, G.V., Girma, K., Freeman, K.W., Teal, R.K., Martin, K.L., 2006. Relationship between coefficient of variation measured by spectral reflectance and plant density at early growth stages in winter wheat. Journal of Plant Nutrition 29, 1983–1997.

    Babar, M.A., van Ginkel, M., Klatt, A.R., 2006. The potential of using spectral reflectance indices to estimate yield in wheat grown under reduced irrigation. Euphytica 150, 155–172.

    Barbieri, P.A., Sainz Rozas, H.R., Andrade, F.H., Echeverria, H.E., 2000. Row spacing effects at different levels of nitrogen availability in maize. Agronomy Journal 92, 283–288.

    Bayat, B., Tol, C.V., Verhoef, W., 2016. Remote sensing of grass response to drought stress using spectroscopic techniques and canopy reflectance model inversion. Remote Sensing 8(7), 557.  

    Cakir, R., 2004. Effect of water stress at different development stages on vegetative and reproductive growth of corn. Field Crop Research 89(1), 1–16.  

    Ceccato, P., Gobron, N., Flasse, S., Pinty, B., Tarantola, S., 2002. Designing a spectral index to estimate vegetation water content from remote sensing data: Part 2. Validation and applications. Remote Sensing and Environment 82(2-3), 198–207.  

    Chai, Q., Gan, Y., Zhao, C., Xu, H.L., Waskom, R.M., Niu, Y., Siddique, K.H., 2016. Regulated deficit irrigation for crop production under drought stress. A review. Agronomy for Sustainable Development. 36, 3. https://doi.org/10.1007/s13593-015-0338-6.

    Chandel, N.S., Tiwari, P.S., Singh, K.P., Jat, D., Gaikwad, B.B., Tripathi, H., Golhani, K., 2019. Yield prediction in wheat (Triticum aestivum L.) using spectral reflectance indices. Current Science 116(2), 272–278.

    Draper, N.R., Smith, H., 1996. Applied regression analysis John, Wiley and sons, New York, 29.

    Gao, B.C., 1996. NDWI-A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sensing and Environment 58, 257–266.

    Genc, L., Inalpulat, M., Kizil, U., Mirik, M., Smith, S.E., Mendes, M., 2013. Determination of water stress with spectral reflectance on sweet corn (Zea mays L.) using classification tree (CT) analysis. Zemdirbyste-Agriculture 100(1), 81–90.

    Ghamisi, P., Plaza, J., Chen, Y., Li, J., Plaza, A., 2017. Advanced supervised spectral classifiers for hyperspectral images a review. IEEE Geoscience and Remote Sensing Magazine 5, 8–32. 

    Haboudane, D., Miller, J.R., Pattey., E., Zarco-Tejada, P.J., Strachan, I.B., 2004. Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies. Modeling and validation in the context of precision agriculture. Remote Sensing and Environment 90, 337–352.

    Hatfield, J.L., Pinter, J.P.J., 1993. Remote sensing for crop protection. Crop Protection 12(6), 403–413.

    Huete, A.R., 2004. Remote sensing for environmental monitoring. Environmental monitoring and characterization. Elsevier (USA). Chapter 11, 183–206.

    Jackson, T.J., Chen, D., Cosh, M., Li, F., Anderson, M., Walthall, C., Doriaswamy, P., Hunt, E.R., 2004. Vegetation water content mapping using Landsat data derived normalized difference water index for corn and soybeans. Remote Sensing and Environment 92, 475–482.

    Jacquemoud, S., Baret, F., 1990. Prospect: a model of leaf optical properties spectra. Remote Sensing and Environment 34, 75–91.

    Jayasree, G., Lingaiah, D., Raji Reddy, D., Narasimha Rao, S.B.S., 2013. Relationship between biophysical parameters and normalized difference vegetation index in maize. Journal of Agrometeorology 15(2), 120–125.

    Kaushal, M., Wani, S.P. 2016. Rhizobacterial-plant interactions: strategies ensuring plant growth promotion under drought and salinity stress. Agriculture, Ecosystem and Environment 231, 68–78. Doi: 10.1016/j.agee.2016.06.031.

    Kumar, A., Jat, S.L., Kumar, R., Yadav, O.P., 2013. Directorate of maize research, Pusa Campus, New Delhi- 110 012, 123p.

    Kumar, B.R., Rao, P.V., Ramulu, V., Avilkumar, K., 2013. Drip irrigation schedule for castor based on pan evaporation. Journal of Research, ANGRAU 41(2), 149–152.

    Ma, Y., Dias, M.C., Freitas, H., 2020. Drought and salinity stress responses and microbe-induced tolerance in plants. Frontiers in Plant Science 11, 591911. Doi: 10.3389/fpls.2020.591911.

    Mirik, M., Ansley, R.J., Michels Jr., G.J., Elliott, N.C., 2012. Spectral vegetation indices selected for quantifying Russian wheat aphid (Diuraphis noxia) feeding damage in wheat (Triticum aestivum L.). Precision Agriculture 13, 501–516.

    Murdia, L.K., Wadhwani, R., Wadhawan, N., Bajpai, P., Shekhawat, S., 2016. Maize utilization in India: An  overview. American Journal of  Food  and  Nutrition 4(6), 169–176. 

    Nilsson, H.E., 1995. Remote sensing and image analysis in plant pathology. Canadian Journal of Plant Pathology 17, 154–166.

    Padilla, F.L.M., Gonzalez-Dugo, M.P., Gavilan, P., Dominguez, J., 2011. Integration of vegetation indices into a water balance model to estimate evapo-transpiration of wheat and corn. Hydrology and Earth System Sciences 15, 1213–1225.

    Panigada, C., Busetto, L., Meroni, M., Amaducci, S., Rossini, M., Cogliati, S., Boschetti, M., Picchi, V., Marchesi, A., Pinto, F., Rascher, U., Colombo, R., 2010. EDOCROS: Early detection of crop water and nutritional stress by remotely sensed indicators. In: 4th International workshop on remote sensing of vegetation fluorescence, 15-17 Nov 2010, Valencia (SPAIN), 1–7.

    Parihar, C.M., Jat, S.L., Singh, A.K., Sai Kumar, R., Hooda, K.S., Chikkappa, G.K., Singh, D.K., 2011. Maize production technologies in India. DMR technical bulletin. Directorate of maize research, Pusa Campus, New Delhi-110 012. 30p.

    Parkash, V., Singh, S., 2020. A review on potential plant-based water stress indicators for vegetable. Crops Sustainability 12, 3945. doi:10.3390/su12103945.

    Payero, J.O., Tarkalson, D.D., Irmak, S., Davision, D., Petersen, J.L., 2009. Effect of timing  of a deficit-irrigation  allocation  on  corn  evapotranspiration,  yield,  water use efficiency and dry mass. Agricultural Water Management 96, 1387–1397.

    Pradhan, S., Bandyopadhyaya, K.K., Joshi, D.K., 2012. Canopy reflectance spectra of wheat as related to crop yield, grain protein under different management practices. Journal of Agrometeorology 14(1), 21–25.

    Prasad, B., Carver, B., Stone, M.L., Babar, M.A., Raun, W.R., Klatt, A.R., 2007. Potential use of spectral reflectance indices as a selection tool for grain yield in winter wheat under great plains conditions. Crop Science 47, 1426–1440.

    Ramachandiran, K., Pazhanivelan, S., Mohamed Amanullah, M., Geethalakshmi, V., Sivasamy, R., Asoka Raja, N., 2016. Quantifying the nitrogen and water stress of maize using spectral vegetation indices. International Journal of Bio-resource and Stress Management 7(1), 092–098.

    Rani, P.L., Sreenivas, G., Raji Reddy, D., 2016. Calibration and validation of CERES-Maize model for hybrid maize under variable plant densities and nitrogen levels in southern Telangana zone of Telangana State, India. International Journal of Bio-resource and Stress Management 7(2), 212–217.

    Rani, P.L., Sreenivas, G., Raji Reddy, D., Praveen Rao, V., Surekha, K., Siva Sankar, A., 2013. Influence of dates of sowing and nitrogen levels on growth and yield of kharif maize under irrigated conditions in south telanagana agro-climatic zone of Andhra Pradesh, India. International Journal of Bio-resource and Stress Management 4(1), 034–042.

    Ranjan, R., Sahoo, R.N., Chopra, U.K., Pramanik, M., Singh, A.,  Pradhan, S., 2015.  Assessment of water status in wheat (Triticum aestivum L.) using ground based hyperspectral reflectance. Proceedings of the National Academy of Sciences, India - Section B: Biological Sciences 87(2), 1–9.

    Raun, W.R., Johnson, G.V., Stone, M.L., Solie, J.B., Lukina, E.V., Thomason, W.E.,  Schepers, J.S., 2001. In-season prediction of potential grain yield in winter wheat using canopy reflectance. Agronomy Journal 93, 131–138.

    Singh, N., Rajendran, R.A., Shekhar, M., Jat, S.L., Kumar, R., Sai Kumar, R., 2012. Rabi maize opportunities challenges, directorate of maize research, Pusa Campus, New Delhi -110 012, Technical Bulletin, No. 9, 32p.

    Snedecor, G.W., Cochron, W.G., 1967. Statistical methods. Oxford and IBH publishing company 17, Parklane, Calcutta, 172–177.

    Szeles, A.V., Megyes, A., Nagy, J., 2012. Irrigation and nitrogen effects on the leaf chlorophyll content and grain yield of maize in different crop years. Agricultural Water Management 107, 133–344.

    Tollenaar, M., Aguilera, A., 1992. Radiation use efficiency of an old and a new maize hybrid. Agronomy Journal 84, 536–541.

    Van Dobben, W.H., 1962. Influence of temperature and light conditions on dry matter distribution, development rate and yield in arable crops. Netherlands Journal of Agricultural Science 10, 377–389.

    Verma, D., Pandey, V., Karande, B.I., Lunagaria, M.M., Patel, K.I., 2010. Spectral characteristics of wheat as influenced by nitrogen stress. Journal of Agrometeorology 12, 89–93.

    Verma, N.K., Pandey, B.K., Singh, U.P., Lodhi, M.D., 2012. Effect of sowing dates in relation to integrated nitrogen management on growth, yield and quality of rabi maize (Zea mays L.). Journal of Animal and Plant Sciences 22(2), 324–329.

    Vina, A., Gitelson, A.A., Nguy-Robertson, A.L., Peng, Y., 2011. Comparison of different vegetation indices for the remote assessment of green leaf area index of crops. Remote Sensing and Environment 115, 3468–3478.

    Wu, C.Y., Niu, Z., Tang, Q., Huang, W.J., 2008. Estimating chlorophyll content from hyperspectral vegetation indices: Modeling and validation. Agricultural and Forest Meteorology 148, 1230–1241.

    Zarco-Tejada, P.J., Rueda, C.A., Ustin, S.L., 2003. Water content estimation in vegetation with MODIS reflectance data and model inversion methods. Remote Sensing and Environment85, 109–124.

    Zhao, L., Shi, Y., Liu, B., Hovis, C., Duan, Y., Shi, Z., 2007. Finer classification of crops by fusing uav images and sentinel-2A data. Remote Sensing 11, 3012.

    Zhao, L., Shi, Y., Liu, B., Hovis, C., Duan, Y., Shi, Z., 2007. Finer classification of crops by fusing UAV images and sentinel-2A data. Remote Sensing 11, 3012.


People also read

Short Research

Preference Towards Online Mode of Distance Education Courses–conjoint Analysis

M. Malarkodi, V. M. Indumathi and S. Praveena

Conjoint analysis, distance education, preference, online learning

Published Online : 07 Feb 2018