Photothermal Quotient-An Effective Thermal Index for Evaluation of Lentil (Lens culinaris L.) Phenology and Yield in the Lower Gangetic Plains

Any change in ambient temperature range during the crop growth period alters the phenological development, thus affecting the dry matter accumulation and productivity. Thermal indices such as growing degree days (GDD) and heat use efficiency (HUE) have been widely used to quantify the effects of temperature on phenology and yield performance in several crops (Pandey et al., 2010). On the contrary, photothermal quotient (PTQ) summarizes a larger volume of weather information by providing the dual effects of temperature and solar radiation on crop growth (Kumar et al., 2016). Since global solar radiation influences air temperature, inclusion of solar radiation in thermal indices is crucial for comprehensively evaluating the temperature effects on crop phenological developmental rate and productivity. The nature of heat and solar energy utilization also depend on the crop genetics and growing environment. A detailed understanding of the association between these weather parameters and yield of a particular crop helps in breeding climate-smart varieties. 
       
Lentil (Lens culinaris L.) ranks second in India among food legumes, covering an area of 1.51 m ha with 1.56 m t and 1.03 t ha^-1 of production and productivity, respectively (DES, 2020). It is grown as a rainfed winter crop in northern and eastern India. However, the shifting monsoons and erratic weather events have made winter agriculture prone to severe yield losses, especially in eastern India. Temperature abruptions associated with late sowing, especially during the reproductive stage, severely affect the floral and pod biologies, overall heat and solar energy utilization, yield components and productivity in winter food legumes (Maji et al., 2019). The genetic makeup of variety also influences the temperature sensitivity of vegetative and reproductive growth traits determining yield (Chaudhury et al., 2020). Therefore, the lentil genotypes with consistent yield performance across diverse weather conditions are likely to fit better in eastern India’s short and erratic winters. Hence, this study was taken up with two objectives: (i) to identify the optimal sowing window and lentil varieties suitable for the lower Gangetic plains and (ii) to evaluate the effects of GDD and PTQ on lentil phenology and yield.

A three-year field experiment was conducted during the winter seasons of 2015-16, 2016-17 and 2017-18 at the District Seed Farm (AB block), Bidhan Chandra Krishi Viswavidyalaya (latitude-22°58'45"N, longitude-88°25'15"E and altitude-9.75 m above sea level). The experimental soil was sandy clay loam in texture (sand-47.7%, clay-24.3% and silt-27.9%) with a neutral pH (7.1) and had medium organic carbon (0.52%), available nitrogen (258.2 kg N ha^-1) and available potassium (153.5 kg K₂O ha^-1) but high available phosphorus (29.4 kg P₂O₅ ha^-1) status. The mean maximum temperatures during the experimental period were 29.2, 28.6 and 28.3°C, respectively in the three consecutive experimental years. The mean minimum temperatures in the same years were 16.7, 14.9 and 15.0°C, respectively. Total rainfall received during the crop growing period in 2015-16, 2016-17 and 2017-18 were 57.9, 22.8 and 52.5 mm, respectively. Year-wise weekly distribution of the mean atmospheric temperatures and the total rainfall during the experimental period are presented in Fig 1.
 

       
The experiment comprised four sowing dates in the main plots (D₁-31^st October, D₂-7^th November, D₃-14^th November and D₄-21^st November) and five varieties (V₁-HUL 57, V₂- WBL 77, V₃ -KLS 218, V₄ -WBL 58 and V₅ -B 256) in the sub-plots, laid out in a split-plot design replicated thrice. Different varieties were sown after seed inoculation with recommended Rhizobium strain. A 25 cm (row-row) × 10 cm (plant-plant) spacing in a sub-plot area of 5 m × 4 m was maintained. A basal dose of 20 kg N, 40 kg P₂O₅ and 40 kg K₂O ha^-1 were applied. Single pre-sowing irrigation was given to ensure uniform germination and crop stand.
       
The dates of emergence (E), 50% flowering (50%F), 100% flowering (100% F) and Maturity (M) in each lentil variety under all the sowing dates were recorded. Pods per plant and hundred seed weight (HSW) were observed from randomly tagged ten plants from each treatment combination. Stover (STY) and seed yields (SY) were measured from a net area of 12.0 m2 and expressed in t ha^-1. The harvest index (HI) is the ratio of SY and biological yield. The phenophase-wise growing degree days (GDD) considering a base temperature of 5.0°C and heat use efficiency (HUE- kg ha^-1 °C day^-1) were calculated as per Kaur et al., (2019). Since the daily solar radiation (SR) of the experimental location was not available, it was computed from the Bristow and Campbell (1984) algorithm using maximum temperature (MAXT), minimum temperature (MINT), Julian days and latitude of the site. The photothermal quotient (PTQ) was calculated using the formula proposed by Ortiz-Monasterio et al., (1994), with the base temperature of 5.0°C in lentil. Where;
AVGT = Daily average temperature (°C). {(MAXT + MINT)/2}.
SR= Daily solar radiation (MJ m^-2).
PTQ = Expressed as MJ m^-2 day^-1 °C^-1.   
       
The yield losses due to deferred sowing dates were worked out using the regression equations proposed by Ortiz-Monasterio et al., (1994) :
Equation (1)- Absolute loss: Yi= a + bSDi + u 
Equation (2)- Relative loss:  ln (Yi)= a + bSDi
Where,
Yi (kg ha^-1) = Yield of the sowing date.
SDi= Julian date for the sowing date.
       
Equations 1 and 2 provide an estimation of the yield losses on absolute (b= kg ha^-1 day^-1 yield loss) and relative (100 × b = percent day^-1 yield loss) basis, respectively.
       
The data were subjected to analysis of variance (ANOVA) for split-plot design. Tukey’s HSD (honest significant difference) was used to compare the treatment means using SAS version 9.1 for windows. The interaction effect between sowing dates and varieties was non-significant (p>0.05) (data not shown).  Pearson’s correlation coefficients was used to quantify the association among the thermal indices, yield and yield components. Regression analysis was employed to quantify the impacts of GDD, PTQ and yield components on yield.

Yield and yield components
 
The 7^th November sown crop recorded the highest SY (1.41 t ha^-1), STY (2.85 t ha^-1), pods per plant (155.5 plant^-1) and harvest index (0.33) (Table 1). Delayed sowing beyond first week of November adversely affected yield traits and yield in lentil. Singh et al., (2009) also reported that deferring sowing from 20^th October to 10^th November reduced lentil yield by approximately 10.4%, implying that early sowing ensures adequate sink partitioning by providing favourable growing conditions for optimal vegetative and reproductive growth.
 

       
WBL 77 yielded the highest among the varieties with better yield attributing traits, closely followed by HUL 57. However, delaying sowing from 7^th November to 21^st November reduced the SY of HUL 57, WBL 77, KLS 218, WBL 58 and B 256 at the rates of 17.9, 13.6, 20.5, 22.5 and 22.1 kg ha^-1 day^-1 on absolute basis, respectively; the corresponding values for the same varieties were 1.4, 0.9, 1.7, 1.8, 2.1% ha^-1 day^-1 on relative basis.
 
Relationship between yield and yield components
 
Correlation analysis showed that pods plant^-1, 100 seed weight and stover yield significantly influenced lentil yield (Kumar et al., 2020) (Table 2). Stepwise multiple regression analysis showed that R2 values increased significantly with the addition of each yield component (according to the strength of degree of association) as predictor variable, indicating that yield is cumulative effect of the various yield components and no single trait emerged as a prime determinant (Table 3). The regression equation with 100 seed weight, stover yield and pods plant^-1 explained 97.0% of the total variation in lentil yield.
 

 

 
Crop phenology
 
The early sowings (31^st October and 7^th November) took 2-4 days less for emergence than the other two sowings (Table 4). The late sown winter crops usually require a longer time for seedling emergence due to low-temperature exposure decelerating the crop growth (Richards et al., 2020). Deferred sowing beyond 7^th November reduced the overall crop duration by 7-13 days. Forced maturity due to high temperature-induced senescence shortened the life cycle of late sown fababean (Prakash et al., 2018). The results also reveal that the 31^st October and 7^th November sowings with more days to flowering culminated in significantly higher yield, evidencing the paramount importance of adequate vegetative window. A shorter vegetative phase limits biomass accumulation resulting in yield losses because seed filling depends on current assimilation and redistribution from pre-stored vegetative pools (Sehgal et al., 2018).
 

       
The varieties HUL 57, WBL 77 and KLS 218 recorded significantly longer duration to achieve various phenological stages. Pooled analysis revealed that the crop duration of HUL 57, WBL 77 and KLS 218 ranged from 113 to 116 days.
 
Growing degree days and heat use efficiency
 
The 14^th and 21^st November sowings recorded significantly higher thermal time requirement for seedling emergence (Table 4). Relatively lower mean temperatures delayed the emergence of late sown lentils in the current study, increasing the heat unit consumption as reported by Richards et al., (2020). The trend of heat unit accumulation followed the pattern of variation in phenological development during the crop growth period. Sowing after 7^th November reduced the crop duration and heat unit consumption from sowing to maturity by 7-13 days and 7.0-12.8%, respectively compared to early sowings. Late sowing lowers heat unit consumption due to supra-optimal thermal regime during the reproductive phase, which accelerates senescence.
       
The varieties HUL 57, WBL 77 and KLS 218 recorded significantly higher heat unit consumption to attain various phenological stages and no significant difference was observed between them (Table 4). The variation emerging from the characteristic genetic makeup of crop varieties had a marked influence on phenology and heat unit consumption, eventually affecting yield (Pandey et al., 2010). Regression analysis revealed that GDD caused 82.0-89.0% of the total variation in the yield of lentil varieties (Fig 2a).
 

       
The 7th November sowing had the highest HUE (0.793 kg ha^-1 °C day^-1), which was possibly due to a proportional increase in dry matter per heat unit absorbed (Table 4). High-temperature exposure during the late vegetative and reproductive stages might have adversely affected dry matter accumulation, reducing the HUE in the late sowing. Among the varieties, WBL 77 recorded the highest HUE, indicating its wider adaptability under varying thermal environments. 
 
Photo-thermal quotient
 
The crop sown within the first week of November recorded significantly higher total PTQ during the crop growing period (Table 4). The varieties HUL 57, WBL 77 and KLS 218 recorded significantly higher PTQ, resulting in a higher yield. Per unit increase in PTQ increased the yield by 20.9, 16.0, 27.1, 28.0, 40.7 kg ha^-1 in HUL 57, WBL 77, KLS 218, WBL 58 and B 256, respectively. Yield increases with increasing PTQ since low PTQs indicated elevated mean temperatures and radiation intensity (Ortiz-Monasterio et al., 1994). Heat stress reduces lentil seed filling duration and yields by 5.5-8.1 days and 38.0-58.0%, respectively (Sita et al., 2018). The regression analysis showed that the PTQ caused 92.0-96.0% of the total variation in the yield (Fig 2b). Thus, PTQ was a better yield predictor than GDD since the former quantifies the combined effects of both radiation and temperature (Kumar et al., 2016).
 
Effect of temperature and PTQ (phenophase-wise) on yield components and yield
 
The temperature during E-50% F exhibited no significant effect on the stover yield of WBL 77 and HUL 257, implying lower sensitivity of these two varieties to temperature fluctuations during the vegetative stage (Table 5). Among the yield components, 100 seed weight and pods plant^-1 were more sensitive to maximum than minimum temperature during E-100%F. However, both maximum and minimum temperatures during 100%F-M showed strong negative correlations with these traits, irrespective of varieties. The maximum and minimum temperatures during the 100% F-M also had a comparable negative impact on the final yield. Since heat stress affects reproductive biology, biomass accumulation and leaf function-associated traits in lentils (Bhandari et al., 2020), late sowing-associated elevated temperatures are likely to reduce lentil productivity as observed herein.
 

       
The PTQs during E-100%F and 100% F-M had a stronger influence on yield and yield components than either maximum or minimum temperature (Table 5). These findings align with Ortiz-Monasterio et al., (1994), who reported that increased PTQ during 20 and 10 days before and after heading improved grains m^-2 in wheat.

The optimal sowing window in the Lower Gangetic Plains was the first week of November for intensifying lentil productivity through improved phenology and thermal energy use. The two promising varieties viz. HUL 57 and WBL 77 exhibited lower sensitivity towards temperature fluctuations during the vegetative stage with consistent performance across varying thermal regimes. Temperature significantly influenced yield components, with pods plant^-1 and 100 seed weight being more sensitive to maximum temperature during vegetative and early reproductive stages and both maximum and minimum temperatures during the late reproductive phase. However, the stronger correlation between photo-thermal quotient and yield components made it a more reliable yield predictor. Hence, photo-thermal quotient might serve as a more effective thermal index in identifying temperature-sensitive traits than the widely-used growing degree day for breeding climate-smart lentil varieties with wider adaptability.

None.