Determination of base temperature (Tb) is a pre-requisite for a further calculation of growing degree days (GDD). Tb was calculated based on the method of least standard deviation (in GDD)
(Yang et al., 1995). Daily mean air temperature (MAT) of the respective months till flower initiation as well as days from planting to flower initiation (FTDAP) of the respective genotypes were used to calculate base temperature (Tb). Base temperature calculated for all the genotypes varied from 9.53°C to 12.09°C (Table 3).
Combined analysis of variance
Combined analysis of variance was performed to detect genetic differences among genotypes for temperature induced sensitivity to flowering time (Table 1). The significant differences in planting dates for FTDAP, MAT, GDD and yield, indicates that the staggered date of plantings had significant variation in flowering days, air temperature, heat units accumulated for flowering initiation and also there is variation for yield at different planting dates. There is also significant differences of genotypes for FTDAP, MAT, GDD and yield, which means all the genotypes differed significantly for flowering days, air temperature, for heat units accumulation as well as for yield.
Flowering, GDD and base temperature (Tb)
The number of days required for the flower initiation varied across the season for individual genotypes. The flowering time from day of the planting (FTDAP) for DC 15 × DCS 47-1 varied from 48.50 days to 55 days, while for both C152 × CP206 and DC 15 × SHUBRA the FTDAP varied from 49 to 55 days. The germplasms, IC97856, IC257428, IC202926, IC198326 and IC259064 varied from 44 to 50, 44 to 48, 44 to 49, 43.50 to 48 and 44 to 48 days across all the season. Among the checks DC 15, DC 16 and DCS 47-1 registered 50 to 57, 47.5 to 57 and 48 to 56 days of variation across the season for flower initiation, respectively.
Among all the genotypes, the GDD required for flower initiation was varied for individual genotypes across the seasons. In F6 crosses, DC 15 × DCS 47-1 had varied from 897.2 to 595.1°C day across the year. Similarly, C152 × CP206 and DC 15 × SHUBRA varied from 928.3 to 589.3°C day and 902.9 to 557°C day across the year, respectively. All the germplasms, IC97856, IC257428, IC202926, IC198326 and IC259064 varied from 770.9 to 417.3°C day, 861.3 to 573.6°C day, 818.7 to 484.7°C day, 821 to 515.7°C day and 798.5 to 432.5°C day across the year, respectively. In the checks DC 15 registered 916.5 to 555.2°C day of GDD. Whereas DC16 and DCS 47-1 registered 890.2 to 550.3°C day and 889.1 to 544.5°C day throughout the year, respectively.
The overall mean of FTDAP and growing degree days (GDD) required by the cowpea genotypes for flower initiationdiffered for individual genotypes. The F6 crosses, DC 15 × DCS 47-1 require 50.71 days and 725.44°C day for flowering initiation with its base temperature being 10.39°C. Similarly, C152 × CP206 and DC 15 × SHUBRA require 51.08 and 51.50 days with 739.33 and 683.90°C day for flowering initiation with their base temperatures 10.17 and 11.38°C, respectively. The germplasms, IC97856, IC257428, IC202926, IC198326 and IC259064 require 46.75, 46.21, 46.63, 45.17 and 45.92 days with their GDD of 594.64, 699.09, 627.51, 660.57 and 582.64°C day for flowering initiation with their base temperatures of 11.87, 9.53, 11.17, 10.04 and 12.09°C, respectively. Whereas, DC 15, DC 16 and DCS 47-1, registered 52.21, 50.67 and 50.79 days with 720.82, 705.52 and 692.10°C day for flowering initiation with base temperatures being 10.82, 10.70 and 11.03°C, respectively (Table 3).
Growing degree days (GDD) varied across the seasons, the genotypes which shown least variation of GDD across the seasons may be considered as insensitive one. IC202926 is the genotype with least variance of 8402.17 (Fig 1f) for GDD across the seasons, followed by IC198326, IC97856, IC257428 and IC259064 with variance value of 8706.84, 10012.74, 12878.21 and 12878.21 (Fig 1g, d, e and h). Genotypes like DC15×DCS47-1, C152×CP206 and DC15×Shubra exhibited higher variance (Fig 1a, b and c).
The genotypes which differed for time to flowering (FTDAP) and GDD for flower initiation across the seasons with respect to temperature were regarded as sensitive genotypes. In contrary, the genotypes where there performances were comparable across the seasons for flowering were regarded as insensitive genotypes (Keerthi., 2015). In the present study, three genotypes (IC202926, IC198326 and IC257428) were recognized, where their flowering days were comparable, they shown least variation for flowering across the season (Fig 2e, f, g). And the other genotypes (DC15×DCS47-1, C152×CP206, DC15×SHUBRA, IC97856, IC259064) shown higher variation (Fig 2a, b, c, d and h).
On an average, sensitive genotypes required higher Tb and average GDD for flowering than the insensitive genotypes (
Keerthi., 2015). Among the three temperature insensitive genotypes recognized (IC257428, IC202926 and IC198326) had registered relatively lower Tb and average GDD of 9.53, 11.17 and 10.04°C of Tb and 573.6, 484.7 and 515.7°C day of GDD.
As expected temperature sensitive genotypes differed widely in flowering time as a function of temperature for flowering, whereas temperature insensitive genotypes maintained relatively constant number of flowering days irrespective of temperature, similar reports obtained by
Keerthi (2015).
Flowering and yield
All the genotypes varied significantly for flowering as well as for yield across all the seasons. Flowering occurred earlier in the warm seasons and it took some additional days to flower in cool seasons
(Hadley et al., 1983). As the flowering differed, yield also varied. Yield was high in some of the months and it was low in some of the off seasons, genotypes with early flowering and high yielding are desirable. Here the 11 genotypes varied significantly for flowering and yield across the seasons. The genotype DC15 and DC16 took 51 and 50 days for flowering and recorded high yield of 35 and 36.7 grams per plant in the month of July. Whereas DCS47-1 registered 48 days of flowering and high yield of 37.6 grams per plant in the month of July. All these genotypes shown more yield fluctuation at different dates of plantings. Among the crosses, DC15 × DCS47-1 has registered 51 days for flowering and high yield of 39.4 grams per plant in the month of July, it also registered high yield (40.1 grams per plant) in the month of January with 52 days for flowering. Similarly, C152 × CP206 and DC15 × SHUBRA both took 51 days for flowering and registered their highest yield of 39.7 and 44.1 grams per plant, respectively in the month of July. They also registered high yield of 39.6 and 54.5 grams per plant, respectively in the month of January. All these exhibited maximum fluctuations for yield as well as for flowering.
The genotypes IC97856 and IC259064 exhibited higher yield of 45 and 42.5 grams per plant in the month of July with 45 days for flowering. Their yield was also high in the month of January (42.8 and 42.5 grams per plant) with 48 and 49 days for flowering, with all their yield fluctuated throughout the year. Whereas IC257428, IC202926 and IC198326 registered least variation for flowering and yield across the seasons. These genotypes took 44, 48 and 44 days for flowering, with yield of 37.2, 41.4 and 34.7 grams per plant in the month of July. Among all the genotypes these three registered least variance for flowering (1.48, 2.10 and 1.61 respectively) as well as for yield (32.15, 36.50 and 23.40) across the seasons (Fig 2e, f, g).
Critical minimum GDD and temperature requirement
The graph of GDD on Y-axis and FTDOY on the X-axis indicated requirement of critical minimum GDD for time to flowering (Fig 1). The critical minimum GDD required for flowering initiation by genotypes ranged from 417.3°C day (IC97856) to 595.1°C day (DC15 × DCS47-1) with an average of 506.2°C day. The minimum temperature required for flowering initiation ranged from 19.5 to 19.9°C with average minimum temperature of 19.7°C, on the other hand the optimum temperature required for flower initiation ranged from 24.3 to 25.2°C with average optimum temperature of 24.75°C (Table 4).
In the present study, the identified germplasms IC257428, IC202926 and IC198326 attained critical minimum GDD of 573.6, 484.7 and 515.7°C day in the month of December. This indicate that cooler months lead to the minimum heat units accumulation. Similar results were obtained by
Keerthi (2015) where dolichos bean require 411°C day as critical minimum GDD for flowering. Similarly, in faba bean and pea critical minimum GDD of 833 and 770°C day are required for time to flowering
(Iannuci et al., 2008).
Correlation studies among different indices
Correlation studies had been carried to assess the genetic correlation between the parameters. The parameters (FTDAP, GDD, MAT and yield) were significantly correlated. The FTDAP is significantly and negatively correlated with GDD and MAT (-0.41** and -0.44** respectively) and positively correlated with yield (0.38**), which indicates that as temperature increases the days to flowering (FTDAP) decreases, flowering initiation occurs earlier in case of warm seasons
(Hadley et al., 1983) and late in case of cool seasons these additional cool days decreases the heat units accumulation (GDD), Whereas GDD is significantly and positively correlated with MAT (0.97**) and negatively correlated with yield (-0.50**), this indicates that as the GDD is increased by increase in temperature, but high temperature has deleterious effect on the yield (Table 2).
Prediction of time to flowering of cowpea genotypes
The timing of flowering and in particular the degree to which it is responsive to temperature as a key factor in the adaptation of species to different eco-geographical locations (
Weller and Ortega, 2015) and cowpea is not exception to this. Hence, accurate prediction of time to flowering helps in making decisions on appropriate sowing and harvesting time to optimise productivity. The linear regression model described by
Robberts and Summerfield (1987) are considered useful for predicting the rate of progress towards flowering. The regression constants (a and b) of all the genotypes helps us to predict the flowering time with respect to the average air temperature prevalent in that area, thereby helps us to plan the sowing and harvesting date. Regression of temperature on the rate of development in all the genotypes were significantly (P<0.01) correlated to mean air temperature, in this regard an attempt has been made to predict the flowering time in all genotypes using thermal model. The genotypes like IC202926, IC97856, IC257428 and DC 16 were found highly significant (P<0.01) with R2≥0.57 (Table 4).
