Four sub-populations with special reference to silking and maturation periods were derived from a broad based CIMMYT population- 31 received from Maxico in the National Coordinated Maize Programme, NARC, Islamabad. The derived populations were LL late silking and late maturity, LE late silking early maturity, EL early silking late maturity and EE early silking early maturity. The crop duration was partitioned into silking and maturity phases to identify short duration maize cultivars with high potential which could be well fitted in various cropping patterns. Maturity was based on black layer formation (BLF). The base population was used for crossing between plants that silked after 47 days of planting and named as early silking. Crosses were also made between plants that silked after 55 days of planting and named them as late silking plants. On the bases of BLF both early and late silking plants were grouped as early and late BLF plants during the following year. Selected groups were crossed to raise 4 sub-populations and finally 40 ears of EE, 20 ears of EL, 16 ears of LL and 39 ears of LE groups were selected and planted during spring 1995. Plants were crossed in all possible combinations within each population and same cycle of selection and hybridization was repeated during Autumn 1995 by selecting 65 ears for EE 40 ears for EL, 30 ears for both LE and LL on the basis of BLF. Crosses were made between the plants silked on the same date in each group and 10 ears from each group were selected for head to progeny rows during spring 1996 and bulk pollination was made among the plants in each group and these populations along with base population were used as experimental material for investigation of variance, genetic gain, heritability, correlation, path coefficient and stability analysis.
The experiment was conducted at three locations; at location I, both agronomic and physiological traits were recorded, whereas at location 11 and location Ill, only agronomic traits were recorded. Significant differences were observed for all the traits at all the three locations except number of ears and oil contents at location I where populations did not differ significantly. The population LL exhibited the highest kernel yield, GFO, EGFO, kernel rows, kernels per plant, leaf area and moisture at harvest. It displayed more kernels per row, plant height, whereas moderate ear height and lowest OMAR. Second population LE gave high kernel yield, kernels per plant, plant height, ear height and highest OMAR and kernels per row, whereas it displayed moderate kernel weight, leaf area moisture at harvest and lowest GFO and EGFO. The population EL had low kernel yield per plant, kernels per row, kernels per plant, plant height, ear height and OMAR, whereas it .showed higher GFO, EGFO, kernel weight and moderate kernel rows, leaf area and moisture at harvest. Population EE depicted lowest kernel yield, kernel weight, leaf area, plant height and MAH, it possessed lower GFD, kernel rows, kernels per row, kernels per plant, ear height and EGFD, whereas it possessed higher DMAR.
High heritability alongwith high genetic advance was recorded for days to silking, black layer formation, GFD, kernel rows, kernels per row, kernels per plant, kernel weight, leaf area, plant height, ear height, kernel yield, EGFD, DMAR, moisture loss rate, moisture loss duration and for moisture at harvest. This indicated the scope of improvement for these characters from broad based population by crossing and simultaneous selection. Moderate heritability and low genetic advance was recorded for LPD and low heritability alongwith low genetic advance were recorded for ears per plant, protein content and oil content. Improvement for these characters from the present material was limited, therefore some novel technique (mutation breeding, wide hybridization & transformation).are suggested to be adopted to improve these important traits.
Interrelationship among various quantitative traits alongwith direct and indirect influence of yield components is an important study for the breeders. Such information may be used in the prediction of the correlated responses to directional selection and in the detection of some traits which may have no value in themselves but are useful as indicators of the more important ones under consideration. Path analysis revealed that kernels per plant showed the highest direct effect on kernel yield followed by DMAR, plant height and EGFD at location-I. Days to black layer formation had highest positive direct effect on kernel yield followed by ear height and kernels per row at location III, whereas at location-II only leaf area expressed high positive direct effect on kernel yield. All these characters are multigenic in nature and influenced by the environment, therefore correlation and path analyses are suggested to be conducted under specific environments for particular material.
Significant correlation between the traits in different populations at all locations revealed positive association of days to silking with maturity period in EE and with kernel rows in EE & EL. Days to black layer formation showed positive association with kernel rows in EE & EL and with GFD in BP, EL & LL. It was observed that both developmental stages had no correlation with kernel yield directly in any population, only significant correlation of days to black layer formation with kernel yield in LE & BP at one location showed high influence of environmental factors. Kernel rows had positive association with kernels per plant in BP, EE, LE & LL and with kernel yield in LE. Kernels per row had positive correlation with kernels per plant in BP & EL and with kernel yield in BP, kernels per plant with kernel yield in all populations and with leaf area in BP, whereas kernel weight displayed negative association with kernels per plant in all populations except BP. Kernel yield could be improved by increasing kernels per plant in all populations, whereas negative association of kernel yield with kernel weight in all the populations needs special breeding techniques to break this undesirable linkage.
Stability analysis indicated that location and population x location interaction were significant for days to black layer formation, kernel weight, leaf area, plant height and ear height. Population-location interaction was significant for days to silking and kernel yield, whereas non-significant variance of location and population-location interaction was observed for kernels per plant. None of the populations possessed bi equal to unity, whereas populations with regression closer to unity were considered stable population. The LE population showed bi value closer to unity for days to silking, days to black layer formation, kernel yield and plant height. The EE population showed bi value closer to unity for days to black layer formation kernel weight, ear height and kernel yield, BP showed for kernels per plant and plant height, and LL for kernel weight and plant height. Population showed bi value greater than unity were LL for days to black layer formation, and kernel yield, LE for kernels per plant, and ear height, EL for days to black layer formation, leaf area, plant height and kernel yield, EE for kernels per plant, leaf area, and ear height, and BP for days to silking, kernel weight, leaf area, and ear height. Populations that displaying negative bi values, were LL for days to silking, leaf area & ear height, LE for leaf area, EL for days to silking, EE for days to silking, and BP for days to black layer formation & kernel yield. Significantly higher S2d were observed for days to silking in EL & BP for kernel weight in LL, EL and EE, for leaf area in LL, EL and BP, for plant height in EL & BP, for ear height in LL, LE, EL & BP and for kernel yield in LE, EL & BP.