Evaluation of crop traits range of dryland chickpea (Cicer arietinum L.) in cold areas of Iran

Pulses play an important role in human nutrition. Among the pulses, chickpea (Cicer arietinum L.) is the valuable food in terms of carbohydrates and protein. Chickpea with more than 8 million tons per year ranks third crop in the world. It is planting in 48 countries with more than 12.11 million hectares. The aim of any breeding program working for unpredictable and rainfed environments is to develop varieties with high and stable yields. Breeders take advantage of the selection for several traits to achieve maximum economic yield. The selection of genotypes based on indices using yield components was used by breeders for a long time. Breeders were believed that obtaining a linear function of traits or selection index could lead the selection of genotypes with better genetic values but including economic weights in this function have been expressed by many researchers. Simultaneous selection using characteristics with important and heritable economic values is more effective. Crop yield is a function of multiple interrelated variables and cannot be defined only by a single-variable equation. One of the most effective method is boundary lines that was introduced by Feiziasl et al, (2003 and 2010) for the first time to determine the optimum levels of plant (dryland wheat) characteristics and yield stability analysis. In this paper, “Boundary Lines” and averaging methods and PCA are used to determine optimum levels for some traits of dryland chickpea in cold areas, Iran.

More than 8000 data for each trait were collected from national and international trials conducted in Dryland Agricultural Research Institute (DARI) experimental stations in Maragheh, Kurdestan, Zanjan, Uromieh and Ardabil from 1996- 2016. The traits were days to flowering, days to physiological maturity, grain filling period, plant height, 100 seed weight and grain yield. The Excel software was used to develop a scatter diagram showing therelationship between each trait with grain yield in each location. Two methods were used to determine the optimum value for a given trait. One is based on the boundary lines method where the maximum grain yield and the optimum value for the trait considered coincide with the crossing point of the two boundary lines. “Boundary Lines” method was used to determine the maximum limit for each trait. The scatter diagram is surrounded by two regression lines, one on the left and the other on the right called boundary lines. Then Maximum yield was obtained at the intersection of both boundary lines. The other approach, called averaging method, is based on subdividing the data into two groups: high and low yielding groups.

The boundary lines method determined the optimum levels for days to flowering, days to physiological maturity, grain filling duration, plant height and 100 seed weight which were 56.2 days for spring and 89.2 days for autumn, 89.4 days for spring and 120.8 days for autumn, 33.1 days, 29.8 centimeters and 34.0 grams, respectively. By averaging of high yielding group method, optimum values for days to flowering, days to physiological maturity, duration grain filling duration, plant height and 100 seed weight characteristics were 75.6 days, 108.8 days, 37.7 days, 30.2 centimeters, and 26.0 grams, respectively. The optimum values for plant height and 100 seed weight were almost the same in both methods while for other traits optimum levels were so different. Principle component analysis (PCA) show that, the most important traits to select high yielding chickpea varieties were days to flowering, days to physiological maturity, and plant height. These methods could help breeders to determine the range of optimum value for traits determining the adaptation of genotypes to a given environment. Boundary lines method is more reliable to determine of characteristic’s optimum levels in comparison with the averaging method.

It can be concluded that, determining the optimum levels of some dryland chickpea characteristics were closely equivalent in boundary lines and averaging methods while, for some characteristics, the estimated optimum levels by these methods were so different. Because boundary lines method is considered the data distribution process and gap data in databank, therefore, its estimations for the optimum levels of the characteristics are more accurate and reliable than the averaging of the high yielding group.