Modeling of Various Developmental Stages in One Spring Safflower by Temperature and Day Length

Introduction
Safflower is an annual, broad leaf oil seed crop of the family compositae adapted chiefly to dry land. Safflower is originally a cold season and long day crop. It is believed to have originated in southern Asia and is known to have been cultivated in China, India, Iran and Egypt almost from prehistoric times. During the middle ages, it was cultivated in Italy, France and Spain and soon after discovery of America, the Spanish took it to Mexico and then to Venezuela and Colombia. Safflower was originally grown as the flowers that were used in making red and yellow dyes for clothing and food preparation. Today also safflower can be cultivated for oil, foots medicinal purpose and cosmetic products. The planted area has reached 795118 ha and resulted in 731425 tons of production in the world during 2004. Planted area in Iran is about 6000 ha with 1 ton seed yield per ha. The highest planted areas in Iran belong to Isfahan, Khorasan and Yazd provinces respectively. Plant development can be defined as a programmed qualitative change in the plant form, which leads plant to maturity, and researchers call it the phases of development or phenology. Plant development is different from plant growth, because plant growth is the result of photosynthesis process during which dry matter accumulates. Water, nitrogen and photosynthetic matters affect growth and development. Under irrigated conditions, all factors except climatic elements can be controled. Thus, only the climatic elements can change the growth and the development of plants and in this case, temperature and day length have the main role.

Materials and Methods
Field experiments were conducted in 2006-2007 and 2007-2008 at the Kaboutar Abad Agricultural Research Station, to model the response of Arak variety to day length and temperature changes in different planting dates. Developmental rate of each stage was estimated using the inverse of developmental duration as the dependent variable and various temperature variables derived from daily maximum (Tmax), minimum (Tmin) and mean (Tmean), day length (DL) and the combination of these variables as the independent variables in stepwise regression models. A step of the regression was considered appropriate if the highest R2 was accompanied by the significant (R<0.1) regression coefficient and partial coefficient of determination.

Results and Discussion
The number of days from planting to emergence, emergence to head visible, emergence to ripening and flowering to ripening were affected by planting dates. Developmental stages reduced as temperature increased. Day length had the most effect on emergence to head visible period and reduced as day length increased. Tmin was the only variable which entered the regression model and explained about 76% of variation in the rate of emergence of Arak. About 84% of variation in the rate of emergence to head visible was explained by Tmax × DL. Most of the variation (about 66%) in the rate of emergence to ripening period was explained by T2mean × DL2, Tmean × DL, T4min, T2min, and Tmin were the other variables that entered the model and reached the precision of the model by about 80%. T4min was the only variable that entered the regression model and explained 55 % of variation in rate of flowering to ripening period.B ased on the ration of absolute value of maximum deviation from the regression line , to actual period of developmental stages (expressed in percentages), long developmental stages were well estimated that indicate the efficiency of the calculated models.

Conclusion
The results of this study indicate that the contribution of temperature and day length to various stages of development do not have the same differences in physiological nature of developmental stages and differences in the responses of developmental stages to climatic, edaphic and agronomic factors and the interactions of these factors with each other and with plant genotypes may be responsible for these reactions. Thus, a separate model may be needed for each cultivar and each developmental stage. Dividing the country to homogenous groups, with respect to temperature and day length, may increase the accuracy of estimating models. This requires availability of sufficient phonological information from various climatic zones.