Cardinal Temperatures for Germination in Three Millet Specieses (Panicum miliaceum, Pennisetum glaucum and Setaria italica)

MATERIALS AND METHODS

Seeds of Common millet (Panicum miliaceum), pearl millet (Pennisetum galucum) and foxtail millet (Setaria italica) were obtained from South East of Iran (Birjand, 59°12’N longtitude, 32°52’ E latitude, 1491 m altitude). All empty seeds were cleaned by soaking in water for 2 min. Experiments were conducted in germinators at the Instituto de Agricultura Sostenible (CSIC, Cordoba, Spain) from March to August 2004. Seeds were germinated at constant temperatures ranging from 10 to 45°C with 5°C intervals in petridishes, lined with two layers of germination papers which were wetted periodically with ordinary tap water as when required. Tests were conducted in total darkness. The experiment was replicated twice with 60 seeds that were drawn randomly from the original seed stocks. In each experiment, 4 randomized replications were used. The time was noted when a replicate had been prepared and placed in germinator. After beginning of germination, the seeds with a radicle equal the length of the seed (approximately 4 mm) were scored as germinated and removed. Collected data used to plot mean cumulative germination curves against time. From these curves, time to 50% of germination was determined by interpolation as Dumur et al. (1990). Intersected lines models were fitted to the data of the reciprocals of time to 50% of germination versus temperatures using below equations:

Where, y = germination rate, T = mean temperature (°C), Tb = base temperature (°C), Tc = ceiling temperature (°C) and To = optimum temperature (°C), a1, a2, b1 and b2 are regression coefficients.

The model was fitted using nonlinear procedure in sigmaplot for windows version 7.0.2001.

RESULTS

Table 1 shows time to 50% of germination at different constant temperatures. Results indicated that time to 50% of germination changed as a U shape curve that is typical of the temperature reaction of many developmental stages, including germination (Olivier and Annandale, 1998).

Comparison of germination rate at different constant temperatures indicated that germination rate of all specieses increased with increase in temperature to 35°C. Regression of the rate of germination versus temperatures separately for the sub and supra-optimal temperatures ranges indicated highly significant linear relationship in each species (Table 2). Figure 1 shows the results of model fitting. Calculated cardinal temperatures, regression equations and R² are presented in Table 2. The slope of line showed positive response for temperature between Tb and To and negative response between To and Tc. Small differences between calculated cardinal temperatures in the two independent replications of experiments was related to extrapolation of two fitted lines to intersect X-axis (where germination rate is equal to zero) to determine base and ceiling temperatures.

Table 1:	Time to 50% of germination (hours) against different constant temperatures

Table 2:	Equations, calculated cardinal temperatures and R2 of fitted intersected lines models
Tb, To and Tc are base, optimum and ceiling temperatures, respectively. R2 is correlation coefficient

Fig. 1:	Fitted intersected lines models to germination rate against different constant temperatures in studied millet specieses

Therefore these small differences can be ignored and mean of the two estimates of cardinal temperatures can be used.

DISCUSSION

The results of the present study confirmed that in the normal conditions, the germination rate of the specieses is highly affected by temperature and findings are consistent with past works on pearl millet (Garcia-Huidobro et al., 1982) and other crops (Covell et al., 1986; Elli et al.,1986; Mwale et al., 1994). Also, results indicated that cardinal temperatures changes between different specieses. Other studies have shown that the cardinal temperatures for germination depend on species and within specieses vary significantly between genotypes (Ann and Binning, 1987). Comparison of cardinal temperatures for different specieses showed that cardinal temperatures in common millet are more than two others one.

Intersected line model was successful in describing germination response to temperature. Consistency of calculated cardinal temperatures for pearl millet with previous studies (Garcia-Huidobro et al., 1982) revealed that calculated cardinal temperatures for common millet and foxtail millet (that there is not any information about their cardinal temperatures) should be exact and reasonable. Accurate control of temperature by using germinators (±0.3°C) and the consistency of the results of two separate experiments ensured us that the cardinal temperatures are valid and acceptable and can be used in other researches with different objects. For example, the cardinal temperatures could be used for prediction of subsequent developmental stages of the crop (Covell et al., 1986; Hadley et al., 1983).

ACKNOWLEDGMENTS

The authors wish to acknowledge the Instituto de Agricultura Sostenible (CSIC), Cordoba-Spain and especially Prof. F.J. Villalobos for all supports and providing all needed equipments to complete this research.