The Effect of Constant Temperatures and Osmotic Potentials on the Germination of Sugar Beet

Gummerson, R. J. 1986. The effect of constant temperatures and osmotic potenl!als on the germination of sugar beeL-J. exp. Bot. 37: 729-741. The germination-time curves for a series of experiments at constant temperatures and osmotic potentials were analysed to produce a relat10nship between germination, time, temperature and osmotic potential. The analysis involved the concept of hydrothermal lime which is a combination of potential above a base potential, temperature above a base temperature and time. For most of the population of sugar-beet seeds examined, the hydrothermal time required for germination was constant. The base temperatures of the seeds were similar but the base potentials varied between seed s. Analysis of the times to germination of fract10ns of the seed population, grouped according to the order in which they germmated, showed that the base potentials were almost normally distributed An equation using this distribution can be used to predict the lime course of germination over a wide range of temperatures and potentials. The equation uses five parameters to descnbe a seed population: the proportion of live seeds, the hydrothermal lime required for germination, the base temperature and the mean and standard deviation of base potential. Key words-Germination, temperature, water potential. Correspondence to: Broom's Barn Experimernal Station, Higham, Bury St. Edmund5, Suffolk, U.K. INTRODUCTION In the absence of any restriction in aeration, the rate at which sugar-beet (Beta vulgaris) seeds germinate is controlled largely by their temperature and the amount of water available to them. The water present in soil can be withheld by matric potential and by osmotic potentials created by the solutions of fertilizers placed near the seed. The seeds within a sample will germinate over a range of times, and a graph of accumulated germination fraction against time for seeds in constant conditions is a sigmoid curve. Germination of seeds at different constant temperatures produces curves with similar shapes but with different displacements along the time axis and covering different ranges of time. In some species, there is a temperature range over which the rate of germination of a particular fraction of seeds (the reciprocal of the time taken for the fraction to germinate) is linearly related to temperature. With species in which the base temperature (the intercept on the temperature axis) is the same for all fractions of the seeds, the different germination-time curves for experiments at different temperatures can be brought together on a single curve by plotting germination against thermal time, the product of the temperature minus the base temperature and time (Garcia-Huidobro, Monteith, and Squire, 1982; Washitani and Takenaka, 1984; Washitani, I 985). In this way the germination curves from experiments at different temperatures can be compared and the effect of temperature eliminated. © Oxford University Press 1986 730 Gummerson-Germination of Sugar Beet The effect of differences in water availability on the time course of germination is similar to that of temperature. Hegarty {1976) noted that for carrots and onions, the rate to 50% emergence was linearly related to water potential. This paper will show that a linear relationship exists between germination rate and potential for sugar beet. This leads to the concept of hydrothermal time which is an extension of thermal time to include a factor for the effects of variations in the availability of water. Central to this analysis is the assumption that seeds possess intrinsic characteristics which govern the way they germinate in various conditions of temperature and water potential. These characteristics are described in the analysis which follows, and are used with the concept of hydrothermal time to produce a single equation which describes the time course of germination in a range of temperatures and potentials. MATERIALS AND METHODS The experiments were designed to produce time courses of germination at a range of constant osmotic potentials and a range of constant temperatures. The same batch of seeds, from a commercial bulk of pelleted sugar beet, variety Regina, were used throughout. One hundred seeds were placed in a moistened fluted filter paper in a polythene box following the technique described by Hibbert and Woodwark (1969). There were six replicates of each treatment. The experiments were carried out in refrigerated incubators. Temperature was monitored at 15 min intervals using thermistor probes at up to six places in each incubator. Temperature variation in time was generally less than 0-5 °C, except when the boxes were removed for counting. As there was some spatial variation of temperature in the incubators, treatments were distributed in blocks and the temperature of each block was monitored. Temperatures given are the averages of all readings. Germinated seeds were counted and removed at intervals ranging from a few hours at 20-5 °C to several days in the later stages of the slower experiments. A seed was taken as having germinated as soon as any part of the seedling appeared beyond the pellet. For the longer experiments at low potentials the seeds were periodically transferred to fresh boxes to ensure that drying of the filter papers would not cause serious changes in potential. A range of osmotic potentials was created using five concentrations of sodium chloride (0, 0-051, 0-095, 0-184,0·273 moles kg')which gave nominal potentials ofO, -0-2, -0·4, -0-8, -1 ·2 MPa but as the same solutions were used at different temperatures, the true potentials vary by up to 5%. The true potentials at 9·6 °C were 0, -0-226, -0-418, -0-798, l · 178 MPa. (Calculated using data from Lang, 1967). Each of these concentrations was included at 7· 7 °C, 9·6 °C, 14· 3 °C and 20-5 °C. The three lower concentrations were used at 5 °C and a single experiment was carried out at 6-7 °C with no salt present. Analysis This paper takes as its starting point the theory used by Garcia-Huidobro et al. {1982) to analyse germination data from experiments at different constant temperatures. The theory is based on the assumption that the order in which seeds germinate is independent of temperature. Each seed can, therefore, be assigned a value of G, the germination fraction at which it germinates and this value of G can be used to identify each seed or small group of seeds. The application of thermal time theory to germination is based on the observation that for some species there is a temperature range over which the germination rate for a particular fraction of the seed population is linearly related to temperature. This can be represented by the equation: 1/t(G) = (T-Ti,)/8(G) (1) where tis the time to germination and 8 is the thermal time to germination. The notation t(G) and 8(G) is used to show that these variables are functions of G and vary from seed to seed. The intercept on the temperature (T) axis is the base temperature Tb. If all the seeds in a population germinated at the same time, the germination-time curve would be a step function and t(G) would be constant. The variation in the time to germination which produces the sigmoid shape of the germination time curve could be caused by the seeds having different base temperatures or different thermal time requirements for germination or both. If both varied independently, it is poss ible that the order in which seeds germinated would depend on temperature. The germination fraction, G could not then be used to identify individual seeds. Gummer son-Gennination of Sugar Beet 731 The results from the experiments described in this paper are 24 germination-time curves, each for a different combination of temperature and osmotic potential. If it can be shown that the rate of germination is linear with temperature and that the base temperature is independent of germination fraction and of potential, the germination-time curves can be reduced to five germination-thermal time curves, one for each potential. The next stage is to show that the rates in thermal time are proportional to potential. This allows an equation similar in form to equation 1 to be written: