How can we possibly know when and how many seeds will germinate, in the environment, when we have only imperfect models developed in predictable lab conditions and are dealing with complex ‘biological clocks’ that slow down or speed up depending on a diverse range of factors? In this issue, Rowse & Finch-Savage (see pp. 101–108) address this area, providing significant advances related to predictions of seed germination, and do this using hydrothermal models in which the effects of temperature and water potential are considered simultaneously.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Beyond basic hydrothermal models at supra-optimal temperatures
Hydrothermal models form the basis for many current efforts to predict seed germination. A key feature of these models is that each seed accumulates hydrothermal time (quantified progress toward germination) according to the temperature and water potential of incubation in relation to the minimum (base) temperature and water potential at which the seed can progress. A key advantage of hydrothermal models is that their equations apply to the entire seed population and lead to simultaneous predictions of germination rate and percentage. Basic hydrothermal models aren’t applicable at temperatures above the optimum, where germination is progressively inhibited, and it is this problem area which is addressed by Rowse & Finch-Savage. They use an upward shift in base water potentials to explain the reduction in germination rate and percentage that occurs at supra-optimal temperatures.
Several lines of evidence suggest that base water potentials, which are distributed approximately normally within the seed population, are both physiologically and ecologically relevant. Alvarado & Bradford (2002) found that true potato seeds incubated in the supra-optimal range also showed an upward shift in base water potentials. Shifts in base water potentials are associated with control of germination via abscisic acid (ABA) (Toorop et al., 2000), weakening of tissues that form a barrier to radicle emergence (Welbaum et al., 1998), loss of seed dormancy during after-ripening (Meyer et al., 2000; Alvarado & Bradford, 2002) and sensitivity to decreasing water potential for species from contrasting habitats (Allen et al., 2000a). The emerging focus on models with physiologically relevant parameters (i.e. rather than strictly empirical) will undoubtedly prove critical in making better predictions of seed germination and in related processes such as dormancy loss and seedling emergence (Benech-Arnold et al., 2000; Forcella et al., 2000).
How hydrothermal threshold models work
The biological clock for many important plant processes (e.g. flowering, bud opening and seed germination) speeds up and slows down according to the sum of all reactions involved. Because temperature plays such a critical role in the timing of many plant processes, ‘thermal time’ is frequently considered an acceptable proxy for the more complex ‘biological time’. With seeds, three ‘cardinal’ germination temperatures – minimum, optimum and maximum (e.g. Fig. 1a in Rowse & Finch-Savage) – have been recognized for more than 150 yr (Alvarado & Bradford, 2002). To quantify the influence of temperature on germination, thermal time is mathematically expressed as the degrees above the minimum temperature (i.e. the base or threshold below which germination doesn’t occur) multiplied by the actual time to germination. Thermal time is possible where germination rates increase linearly between the minimum and optimum temperatures, which appears to be the case for seeds of most species.
Water uptake is clearly essential for seed germination yet seeds are frequently subjected to great fluctuations in water availability. Initial water uptake into dry seeds is a physical process, viewed as satisfying the seed matric potential. If imbibing seeds are exposed to a reversal of the water potential gradient that drives water uptake, such as when the seed zone dries, seeds will lose water. Analogous to thermal time, ‘hydrotime’ quantitatively describes the rate of progress toward germination as a function of seed water potential. At reduced water potentials, progress toward germination is progressively restricted. For purposes of modelling seed germination, the effects of temperature and water potential can be combined into a single ‘hydrothermal time’ (for mathematical equations see Rowse & Finch-Savage, eqns 1–3).
Once the parameters for hydrothermal time (the constant hydrothermal time required for germination, mean base water potential for the seed population, base temperature for germination, and standard deviation of base water potentials) are known, germination time courses can be predicted for any temperature and water potential combination. These predictions permit germination progress to be summed across variable environments, such as those that include fluctuating temperatures and water potentials (Bradford, 2002). The fact that base temperature and base water potential thresholds are limits below which radicle emergence does not occur greatly simplifies the mathematical solutions required in predicting seed germination.
Hydrothermal modelling efforts provide a theoretical framework for quantifying the effects of temperature and water potential on seed and seedling processes under both controlled and natural conditions. In the field, predictions related to seed germination span all the processes that begin with inputs related to weather and soil, and end with outcomes for establishment and survival (Fig. 1). The seedzone, which is located near the soil surface, is characterized by the widest temperature and water fluctuations that occur within the soil profile. These fluctuations necessitate accurate measurements or estimates of seedzone microclimate. The difficulty associated with obtaining accurate knowledge of soil conditions near the soil surface, compounded by the fact that soil water potential may not accurately reflect seed water potential under fluctuating conditions (Allen et al., 2000b), is at least one motivation for adaptations to hydrothermal models used to predict seed germination. Finch-Savage and coworkers successfully predicted germination and emergence of carrot (Finch-Savage et al., 1998) by basing predictions on thermal time that accumulated only during periods when soil water potential was above the average threshold for the seed population. Additional modifications of the basic hydrothermal time model account for germination-directed physiology that occurs at water potentials too low to permit radicle emergence. Examples of such modifications include employing the concept of hydrothermal ‘priming’ time (Bradford, 2002) or virtual osmotic potential (‘virtual’ because values are empirical and not measured) (Rowse et al., 1999). The virtual osmotic potential model continuously integrates the history of seed osmotic potential, a concept that has been expanded to incorporate sub- and supra-optimal temperature effects and fits within the hydrothermal time model (Rowse & Finch-Savage, pp. 101–108).
It is often unclear how data from experiments conducted under controlled, usually constant, temperature and water potential conditions in the laboratory relate to the highly variable field environment. Simulation modelling provides the best possible test of the hypothesis that seeds and seedlings integrate their successive water potential and thermal histories under field conditions in predictable ways, a hypothesis that has been verified with hydrothermal time concepts for seed dormancy loss in Bromus tectorum (Bauer et al., 1998) but has so far received limited attention in modelling seedling emergence in the field (Finch-Savage et al., 1998; Roman et al., 1999; Shrestha et al., 1999). The advancements in hydrothermal modelling proposed by Rowse & Finch-Savage will fit nicely into field modelling efforts, leading to considerable progress in developing hydrothermal threshold models that more accurately predict seed behavior in the field.