This volume presents a conceptual approach to plant cell differentiation that differs in a number of respects from those already present in the literature. We seek to show how every cell has an individual competence to respond to the signal inputs that may impinge upon it and how every cell then has an individual qualitative and quantitative response. Central to this target cell concept is the premise that each cell is selective and can therefore discriminate amongst the many incoming signals to which it is exposed by an ability to perceive them and to respond to them.

Because each cell occupies its individual position within the plant body, the intensity or diversity of the signal inputs that it receives are not themselves identical. Hence, each cell is a unique individual and displays a unique target status even though it may also possess considerable commonality with its neighbours. We define this target status of a cell as the selectivity of its response to a signal and the intensity of that response.

The target cell concept arose originally from notions that were current amongst insect and mammalian scientists stating that a regulatory chemical produced in one organ would be perceived and activated upon by the cells in a distant organ – a specificity that operated between two distinct cell types. As the evidence for specificity of response to hormonal inputs increased during the twentieth century, developmental biologists saw this ability of cells to discriminate amongst the multitude of chemical signals to which they were exposed as a marker of the cell's ability to discriminate between them.

Despite enormous biological diversity, uniformity within a species remains remarkably constant. Trees can be instantly recognised by their shape or form, and leaf and flower structures are a basis of identification and classification. Roots conform to each species expectation, and the natural mutant that does not conform to the species type is a rarity.

This means that the society of cells that make up the plant body is under very strict control. No one branch can outgrow its neighbours and distort the overall shape of the tree. The buds that are terminal on any branch keep pace with, but do not outstrip those on the neighbouring branches. The phenomenon of apical dominance, used by crop growers from time immemorial as the basis to reshape by pruning or to improve yields, is intertissue signalling at its most evident. The consistency with which lateral buds will grow when the terminal bud is removed is central to plant culture and pruning systems throughout the world. Not until Frits Went demonstrated the presence of a chemical substance (indole-3-acetic acid) in Avena coleoptile tips and then showed that it would replace the terminal bud in inhibiting the growth of laterals was the first intertissue signal molecule properly established. With the knowledge that the auxin molecule is transported in a polar fashion from cell to cell as it passes from terminal bud to tissues below, it is not difficult to understand how a terminal bud can control the growth of fellow buds below over distances that are relatively short.

In the previous chapter, we considered the perception of auxin and ethylene in different target tissues. While these hormones were treated separately, the essential mode of action of ethylene and auxin is to relieve a pre-existing repression of response (see Figures 7.2 and 7.4). It is now clear that this mode of action of hormonal signals is widespread amongst plants. In this chapter we look at three further examples, the cytokinins, gibberellins and brassinosteroids, and again examine the evidence for perception of these signals via binding proteins or receptors in different target cells. We additionally examine the evidence that these developmental cues operate through the relief of pre-existing repressions of molecular responses in each cell type examined.

Cytokinin perception in the context of receptors and target cells

Two spectacular events in plant development are attributable to the action of cytokinins. The first, is the conversion of the cells of a callus culture into the organisational complexities of a shoot meristem. Cytokinins act not alone, but in concert with auxin, the ratio of one to the other being critical for optimal organ development. The second is the maintenance of the non-senescent state in specific tissues. Here, cytokinins act as repressors of cell death programmes in many target cell types, most notably those of the leafy tissues of herbaceous plants. Seemingly, there is a requirement for cytokinins primarily synthesised in the root meristems for the retention of metabolic function in the green shoot.

Plants are remarkable in that during evolution from the single cell to the multicell state they developed centres of cell division, the meristems, as the principal repositories of all genetic information. In essence, whether it be the single apical cell of a liverwort or the multicellular dome of tissue of the higher plant, the meristem holds the blueprint of the species. It is only from the divisions of their meristematic cells that the plant body can continuously enlarge and reproduce. Whereas an embryo generates polar identities and a meristematic initiation from the two opposing ends of the zygotic cell, whole plants do not have a restriction to two meristems and they develop unlimited numbers of new primordia as the body of the plant continues to increase in size and cell number.

Anatomists, biochemists, molecular biologists and developmental botanists each see the same plant from different viewpoints. The anatomist studies the structural and visible characters of cells and tissues and describes them accordingly. The biochemist homes in on the functional processes of metabolic control and enzymatic activity attributable to specific plant parts – photosynthetic activity in leaves, for example. The molecular biologist seeks the genetic control of biochemical processes and is particularly attracted to the performance of mutant plants with abnormal behaviour, using them to probe the genetic control of the normal.

In 1985 one of us published a small book called Seed Ecology. It contained 42000 words and cited 334 references. It was successful in introducing a generation of ecologists to our subject, but it is now seriously out of date and has been out of print for some time. The book you are now holding contains 94000 words and cites 1117 references. Only a small part of this expansion can be attributed to covering any part of the subject in more detail; nearly all of it reflects simply the massive increase in interest in seed ecology in the past 20 years. One sign of this expansion was the launch in 1991 of the journal Seed Science Research, providing a major platform for fundamental work in seed biology, including ecology; 37 of our cited references are from that journal. More recently, the International Society for Seed Science was founded in 2000. This society sponsors meetings on all aspects of seed science, including, for the first time in 2004, a major international meeting on seed ecology. Our cited references also reflect this recent growth: 82% are from the past two decades, while 15% are post-1999.

Recent work in this field has transformed our understanding of many aspects of seed ecology, especially dispersal, storage in the soil and the ecological role of seed dormancy.

Seedling establishment represents the final hurdle in the process of regeneration. The start of the seedling phase may be defined by the completion of germination. In most cases, this is marked by the extrusion of the radicle (root), which anchors the seedling in the soil, followed by the plumule (shoot), which grows towards the light. If the seed is buried, the plumule has to push its way through the soil to the surface, a process that expends energy from the seed's reserves. In most field experiments, the appearance of the shoot at the soil surface (emergence) is the first sign that germination has taken place and is usually taken as the starting point in demographic studies. However, although it is seldom measured, mortality between germination and emergence is probably quite high, especially if the seeds are emerging from any depth (See Section 7.3). The emerging seedling faces a new set of hazards. Whereas a lack of light, water or nutrient has little or no effect on seed survival, these become major causes of death in seedlings. The predators and pathogens that menaced the seed are replaced by a different set at the seedling stage.

Early growth of seedlings

The term ‘seedling’ is used very loosely in the literature to cover young plants generally, and it is seldom defined strictly, even within the contexts of individual studies (Fenner, 1987; Kitajima & Fenner, 2000). The main problem is defining the end point: when does a seedling cease to be seedling?

The period of seed development on the parent plant can be one of the most hazardous phases in a plant's life cycle. In many plants, only a very small proportion of the ovules eventually mature into viable seeds. This is because many flowers fail to produce fruits, and many of the ovules in fruits fail to produce seeds. Studies on a wide range of species have recorded huge variations in fruit set and seed set (Wiens, 1984; Sutherland, 1986). Seed losses in the pre-dispersal stage may be due to pollination failure, genetic defects, lack of resources for development or seed predation. This chapter considers both the proximate causes of mortality and the evolutionary consequences of mortality in seeds before they have been shed by the parent plant.

Fruit and seed set

Fruit set is characteristically very low in certain species. In Yucca elata, only 6.6% of the flowers were recorded as producing mature fruits under field conditions (James et al., 1994). In Aesculus californica, fruit set in nature was shown to be about 10% (Newell, 1991); in Cornus sanguinea, it was shown to be between 8 and 22% among different populations (Guitián et al., 1996). The Proteaceae as a family are notable for their low fruit set (Charlesworth, 1989; Ayer & Whelan, 1989; Wiens et al., 1989). In a survey of 18 species growing under natural conditions, Collins & Rebelo (1987) recorded fruit set values that ranged from only 0.1 to 7.2%.

Few things are more important for plants than ensuring that germination takes place in the right place and at the right time. Sometimes this requirement is satisfied by germination as soon as seeds are shed, but in most plants there is a delay of anything from days to decades. One important mechanism for achieving this delay, although not the only one, is seed dormancy.

Types of seed dormancy

There are three fundamentally different types of seed dormancy, at least two of which have evolved on several separate occasions (Baskin & Baskin, 1998). These dormancy types are morphological, physical and physiological. In morphological dormancy, the seed is immature when shed and a period of growth and/or differentiation is required before germination can take place. Seeds with physical dormancy have impermeable testas or pericarps; the embryo is therefore dry until the seed coat is broken and water enters. Physiological dormancy prevents germination until a chemical change takes place in the seed. Dormancy types may be combined in the same seed – a combination of morphological and physiological (morphophysiological) dormancy is very common, but physical and physiological dormancy are rarely combined. The combination of physical and morphological dormancy is, of course, impossible. Note the crucial distinction between physiological dormancy, which is reversible, and the other two types, which are not. Thus, generally speaking, physiological dormancy permits a more flexible response to the environment than the other two types of dormancy. Morphological dormancy seems to be the most primitive type (Baskin & Baskin, 1998).

Seed dispersal has long been an object of fascination to biologists and the general public alike. Examples abound of structures that have clearly evolved to promote dispersal by wind or on the outside or inside of animals, but it is only recently that attention has turned to the question of just how well these structures work and what happens to the seeds of all those species (the majority) with no obvious adaptations for dispersal. Few things in seed ecology have changed more in recent years than our understanding of seed dispersal.

Wind dispersal

Any structure that increases air resistance of the dispersule is likely to improve dispersal by wind. Some morphological adaptations impart lateral movement directly, but the great majority merely slow the rate of fall, relying on wind to provide the lateral motion (Augspurger, 1988). Wind dispersal has probably received more attention than all other dispersal modes, since it can be investigated (even if not totally satisfactorily) in the laboratory and is relatively amenable to mathematical models of varying complexity (Sharpe & Fields, 1982; Green, 1983; Matlack, 1987; Greene & Johnson, 1989, 1990, 1993, 1996; Hanson et al., 1990; Andersen, 1991).

These models are essentially of two sorts: (1) analytical models that describe seed densities directly (e.g. Greene & Johnson, 1989) and (2) individual-based models that simulate the movement of individual seeds (e.g. Andersen, 1991). Seed shadows are then produced by summing simulations for large numbers of seeds. See Jongejans & Schippers (1999) for a relatively simple individual-based model.

Relatively little is known about the cause of seed loss in the soil once initial dispersal has taken place. The vast majority of dispersed seeds fail to emerge as seedlings. Seeds buried in soil tend to have a more or less exponential decay (Roberts & Feast, 1973). Some may be eaten; others may be attacked by pathogens. Another possible fate is germination at depths that are too great to permit emergence. A large fraction simply may lose viability in the course of time and die of old age. We will examine each of these possibilities in turn.

Post-dispersal predation

Post-dispersal seed predators are typically granivorous mammals (e.g. rodents), birds (e.g. finches) and insects (e.g. beetles and ants), but the taxonomic range of seed-eating organisms is wide and includes slugs (Godnan, 1983), earwigs (Lott et al., 1995), fish (Kubitzki & Ziburski, 1994) and crabs (O'Dowd & Lake, 1991). Seed predation can be regarded as a specialized form of herbivory. Because it impinges directly on the capacity of plants to regenerate, it can play a key role in population dynamics. The proportion of seeds eaten varies greatly between species, locations and years, but it is often extremely high. For example, capuchin monkeys were recorded as eating 99.6% of the seeds of the wind-dispersed forest tree Cariniana micrantha at a site in Amazonia (Peres, 1991). In a Costa Rican tree, Ocotea endresiana, rodents ate 99.7% of the dispersed seeds within 12 months (Wenny, 2000b).

Germination involves the imbibition of water, a rapid increase in respiratory activity, the mobilization of nutrient reserves and the initiation of growth in the embryo. It is an irreversible process; once germination has started the embryo is committed irrevocably to growth or death. Externally, germination is marked by the bursting of the testa and the extrusion of the plumule or radicle. In this chapter, we examine the influence that various environmental factors have on the process.

Temperature and germination

Constant temperatures

Quite apart from its well-documented effects on the induction and breaking of dormancy, temperature has important effects on germination itself. These may be divided, conveniently but rather arbitrarily, into effects of constant and alternating temperatures. The latter are considered later. In seasonal climates, temperature is of course a good indicator of the time of year and is therefore implicated strongly in determining the timing of germination. Washitani and Masuda (1990) conducted a remarkably detailed study of germination in a Japanese grassland, in which germination of almost all species was confined to the spring–early summer period. They found that the temperature at which seeds began to germinate, when subjected to gradually increasing temperatures in a standardized screening programme, was linked closely to the observed timing of emergence in the field (Fig. 6.1). Interestingly, emergence timing was not correlated at all with presence of dormancy or requirements for dormancy breaking, illustrating the important point that dormancy normally plays little part in determining germination timing.

A seed is an embryo plant wrapped in a protective covering of maternal tissue (the testa). It is generally provided with a supply of nutrients contained in a separate tissue (the endosperm), though in many cases all the nutrients are absorbed by the seed leaves (the cotyledons) during the course of development. The primary function of the seed is reproduction. This does not necessarily result in an increase in numbers of the species. In a stable population, each adult is eventually replaced by another adult. This is achieved by the production of large numbers of offspring, most of which will die before reaching maturity. A seed therefore has several functions in addition to multiplication. Its small size (at least in comparison with its parent) renders it well suited for dispersal and the colonization of new areas. In addition, many seeds can withstand a much wider range of environmental conditions than the adult plants, especially extremes of drought and temperature. Their ability to undergo a period of arrested development and persist in a state of diapause is important as a means of persistence for many species, but it is especially crucial for annual plants that do not survive as adults during periods of unfavourable conditions such as seasonal cold or drought.

Sexual vs. asexual reproduction in plants

An important feature of seeds is their genetic variability. This derives from the fact that (except in the case of apomicts, mentioned below) they are the products of sexual reproduction.

In most plant communities with a closed canopy, the establishment of seedlings usually requires at least some degree of disturbance to provide areas free of existing vegetation. Established plants have a clear advantage over seedlings in their ability to intercept light and monopolize other resources such as water and nutrients. Gaps that are created by any agency in vegetation can be considered ‘competitor-free spaces’ that provide opportunities for seedling establishment (Bullock, 2000). The study of gaps and their role in promoting recruitment has been an important focus of investigations into regeneration and species diversity in plant communities over the past few decades.

Gaps, patches and safe sites

A gap is an area that is at least partially free of vegetation, where there are sufficient resources available to permit the recruitment of new individuals. Gaps are not always necessary for regeneration, especially in cases where vegetation itself can ameliorate conditions in a harsh environment (see Section 8.7). The term ‘patch’ is often used in this context to mean much the same as a gap, but it is a less satisfactory term because it suggests something that is stuck on rather than removed. It is perhaps best reserved as a term to refer to the successional vegetation that comes to occupy a former gap. A useful term coined by Harper (1977) is ‘safe site’.

Mature seeds are shed from the parent plant and, sooner or later, find themselves at the soil surface. Germination may then take place immediately or may be delayed for an indefinite period. During this time, the seeds on or in the soil are said to form a soil seed bank. A number of schemes have been published with the aim of describing and classifying the different types of seed bank (Csontos & Tamas, 2003). Perhaps the most widely employed in temperate regions is the scheme of Thompson & Grime (1979), which recognizes four types (Fig. 4.1). Type I are autumn-germinating species, whose seeds are present during the summer only. Type II are spring-germinating and present mainly during the winter. Seeds of both are often, but not always, dormant when shed, this dormancy being broken by a period of low or high temperatures in types I and II, respectively. Both are described as transient, since normally no seeds persist in the soil for more than one year. In types III and IV, a smaller or larger fraction of the seed output enters a persistent seed bank, which survives for more than one year. In fact, types III and IV are clearly ends of a continuum, and it is now apparent that the same species may behave as type III or IV at different times and in different places (Cummins & Miller, 2002).