As we examine the life history of a plant we need to start somewhere. Because humans start growing plants of interest from propagules, it makes some sense to start with seeds. Of course a seed is part of a life cycle so the question of what came first, the plant or the seed comes to mind. While we cannot resolve the cycle into a linear argument from a development perspective, we CAN resolve the question from an evolution perspective. The answer is different than the one for the chicken and egg question. For plants, the plant evolved before the seed very clearly. However eggs evolved long before there were chickens on our planet.
From a developmental point of view, let us for now just say that a seed germinates into a plant, the plant produces a flower, part of the flower becomes a fruit, and that fruit contains a seed. The seed can be planted to start this life-cycle over again. However, it is seed-chauvinism to start here. Many plants do not reproduce via seeds at all.
A seed consists of three parts: a Dormant Embryo, a Storage Tissue, and a Seed Coat. Not every seed that has evolved on this planet has precisely the same structure. In some seeds, the endosperm is retained as the storage tissue. In other seeds, the endosperm is more or less used up to put storage chemicals into the embryo itself (commonly in the cotyledons). Below is a diagram of two hypthetical seeds. The upper seed shows a dicot that lacks endosperm; its storage material (blue) is held in the cotyledons That would be an apomorphy...a derived condition. The lower seed shows a monocot that has a well-developed endosperm (also blue). Both of these examples qualify as true seeds because they possess all three parts needed to make a true seed.
Below is a photomicrograph of a longitudinal section of a Capsella seed. The seed coat is the three-layered outer covering. The outermost and innermost layers of the seed coat pick up a red dye used in the preparation of this sample. That dye is preferentially held by parts of the specimen that have waxy or brittle biomolecules. Cutin, suberin, lignin are three components of plant cells that pick up the red dye. This leads us to the idea that the integument has a waterproofing (cutin/suberin) function and also possibly a mechanical strengthening (lignin) function that could impede herbivory.
Looking toward the inside of the seed coat, there is a bent embryo (dark looking) surrounded by some loose and light colored cells of the remaining endosperm. Indeed the paucity of endosperm leads us to the conclusion that much of the storage material has been moved to the embryo itself...perhaps much of that in the cotyledons. However, this seed is clearly somewhere between the two extremes diagrammed above.
The embryo has an axis (on the left) with a downward-pointing radicle including a root apex. The radicle ultimately penetrates the seed coat in seed germination, branches profusely, and becomes the primary root system. Toward the top of the embryonic axis, the radicle becomes hypocotyl...transitioning from root-type anatomy into stem-type anatomy as we shall see later. In seeds with hypogeous germination, this hypocotyl is too short and does not grow above the soil. In species with epigeous germination, the hypocotyl elongates rapidly, lifting the cotyledon(s) out of the soil and into the air. The axis has two appendages, the cotyledons; as there are two cotyledons, Capsella is a dicotyledonous plant. There are two classes of flowering plants, the dicots and the monocots, and they are distinguished (in part) on the basis of the number of cotyledons found on their dormant embryos. The part of the axis where the cotyledons attach is called a node because this zone of the axis does not elongate. At the extreme top of the axis is a shoot apex; after germination this shoot apex will elongate, and make appendages to produce the collection of stems and leaves that constitute the plant shoot.
The seeds are carried inside the fruit by animals called dispersers. The animal usually feeds on the fruit (ovary wall), but "tosses" the seeds (or passes them through its digestive system). This assures that seeds end up far away from the "mother" plant, where they can develop without competition from "mom." The fruit may possess special characteristics to attract and encourage dispersal by an animal. Other seeds may be dispersed by catapult (as you observed in the arboretum walk), parachute, or other means. Once dispersed, a seed needs to germinate and grow into a new plant, eventually maturing into a reproductive adult plant.
Some seeds sprout with just water and reasonably warm temperatures. This is true of most common garden plants. Such seeds are often referred to as "seeds lacking dormancy." Wild species usually have some kind of deeper dormancy to avoid sprouting in late summer/fall when the seeds are dispersed. This assures that tender seedlings are not frozen at a young age, but do not appear until warm weather arrives in springtime. For these species it takes something more than just water and warmth to germinate.
The mother plants of our "wild" species put their embryos into dormancy by using chemicals. Perhaps in hardy plants the vast majority use abscisic acid to keep their embryos from germinating too early. This chemical is broken down by enzymes in the embryo over time. The enzymes have a cold temperature optimum. Thus the enzymes are not active until early winter. The abscisic acid keeps the embryo dormant from dispersal to frost. After 6 weeks at about 4° C, the optimum for the enzymes that metabolize abscisic acid, the AbA is metabolized completely. But now the cold soil keeps the embryo dormant until the warmth of spring (20° C). It is an ingenious mechanism evolved in hardy plants. To accelerate this process artificially, seeds of wild plants can be harvested in late summer as they mature, placed in moist soil, and kept in a refrigerator (4° C) for four to six weeks. This treatment is called stratification. After the stratification is complete, then the plants are removed to a lighted greenhouse that is reasonably warm. This treatment is called vernalization. Now the seeds will have "experienced winter and spring" and will germinate.
Some other wild seeds have very thick seed coats that are thoroughly lignified and waterproofed. These species have evolved the ability to pass through the digestive system of an animal or survive pounding in the surf at the shore or go through a fire before they are even able to sprout. The degradation of the seed coat is called scarification, and this process permits water to pass through the seed coat so that the embryo can begin metabolism, elongate its radicle, and germinate. Many wild legumes, such as the Kentucky Coffee tree and the Sea Bean, have such heavy seed coats.
Yet other wild seeds have an exceedingly thin seed coat. The evolution of the thin coat is accompanied by dormancy that is overcome by either light or darkness. The light can penetrate the seed coat and the embryo inside can then detect whether it is deeply buried or right on the surface of the soil. Small seeds with thin seed coats often require light for germination; large seeds with thin seed coats often require darkness before they initiate germination. Think about the amount of storage materials in the seeds and the evolution of the two different signal responses will become apparent. Examples of seeds requiring light include lettuce, seed requiring darkness include garden pea.
Some very small seeds may have a very rudimentary embryo (only a few cells large) and almost no storage material. The orchids you observed in the arboretum are an example. These seeds will not germinate until they are allowed to develop further. The seeds in nature are dispersed into soils that have symbiotic fungi. These fungi degrade detritus in the leaf litter and nourish the orchid embryos in the dark. Once the embryo has "after-ripened" to some sufficient stage, it will germinate and grow into an adult orchid plant.
Seeds of desert species (30 N and 30 S latitude) often have no winter cold signals, few animals for scarification, no fungi for nourishment or degradation, etc. These species often experience a dry season and a wet season in their environment. Not surprisingly desert species have evolved ways to ensure that their seeds sprout only when the wet seasons has arrived. Desert plants invest their seeds with phenolic substances that inhibit seed germination. As long as they are dry these seeds will not germinate. If there is a passing shower during the dry season, some of the phenolic is dissolved and leached out of the seed. But one shower is not enough...there is still enough phenolic left in the seed to prevent germination. After repeated rinsing, the phenolic is sufficiently leached out so that the seed sprouts during a bona fide wet season.
Shown below is the biochemical mechanism for seed germination in two species: barley and lettuce.
With the imbibition of water, the hormone signal in barley, gibberellic acid (GA), is carried from the embryo to the aleurone layer of the endosperm. The GA activates the DNA for the gene encoding alpha-amylase in the aleurone cells. Transcription and translation of that gene results in the production of alpha-amylase inside the aleurone cells. This enzyme is shipped by ER into the Golgi, sorted and packaged into vesicles, and exported through the cell membrane by exocytosis. The amylase is thus dumped into the endosperm area. There the amylase breaks down starch into the sugar maltose which is transported to the embryo. The sugar fuels respiration in the embryo so it can grow. The radicle protrudes from the seed coat, and germination is accomplished. After plant physiologists figured this out, we could easily tell the brewing industry how to produce more beer per bushel of barley. What was the answer?
A similar mechanism exists in lettuce (lettuce has no aleurone and the embryo is a dicot), but the activating chemical is a pigment called phytochrome. This chemical exists in two different forms: Pr and Pfr. Pfr is the form of phytochrome that photoactivates the genes for amylase in lettuce; Pr is inactive. Whether a lettuce seed germinates depends on how much of each of these two forms of phytochrome is present in each cell. Typical lettuce seed batches germinate at 30-60% if placed in darkness because at least this percentage of seeds have enough Pfr to stimulate germination. If, however, you put the lettuce seeds in red light (660 nm), the red light causes all the Pr to change into Pfr. Now 85-95% of the seeds can sprout because they all have an abundance of Pfr inside. On the other hand, if you put lettuce seeds in far-red (730 nm) light, the far-red light causes all the Pfr to change into Pr. In far-red light, then, all the seeds have essentially no Pfr and so very few (0-5%) actually sprout. After plant physiologists explained this, Burpee Seed company understood why customers have trouble getting lettuce seeds to germinate well in their gardens...they tend to "bury" the seeds...they need to be exposed on the surface!