There are many orchard planting and training systems.

The choice of "high density" or "semi-dwarf" is a major decision that takes a lot of thought!

Within each support system there can be many different trimming and training methods used to develop the orchard.  For new fruit growers it can be a very confusing thing to read about all the different systems, their plus and minus aspects, and which system to use.  For the most part many of these systems really are for the experienced commercial fruit grower wishing to meet specific goals for their commercial fruit operation.  Many books and papers have been presented on all these different systems, and from a skeptical and cynical point of view, many of the systems are pipe dreams of researchers looking for something to publish.  As you gain more experience in orchard management and in the fruit industry, then you may want to try some of these different systems, but in order to gain that experience, we suggest you stick with the tried and true systems that have been used for the past 20 years.  Every different variable that you put into the equation, makes the results more difficult to predict, so for starters, stick with a system that has been fairly pro

To be simplistic, there are only three basic tree support systems:
Self-supporting, stand alone trees in low to moderate densities.
Single trees supported by poles or stakes, in higher moderate densities.
High density systems fully supported by trellises, posts, and wire.

Apple Systems:  With the advent of more dwarfing rootstocks, the apple industry went through a revolution in training systems.  Many of the old standard trees on seedling rootstocks, planted at 40x40 feet were pushed out in the 1970's and were replanted into semi-dwarf systems.  In the 80s and 90s, fully dwarf tree systems started to supplant semi-dwarf systems in many of the intense apple growing regions.  Now you cannot find an apple tree on standard seedling root at any of the commercial nurseries, and there are very few in most successful commercial orchards.  In the 2000's the trend for commercial apple production is definitely toward high density systems of between 600-1000 trees per acre, with some systems going well beyond that density.
    Self-supporting semi-dwarf systems:  Usually planted at distances of 10x16 feet to 16x24 feet.  Densities of 115 to 275 trees per acre, with a more normal density in the range of 180-200 trees per acre.  Typically on EMLA 7, 106, and 111 size rootstocks.  For the most part all of these trees are trained to a "central leader" system, where a main leader is allowed to grow up the center of the tree, and 2-4 swirls of permanent scaffold limbs form a tree looking somewhat like a pyramid.  Easy to learn, easy to train, easy to maintain, and quite productive when mature.  This is a good system for new fruit growers to start out with.  The most slow to come into production initially, but the most "forgiving" for novices.
    Single pole systems:  Usually planted at distances of 4x12 feet to 10x18 feet.  Densities of 250 to 900 trees per acre, with more normal densities in the 300-400 trees per acre range.  Typically on EMLA 26 and EMLA 7 size rootstocks.  Trained to a modified central leader system, depending on the density.  More dense trees have less permanent swirls of scaffolds, although usually the lowest one is permanent.  Higher density systems have limbs which are renewed more often and less vigorous, smaller wood is the target.  It is fairly easy to learn these systems, and they can be very successful for the new fruit grower.  Support is usually by a single wooden post, ranging from 2-4", or metal conduit or angle iron stake strong enough to help support the central leader and keep the tree from tipping over in the wind or under crop load.  Usually quite early to come into good production.
    Trellis systems:  Usually planted at distances of 2x10 feet to 6x14 feet.  Densities of 500 to 2200 or more trees per acre, with more normal densities in the 600-900 trees per acre range.  Almost always planted on Malling 9 size rootstocks, since full dwarfing is required to keep these systems in-bounds and productive over their lifetimes.  There are still many different training systems, with many different names, but their main characteristic is usually that a single leader is trained up the trellis as high as it will practically go and fruiting wood is allowed to grow off the main leader, being renewed on a regular basis.  Almost no permanent scaffold limbs in most systems are allowed.  These are very productive systems which bear early and heavily, but require intense, detail management by a skilled and knowledgeable fruit grower.  These types of systems are becoming more the norm in commercial fruit growing, and your may want to "graduate" to them after getting your feet wet.  Training "goofs" on these systems tend to haunt you for the life of the system.  The trellis are usually constructed with major end posts on each end of the row (6-8" min) with regularly spaced support posts about every 30-40 feet (4-6" min).  Any where from one high wire at the top to as many as 6-10 wires in some systems.  Some systems form a "V" shape, some a "T" shape with the posts and wires, but it seems to most successful system usually is a single plane of posts with 2-4 wires.  There may be bamboo or metal conduits used to help support individual trees.  These systems are costly to install, but usually produce lots of high quality fruit early in their life span.  Interestingly enough their maximum production levels can be quite is similar to all the other systems.  In many of the large apple production regions, growers are looking to the future and designing these high density systems with the prospect of using self-propelled "platforms" that employees can work from for pruning, thinning, and harvesting.  Mechanized harvesting is still years away, but is on the horizon.

Pear Systems:  For the most part pears are planted in self-supporting systems very similar to apples in the self-supporting systems above.  There are very few fully dwarf pear rootstocks that have proven themselves to be good.  Pears in general are less vigorous than apple trees, so even pears on standard seedling roots and on the "semi-dwarf" pear roots will grow to about the same size as most semi-dwarf apples.  Most training systems are central-leader based with permanent limbs.

Peaches and Nectarines:  There are really no commercially utilized dwarf or semi-dwarf rootstocks for peaches and nectarines at this time, although there are some new rootstocks being tested that may have potential.  Most trees are grown on seedling rootstocks and can be kept in size ranges close to semi-dwarf apples.  Most training systems for commercial growers are based on an "open center" systems where there are 2-5 main scaffold limbs growing into a vase shape.  Some commercial growers try to utilize central-leader or fan shapes, but often this is working against the normal architecture of the tree.  

Plums, Prunes, and Apricots:  There are no commercially utilized dwarf or semi-dwarf rootstocks to choose from.  Almost all trees are grown on seedling rootstocks of some kind, although some of the same new dwarfing rootstocks being tested for peach may work with other stone fruits too.  Plums, prunes and apricots can be trained very similar to peaches, however, some varieties and cultivars have upright characteristics so that fan or leader systems may prove useful.  The training approach should be flexible based on the variety's growth habit.  Planting distances and densities are very similar to peaches in the 150-250 trees per acre range. 

Cherries:  With the development of "dwarf" rootstocks, sweet cherry culture, in particular, has been going through a revolution.  I say "dwarf", because most of the newer rootstocks, such as Gisela® are still quite new and growers are still learning how much dwarfing character they really impart, and we are finding it may vary according to region.  Also, the variety / rootstock combination seems to take on more importance on cherry than on apple.  Apple rootstocks perform more consistently across the many varieties.  There is still a big learning curve for most commercial growers, but as researchers, extension agents, and other growers work with the new dwarf cherry roots, the more the knowledge base is expanding.  In general, there are three ways to go with on cherry planting systems at this time:
    Standard Mazzard or Mahaleb rootstocks:  In about 10-15 years, these trees will pretty much grow to their full potential size.  Mazzard rootstock is the most vigorous and is mainly used on sweet cherries.  Trees can grow into the 20-30 foot height range if left unchecked.  In well trained systems, they can be maintained in the 15-20 feet high range, and about as much in spread.  The main drawback for cherries on Mazzard root is that they are slow to come into bearing--- 4-7 years.  Mahaleb is most often used on tart cherries and some new "clones" of Mahaleb, like CT 500, can be used for sweet cherries.  Mahaleb is about 75-90% of Mazzard size and is often called "semi-dwarf", simply because it is smaller than Mazzard.  Mahaleb should only be used in well drained soils.  Depending on the variety of cherry, central leader or open center is used.  Usually the variety will lend itself better one way of the other.  Many sweet cherries tend to naturally grow in a central leader type way.  Tart cherries tend to be a little more spreading and bushy.  Planting distances can range from 14x20 feet to 16x24 feet.  Densities of 100-155 trees per acre.  Plant tart cherries at the higher density and sweet cherries at the lower one in most cases.
     Gisela® 6 and 12, Maxma(R) and Krymsk Rootstocks are less vigorous than Mazzard, but are much more precocious.  They will start bearing heavily after just a few years.  Whereas it seemed to take a lifetime to get sweet cherries on Mazzard root into production, on these newer precocious roots it seems like a snap.  The training systems are still being developed for most of the new dwarf cherry roots, varying from central leader type systems, to fan types, to open center systems, depending on the grower and his philosophy of growing.  The one fact that is becoming evident is that proper annual trimming, based on the fruitfulness of the particular variety, is necessary to avoid over-cropping and small sized fruit on many dwarf roots.  Within the next couple years, there will be several well researched training systems published and offered as guides to growers.  I recommend Gisela® 6 and 12 to even new fruit growers, because they can bring you into profitable production within a few years, are somewhat more forgiving than Gisela® 5- full dwarf root, and are less frustrating than standard rootstocks for many sweet cherry varieties.
    Gisela® 5 rootstock is the most dwarfing of the new cherry roots.  At one time it was thought to be the new "Malling 9" of cherry roots, which would allow pedestrian orchards to take over from the "forests" which Mazzard root trees used to form.  Gisela® 5 has its place with less productive and/or highly vigorous varieties.  It is very precocious and will start bearing in a couple years.  Under poor management, it can set huge crops of small fruit, so regular and proper annual trimming is required to maintain consistent production of large fruit.  This rootstock is not for the inexperienced new grower who is not willing to learn its ins and outs.