Often, especially in lowland, agricultural country, the limits to a site are clearly defined on the map and on the ground by sharp, man-made boundaries which give a division between ground of high scientific interest and that with low or negligible scientific value. This is perhaps most obviously true of woodlands surrounded by farmland, but applies to most other formations. Such boundaries, defined by hedges, fences, walls or ditches are the obvious limits to draw for many key sites. Sometimes, however, there are no ready-made boundaries of this kind, especially in uplands, where areas of high scientific interest merge imperceptibly into less important ground.

In these cases, boundaries must usually be drawn arbitrarily though they may follow natural features such as streams, or artificial ones such as walls, fences, and roads. The principle followed here, though very largely subjective, is to define an area which is large enough to include the full range of diversity shown by the formation in the particular locality, but not so large as to give an unnecessarily extensive representation of certain features. This principle has some analogy with the concept of 'minimal area' in the plant community, i.e. the smallest area in which a particular community attains its full number of constant and characteristic species.

Sometimes it happens that an area is not of uniform quality, but the omission of the less valuable parts would create an unsatisfactory boundary, and perhaps cause later problems of management and protection. In such instances, areas which appear as natural units are often best treated as such in defining key sites. For some sites it is desirable to have a marginal or adjoining buffer zone to protect a vulnerable ecosystem, e.g. the hydrology of a valley or basin mire is particularly affected by changes in the catchment, and a measure of control over the latter is desirable. Similarly, it may not be possible to protect mobile creatures, notably birds, simply by safeguarding the place where they breed, e.g. species such as the red kite and osprey depend for their food supply on quite a large area beyond the confines of their nesting woods.

Where there is actual geographical continuity between high- quality examples of different formations, a boundary is drawn to include these within a single site. A single geographical unit containing examples of more than one major formation is known as a composite site. Where examples of the same formation lie close together but are separated by different land (especially farmland) of much lower quality, it is convenient to regard these as a single geographical and ecological unit, termed an aggregate site. Decisions on the particular clusters of sites which should be so treated are necessarily arbitrary. Conversely, within certain sites (mainly upland) which form single, continuous units, some parts are much less valuable than others, or certain vegetation types may be unnecessarily extensive. The implementation stage of the Review should consider how best to deal with these situations.

One of the disadvantages of carrying out the Nature Conservation Review according to subdivision into major formations is that geographical entities are often broken up and artificially compartmentalised, instead of being treated as integrated ecological units. As a result, the particular importance given by their diversity does not always emerge clearly from the separate assessments and descriptions. This applies especially to the New Forest, which is not simply an aggregation of important woodlands, acidic heaths and valley mires, but gains extra status as the largest continuous area of undeveloped land remaining in the whole of lowland England, i.e. its total value is greater than the sum of the individual parts. Some areas which once had this quality, e.g. Breckland and the Isle of Purbeck, have partly lost it through recent dissection and depletion by human activities, though their importance is still outstanding. In certain other districts, there are unusual concentrations of important sites, often covering a variety of formations, and these should be regarded as having special importance to nature conservation on this large scale.