Growing Broadleaved Trees in Mixture with Conifers


Wessex Silviculturai Group



A number of mixed stands of various ages were visited in south-western England. These stands are discussed subjectively and tentative conclusions put forward. The paper is a record of the observations of practical men in the forest rather than the result of a detailed investigation.


This is the second study by a group of practical foresters and forest owners. The group endeavours to enlarge its knowledge of particular problems by concentrated observation and discussion within the region. This method is necessarily subjective and unlikely to produce conclusive results, but con¬tinuous attention to a particular subject in the forest can give useful pointers and correct misunderstandings.


Over most of England, pure conifer stands appear, at present, to be the most profitable type of forest. Pure stands of oak and beech are no longer economic to establish by themselves, as they no longer produce an adequate return from early thinnings. Oak and beech form a very large part of our present-day landscape, and many woodland owners wish, both for amenity and for good game management, to establish crops which may eventually form stands that are either mixed or purely broadleaved.

Although most sites will grow conifers more economically than broad-leaved trees, there are, particularly in southern England, large areas of shal¬low, calcareous downland soils which will not support the growth of conifers of large size. If these sites are to be afforested, then the final crop is likely to be limited to beech or sycamore. There is, therefore, considerable scope for the creation of broad-leaved woodlands. It would seem that, both for financial and silvicultural reasons, these woodlands are best established in mixtures with conifers. The problem is how to select the correct broadleaved and conifer species for the site, and how to manage them so that the best possible final crop is produced by the most financially satisfactory method.


Fourteen forests in the region were visited over a period of eighteen months. Any relevant stands were observed and discussed. The majority of stands ranged from 1 to 30 years in age. Few stands were seen in the later stages of development, and most of the older stands seen had already lost their conifer element. Few measurements were made, and the discussion was concerned mainly with past history, present condition, and probable future develop¬ment. These discussions were roughly recorded, and form the basis of this paper. The forests visited were as follows:


Name Owner County Main Soil Material
Cholderton Capt. L. Edmunds Wiltshire Very shallow drifts on chalk
Cranborne Chase Forestry Commission Wilts./Dorset Very shallow drifts on chalk
Pentridge Viscount Cranborne DorsetVery shallow drifts on chalk
Crichel The Hon. Mrs. M. A. Marten Hants/Dorset Variable drifts on chalk
Salisbury Plain Ministry of Defence Wiltshire Variable drifts on chalk
Springhead H. R. Gardiner, Esq. Dorset Variable drifts on chalk
Micheldever Forest Forestry Commission Hants Variable drifts on chalk
Delcombe G. R. Rickman, Esq. Dorset PP Variable drifts on chalk
Cirencester Park Earl Bathurst Gloucestershire Shallow calcareous loams
Woolland D. S. Hughes, Esq. Dorset Calcareous Vale clays
Garnons Sir Richard Cotterell, Bart. Hereford Deep old red sandstone looms
Dymock Forest Forestry Commission Glos./Hereford Deep old red sandstone looms
New Forest Forestry Commission Hants Acid sands and clays


Conifers planted in mixture with broadleaved trees are frequently said to act as 'nurses'. They usually get off to a faster start and give a certain amount of protection to the broadleaved species. This is often especially valuable on difficult, exposed, or frosty sites. Their faster rate of growth may also improve both the form and height growth of the broadleaved species, and may help to suppress herbaceous weeds quicker. Later on they will suppress the broadleaves unless carefully managed.

It quickly became clear that although there were some sites where conifers were necessary to help the broadleaved trees to get established, in the majority of cases the conifers were planted for mainly financial reasons. The conifers provide both a bigger volume of timber per acre, and also earlier saleable thinnings.

The sites visited can be roughly grouped into three main types:

(a) Reasonably fertile, sheltered sites where pure conifers would be much more profitable, and broadleaved trees are required for amenity, game management, or to produce high quality timber quickly, e.g. Garnons, Dymock, Cirencester.

(b) Shallow calcareous soils where beech and sycamore are the trees most likely to attain large size, and where most conifers are short lived due to death from chlorosis — the high alkalinity of the soil induces iron deficiency, e.g. Cholderton, Cranborne Chase.

(c) Exposed, frosty, or heavy-clay sites where broadleaved trees are slow to establish themselves without the help of conifers, e.g. Woolland, New Forest.

In the region studied, the first type of site is the most common, although the Hampshire Downs, the Salisbury Plain, and Cranborne Chase provide large areas of the second type of site. Many of these downland sites are also exposed and difficult to establish, as are the deeper upland soils. Frosty sites may be high or low, depending on the topography, and some heavy valley sites can be very difficult.

Whatever the site, the primary problem is to achieve a satisfactory final crop of broadleaved trees whilst obtaining the maximum possible intermediate returns. This is usually done by selecting one or more suitable conifers to grow with the final broadleaved species. Whenever these mixtures are planted, the main problems are:

1. Compatibility of the species in the mixture.

2. The method of mixing the species.

3. The management of the stand.


In all the stands visited, the success of the mixtures was largely dependent on the degree of compatibility between the broadleaved trees and the chosen conifers. Species which are compatible on one site may not be compatible on another, and species compatible under one type of management may not be compatible under another. To complicate matters further, species may be compatible with each other only when particular provenances are used.

The main criterion for a compatible conifer is that it should have a height growth comparable with but slightly faster than the broadleaved species on a given site. It should also be reasonably columnar in crown shape as well as being a good volume producer, and should be saleable in its early stages of growth.

European conifers, such as larch, Scots pine, and Norway spruce, were, of course, most used in the past and it is probable that many present-day mature oak and beech woods were originally grown in mixture with them. They were seen to be compatible with beech on several sites notably Cranborne Chase, Salisbury Plain, and Cholderton. In the New Forest, however, a vigorously branched provenance of Scots pine threatened to suppress completely oak on poorish sites, while at Garnons on rich loams very suited to oak the wide branching of pine gave less scope to oak than did Norway spruce. Norway spruce seemed to be the most compatible tree with oak and good examples were seen at Woolland, Crichel, and Garnons. European larch, which starts growth quickly, tended to overtop and oppress oak on the more fertile parts of a Crichel plantation, but as the site became poorer, larch was at something of a disadvantage and oak was holding its own. On the richest sites, at Dymock and Garnons, larch tended to outgrow oak. At Garnons it has been used successfully in growing fine ash, but has to be removed early in the rotation to allow the ash free development.

Beech, being more tolerant of competition and of site, suffers less from overgrowth in mixture with European larch, Scots pine, and Norway spruce. It can also recover more readily from partial suppression, especially by larch. Good spruce-beech mixtures in the pole stage were seen at Cranborne Chase. Beech had grown up very well in 40-year-old stands of larch at Dymock, and with Scots pine and Norway spruce at Cholderton.

Of other conifers, Corsican pine tended to swamp oak by lateral branch vigour at Crichel and to overgrow beech at Cranborne Chase. Japanese larch was too fast growing for oak at Woolland on clay, but beech had grown up well with it at Dymock on heavy loans. The vigorous branching and height growth of these two species make them less easy companions for broadleaved species. Douglas fir had been fairly widely used with beech on the lams over chalk, and on these soils its growth is more comparable with beech than on better sites, but its height growth and vigorous branching still presents a strong threat to the beech and constant vigilance is needed to prevent over-topping.

Lawson cypress and western red cedar are likely to prove compatible species with broadleaves, their height growth is similar and their soft columnar branching is favourable. Although no reasonably sized examples of their mixtures were seen, there is a good example of an oak-red cedar mixture at Wytham Wood, near Oxford. Where the incompatible Douglas fir is used with oak and beech, Lawson cypress and red cedar would seem to be useful as buffers.


Compatibility problems can be met by varying the pattern of planting, and the type of mixture will substantially affect the reaction between the broad-leaved and the conifer species. The main methods of mixture are:

(a) Lines and bands.

(b) Groups.

(c) Underplanting.

(d) Intercropping.

(e) Lines and Bands

The simplest mixture is an alternate row mixture. This is the traditional type of mixture, and under good management it can be very successful, but disastrous under neglect, or when the wrong species have been chosen. Unless the species are very compatible, the broadleaved trees may be quickly sup-pressed. This could be seen at the New Forest with the oak-Scots pine mix¬ture, where the Scots pine was of a poor, rough provenance (and the soil was marginal for oak); and at Cranborne Chase with Scots pine-beech, Corsican pine-beech and Norway spruce-beech mixtures. These mixtures were often most successful when chlorosis had reduced the vigour of the pines, or where the Norway spruce had been removed early for Christmas trees. Generally speaking, line by line mixtures contain more broadleaved trees than are necessary to form a final crop, and leave little margin for error in the choice of compatible species. They may, however, on the most difficult sites, give the maximum protection to the broadleaved species, and local experience at Micheldever and Cranborne bore this out.

Band mixtures arose from line mixtures in an endeavour to reduce the amount of tending needed and to avoid the suppression or loss of broadleaves caused by incompatible mixtures or delayed treatment. Some of the earlier mixtures had a high proportion of broadleaved trees. In the New Forest there were several examples of four row oak—one row European larch, and four row oak—one row Scots pine mixtures. These have given rise to reasonable oak crops, particularly where the growth of the larch was poor, or where the larch was removed early for G.P.O. telegraph poles. These mixtures, however, contain far too large a proportion of broadleaved trees to be economic. At Dymock, five row oak—two row Norway spruce, and ten row oak—three row Norway spruce mixtures have given rise to promising crops, but the financial returns are again low. At Woolland, a one row oak—two row Norway spruce mixture was very successful in places, but, although the proportion of conifers is high, the oak is rather too scattered to give an efficient choice of final crop trees. Three row—three row mixtures of Douglas fir and beech where conditions were not very favourable to the conifer looked at 15 years as though the beech would make timber provided that the largest Douglas fir were continually removed in the early thinnings. At Cranborne Chase a similar band mixture had a 10-foot-wide row of hazel between beech and Douglas fir which gave management more scope in the pole stage, though here, too, heavy thinning of predominant Douglas fir will be needed, but at a later date.

Current band mixtures are related to the final crop spacing likely to be required; for oak and beech this is likely to be between 3o and 40 feet. At 5-foot spacing, this would permit three row—three row, three row—four row or three row—five row broadleaf—conifer mixtures, and at 6-foot spacing three row—three row, or possibly three row—four fow broadleaf—conifer mixtures could be arranged. The customary width for the broadleaved species band is three rows. Two row bands are not very successful as there is a tendency for both the rows to be suppressed if thinnings are delayed, whereas with a three row mixture there is a reasonable chance that the centre row will survive a con-siderable amount of neglect. The commonest mixture is the three row—three row one, and several good examples of this type were seen, particularly the oak—Norway spruce mixture at Crichel and the beech—Douglas fir mixture at Micheldever. The ii-year-old Norway spruce—oak mixture looked very well, but the height growth of the Norway spruce was not so greatly different from that of the oak, and so there did not seem to be much 'nursing' effect. There may, however, be a more noticeable 'nursing' effect later on. With the beech—Douglas fir mixture, the height growth difference was greater, so that although there was some 'nursing' effect, there was also a danger that the Douglas fir would suppress the beech unless it was thinned early and vigorously. In both these examples, the proportion of hardwoods is still high. It would appear from some very young plantations that this proportion can be reduced still further by putting alternate narrow crowned conifers such as Lawson cypress or red cedar in the outside beech rows. This reduces the percentage of hard¬woods from 50 per cent. to 33 1⁄3 per cent. An alternative method on good sites might be to plant a three row broadleaf—five row conifer mixture, with a narrow crowned conifer in the outer rows and Douglas fir in the middle.

(b) Groups

A rather different approach is to plant groups of a broadleaved species in a conifer matrix, the usual practice being to plant the groups at a final crop spacing of 30 to 40 feet. At Dymock, moderately successful groups of oak were seen in a larch matrix. The crop was 40 years old, and the oak, which were in groups spaced 16 feet apart, centre to centre, were obviously too close together. They were also very drawn-up due to under thinning in the past. Beech groups of sixteen trees at 30 to 40-foot centres in European and Japanese larch have succeeded well here.

The smallest size of a group is one tree, but this presents too much of a risk, partly because of the unknown genetic make-up of the transplants and partly because of the numerous natural hazards, such as physical damage by frost, and biological damage from browsing and fungal disease. Improved selection may enable better forms of tree to be grown, and thus reduce the numbers necessary, but the natural hazards will remain, and so the size of the group is unlikely to drop below four or five trees.

At present the most satisfactory size of group would appear to be nine, in a straightforward block of three by three at a similar spacing to the rest of the crop. Groups with reduced spacing have been tried, and the most common one is a group of thirteen formed by planting a quincunx at say 6-foot spacing, and then plant¬ing eight more trees in between these, at 3-foot spacing. This gives a larger selection without taking up more space, and might well allow for beneficial competition between the broadleaved trees themselves. These advantages are unfortunately offset by the tendency for the intermediate trees to be cut down during weeding.

Good examples of oak groups were seen in both European and Japanese larch matrices at Crichel. In the case of the oak—Japanese larch mixture, it was clear that early and heavy intervention was necessary to prevent the oak from being too suppressed. The European larch mixture seemed slightly easier to control, and the mixture seemed better silviculturally, though not economically, on the poorer site types, where the growth differences were not so marked.

The group method gives the best theoretical mixtures, and allows the minimum practical number of broadleaved trees to be used at the most eco-nomical distribution. Unfortunately, it is statistically improbable that the central tree will be the best tree, and the final spacing will inevitably differ considerably from the theoretical ideal. Group mixtures demand skilled and intensive management. Their planting requires more skill, especially if the groups are at a different spacing from the matrix, and subsequent weeding is more difficult. There is a tendency to forget about the broadleaved groups once the conifers are out of the weeding stage, and the broadleaves may become badly suppressed by coppice regrowth or climbing weeds. On the other hand, group planting makes the selection of final crop trees easier than in band planting.

A simple and very effective group mixture can be achieved by planting rows of conifers alternately with repeating blocks of conifers and broadleaved trees. This gives rise to a 'tartan' pattern, which in its simplest form consists of a band of three conifers, alternating with blocks of nine broadleaves and nine conifers. The broadleaved tree percentage is thus 25 per cent. If the conifer bands were increased to five rows, and the blocks were of nine broad-leaved trees and fifteen conifers, then the proportion would be reduced still further to 14 per cent., which is about as low as one can go without running too much risk of not obtaining a satisfactory final crop.

(c) Underplanting

On some very difficult sites, it may be that the best method of establishing a broadleaved crop is to establish a conifer crop first and then to underplant once forest conditions have been obtained. Numerous examples of this were seen at Springhead, where beech and sycamore were being extensively intro-duced under first generation conifer crops on chalk downland. On this estate, the groups, which were quite small, were being seriously damaged by deer and rodents. Other problems were those of removing sufficient overhead cover to allow enough light for the satisfactory growth of the underplanted broad-leaved species, and extracting produce without damage.

Small groups with overhead shade are very difficult to manage. It would seem easier to make large groups of a quarter of an acre or more and to re-move all the overcrop from the group, using side rather than overhead pro¬tection. Extraction spaces should be allowed for between the groups. The best time to introduce the broadleaved species into the pioneer crop is probably during the early pole stage, after, say, the first or second thinning.

(d) Intercropping

In some cases it may be possible to take more than one conifer crop off the site during one broadleaf rotation. At Garnons, oaks were originally planted in a Norway spruce matrix in lines at final crop spacing. The Norway spruce were then removed as Christmas trees, and the ground replanted with conifer—broadleaf line mixtures of faster growth. This led to numerous management problems, and was discarded as being too difficult. The current method, which is simple to manage, is to take the early return from densely planted Norway spruce in a normal band mixture, spaced to provide final crop trees.

On a good site where the broadleaved species are capable of growing strongly for over a hundred years it may be possible to take two crops of conifer sawmill timber during the life of the broadleaved crop. If a wide spacing of broadleaves is used, with a final crop spacing of over 4o feet, then it will be necessary to grow a second conifer crop in order to use the site efficiently.


The most economical mixtures occur when the site allows the use of a highly productive conifer. Douglas fir is a good example, but this type of mixture is likely to suffer if neglected, and needs careful management. Mixtures with Japanese larch or western hemlock are also liable to suffer from neglect, and incorrect early thinnings can quickly spoil the broadleaved crop. It is probable that a policy of removing the largest conifers first will be the most satisfactory one when rather incompatible species are grown together. Removal of conifer side branches may help to postpone thinnings, but mixtures with fast-growing conifers tend to be inflexible, and thinnings may have to be carried out at an inconvenient time.

In terms of money returns, at least 30 per cent, of the more productive conifer yield is lost to the desired broadleaf whose early thinnings are less in bulk and rarely equal in price. The conifer matrix is generally felled on a rotation shorter than the optimum in terms of volume and profitability. As the broadleaf is generally 'at risk' to the conifer and the greater proportion of conifer the greater the risk of losing the broadleaf, the thinnings must be made urgently. All this points to the use of a readily marketable conifer which should give a high yield per acre and yield useful by-products. Of the most generally compatible conifers, Norway spruce, Lawson cypress, and western red cedar are all good volume producers and in addition yield Christmas trees from tops or foliage from side branches. They combine silvicultural attractiveness and economic desirability.


It would only be possible to suggest detailed mixtures of species if all the relevant details such as objects of management, site type, the provenance of the species, and their reaction to the site were known in some detail. However, certain broad principles do seem to have arisen from this study, and these are:

1. That groups of nine broadleaved trees in a conifer matrix seem to be the best method of establishment, and that these groups are best distributed in a tartan pattern.

2. That the most productive conifer that the site is capable of supporting should be grown in the matrix, and that if this conifer is not compatible in height growth or branching habit with the broadleaved tree required, then a buffer of a narrow crowned conifer should be placed around the group, or heavy early thinnings carried out.

3. That over a very wide range of conditions, Norway spruce is a most compatible species with both oak and beech. It provides a flexible, easily managed mixture, and the thinnings are readily utilizable at all stages.

4. The simplicity of a mixture of two species may well be better in the long run than a higher yielding but more complicated mixture of three or more species.

5. Except where shallow calcareous soils limit the life of conifers, only enough groups to give some forty final stand broadleaves per acre are needed. This is a spacing of about 35 feet from centre to centre. Suitably placed conifers can be grown on to replace groups that fail. Final spacings of more than 35 feet should be considered only on good sites well capable of supporting broadleaves through two conifer rotations, or where a mixed final stand is required.

6. On very shallow soils that are incapable of sustaining conifers to more than half a rotation of from 80 to 100 years, line mixtures or narrow bands are preferable to groups.

7. That under very difficult conditions it may be better to establish a pure conifer crop first, and then to introduce groups with broadleaves in the pole stage.


The writers wish to thank all owners for permission to visit their woods, and all members of the Group for their observations and help in compiling the report. In particular they would like to thank M. J. Penistan for his unfailing encouragement and advice.