This page describes the design and construction of our kitchen garden project so far. Although I had no idea I would be writing this blog, I did use my mobile phone to keep a somewhat patchy photographic record of the project from the beginning. Now I am glad I did, as it allows me to chronicle the development of the garden here, and discuss some of the design issues that might be of interest to anyone considering such a project themselves. If you find yourself yawning on this page, you can get a quick virtual tour of the garden instead by visiting the gallery. You can also download a scale plan of the garden here: Garden Plan.

There is a lot of stuff on this page; you can use the links below to navigate:

Design influences

Design considerations

Breaking ground

Heavy construction

Glasshouses, coldframes, and polytunnel

Fruitcage and other fruit

Design influences

Whilst a garden, productive or decorative, can be constructed in an ad hoc manner, when there is considerable construction effort – walls, glasshouses, beds, etc.,a detailed planning process is to be recommended. The design considerations that are to be brought to bear on the planning process depend to a great deal on the designer’s philosophy and the sort of garden being laid out. I am an old fashioned sort of gardener looking for a highly cultivated space. Whilst I really appreciate the aesthetics of the cottage garden and the informal potager, indeed, in many ways find nothing more attractive, I am nonetheless more inclined to lay out a well ordered plan – an engineer’s garden – with every element carefully placed exactly where I want it to go. This is quite a different perspective from someone favouring a more informal layout or considering quite different philosophies of cultivation, such as permaculture. Thus, the designer’s initial approach will scope the design considerations.

If resources were no object, my ideal kitchen garden would replicate, on a much more modest scale, a typical walled kitchen garden of the 1800s. I have had a fascination with these gardens for many years and have read many old texts from the period, finding that they have much to offer even modern gardeners. Indeed, the methods of managing soil fertility have much in common with organic gardeners of today. One can go back sufficiently in time to a period before the widespread application of synthetic fertilisers and pesticides. I remember most fondly the BBC television series The Victorian Kitchen Garden, 1989, and the two further series that followed, and have the boxed set on DVD, along with Harry Dodson’s Practical Kitchen Garden: A Personal Guide to Growing Vegetables and Fruit. There was something very special about this programme and, in my opinion, no television gardening programme has come close since. This was also a very sad programme in many ways; Harry, and the knowledge of the old ways that he represented, walled gardens in general, as well as that specific garden, had all but passed away, and I cannot watch it without feeling a profound sense of loss. I like the old ways, the old varieties, and the old gardens. Nonetheless, the resources required to construct and maintain even a small scale version of this kind of garden are far beyond our backyard enterprise. Thus, although I take some influence from these old gardens, I cannot replicate them and must consider various cost and time saving design elements.

Ideas for garden design can be readily gleaned from visiting some of the great gardens, large and small, that are open for public viewing. One of my favourites in our local area is the well restored walled garden at West Dean, near Chichester. Bookshops and garden centres are awash with gardening books and a great many on the cultivation of fruit and vegetables. I have quite a collection of such books, but find that many of the offerings, however well presented, are too similar in content. It is worth spending some time looking for books that offer something different from the usual allotment or kitchen garden formula. There are some that offer useful insights for the prospective organic gardener, some that include more detailed information on soil conditions and climate, others still that offer guides to garden design; all of these are helpful reading before embarking on the development of a new kitchen garden. Just flicking through a good book can provide a great deal of inspiration. You can find a list of some of my favourite books here.


Design considerations

Gardening does not need to be complicated. If you throw some seed at some soil there is every chance that nature will do what it was designed to do and things will grow. If you provide for the basic needs of fruit and vegetable crops – sun, fertile soil, and water in good measure – good things will be sure to happen. A great deal of success can be had simply by digging a few beds and planting ad hoc with whatever crops you favour. However, if considering the construction of a larger kitchen garden, or wondering how to maximise the cropping potential of a smaller space, it might be helpful to plan the garden with some attention to detail. Some details, such as crop rotation, may become quite important in managing pests and diseases without resorting to harmful chemicals, but often they are marginal things, such as whether it might be better to plant a choice pear agaist a west or east facing wall, and it is up to the interests of the individual gardener how far they want to go into these matters. Things will often grow quite happily with but a modicum of care.

Site survey

Before planning begins, a careful survey of the site is to be recommended. A great many factors bear on the layout of a kitchen garden. Amongst others, these might include: size, orientation, and lay of the land; location of any trees or large hedging that may limit light, nutrients and water; protection from prevailing weather; possibility of frost pockets; soil properties; storage for tools and materials; storage for winter produce.

We are fortunate in that our site is near ideal for many of these factors. A south facing aspect is to be preferred, and our plot faces approximately 10 degrees west of south. There is a slight slope to the south and a more pronounced slope to the west, neither of which are sufficient to cause any difficulties in the construction work. A more steep slope to the south would, in fact, be beneficial, as it improves the angle of the plot to the sun, increasing the direct solar radiation and having the effect of shifting the microclimate south; a substantial slope can have a remarkable effect. Most fruit and vegetables are best grown in full sun, although there are crops that can be grown in partial shade. There is generally little shading of the site. On the west side of the plot, a tall privet hedge does result in some late afternoon shade, but not so much as to cause any concern. If we were to grow up to the hedge, this would be problematical, in terms of light, nutrients and water. However, the plot is wide enough to position the western boundary a good distance from the hedge.

In terms of size, the plot is much larger than we need, leaving plenty of space for future projects; for example, we are planning to add a traditional orchard at a later date. It is ideal if the largest dimension runs east to west, rather than north to south, so as to maximise exposure to the sun. In our case, the plot is much longer north to south. However, it is sufficiently large that a reasonable south facing aspect is attainable.

There are large hedgerows and one large oak tree near the plot. The latter is sufficiently distant not to cause any concern. The hedgerows were dealt with by creating the kitchen garden inside the existing boundaries, allowing the hedgerows to remain and flourish without interfering with the kitchen garden. The site is reasonably well protected from the weather: to the north lies the house, trees and the wall of the kitchen garden that we constructed to exploit the southerly aspect; to the east is the garage block and outbuildings and, beyond that, a substantial hedgerow; to the west is a tall and substantial privet hedge; and to the south, a somewhat patchy hedgerow and, just beyond our plot, a row of tall poplars. In all, this offers a good degree of protection to the kitchen garden. A more open site might benefit from some protective plantings or the use of windbreak netting. The outbuildings, four in total, were rather dillapidated attachements to the rear of the garage. Renovated, however, they are ideal to provide a potting shed, storage for harvested fruit and vegetables, and even a forcing shed. 

The soil was our greatest concern when we began to explore the possibilities, and several test holes were dug in various locations. Our home is located on the western edge of the New Forest, just outside of the National Park boundary. Soils within the New Forest are generally acidic and low in fertility. However, we are also influenced by our location near to the river Avon, and fall within a band of much improved fertility offered by the valley. Tests on the soil revealed a composition that is sandy, with a great deal of silt, and a very small amount of clay. Such a soil is reasonably free draining, readily warms up in the spring, and, with the addition of manures and composts, has good potential for cultivating fruits and vegetables. I prefer the acidic soil to an alkaline soil as I feel it is more flexible. Soft fruit enjoys some acidity, especially blueberries, which demand it. With the exception of brassicas, most vegetables will grow quite happily on such a soil, and some, such as potatoes, benefit from it. Brassicas tend to prefer a more alkaline soil. In particular, they often suffer from clubroot, which is much less troublesome on alkaline soils. This can be readily remedied by the application of lime. So far we have not limed the soil and they have grown strong and healthily for us, but at some point we will no doubt need to apply some lime, as repeated applications of manures and composts tend to acidify the soil further.

Although satisfied with the general properties of the soil, and observing from the test holes that we had a surprisingly good depth of top soil, our plot is all but surrounded by lakes that were former gravel pits. Thus, we had concerns that the soil may be excessively stony. Some amount of small stones is needed in the soil, but in excess and in large sizes, they can be detrimental to growing, especially root crops. Although it varied across the site, it was quickly realised that in some areas the soil was very stony indeed. There are three options to deal with this: do nothing about it, and accept that some crops will not grow so well; build deep raised beds filled with bought in top soil, manures and composts; or sieve the soil to remove the large stones. Only a fool would consider sieving the soil on such a large extent, but we were not keen on the cost of constructing deep raised beds and buying in the soil to fill them, so that is what we did. Without the benefit of heavy machinery, it was very labour intensive, and many tons of soil had to be processed, but the end result is very satisfactory. Over time we might pick up further stones from the subsoil, but the largest of these were removed, so the beds should remain in good condition. With repeated application of composts and manures they will also gain somewhat in depth.

All things considered, the general prospects for the site were very good indeed, with many factors close to ideal.


From various decorative plantings made around the house, we quickly realised that some form of enclosure would be essential. The plot was home to a great many rabbits and frequented by deer. Between them they can reap great destruction, and there would have been little point attempting to cultivate anything without offering some protection. Height was needed to exclude the deer and a wall of mesh dug a couple of feet into the ground to exclude the rabbits. I would love to have built an English bond wall of red brick, but on such a scale the material and labour costs are entirely prohibitive, so a more modest approach was sought. The south facing aspect is the most important, and it is here that a wall can offer an incredible difference in the microclimate. We already had, from the garage block, some west facing walls, which helps further with the microclimate. A 2 meter south facing wall was therefore constructed, out of 6″ block and render. This is much cheaper than brick, and also happens to match with the garage block. The rest of the plot was fenced to a height of 2 meters, with mesh dug 2′ into the ground to prevent rabbits from burrowing underneath. The height was a compromise between cost, visual impact, and avoiding the need to seek planning permission. Ideally, the wall would be rather higher, but 2 meters is just enough to allow fruit to be trained against the wall. The fence forms the west and south of the enclosure, and part of the east, where it then meets up with the outbuildings; the garage and the outbuildings forming the remainder of the eastern boundary. Not quite a walled kitchen garden, but a good compromise that we are very happy with. The effect of the walls is certainly readily observed on even a moderately a sunny day.


A good harvest may be obtained without the use of covered growing areas. However, any amount of cover, whether glass or plastic, can help to extend the growing season, allowing an earlier start to the year and the harvests to continue into winter, as well as protecting delicate plants from harsh weather. Large constructions are not the only option here. Coldframes were traditionally used to great effect, although they have become less common and those offered for sale now tend to be rather small. In the absence of frames, cloches can be used for protection on a smaller scale. Glasshouses and polytunnels offer further advantages, allowing crops that suffer outdoors, either due to climate or disease, to be successfully grown under cover. For example, peppers and chillies typically grow better under cover, outdoor grown tomatoes, being of the same family as potatoes, may suffer from late blight, and peaches, unless of a modern resistant sort, can be badly damaged by peach leaf curl when exposed to spring rains that wash spores onto the buds.

Polytunnels are the most cost efficient option, although one must consider the cost and inconvenience of replacing the covers every five to ten years. There is, also, to my mind, something particularly pleasant about a glasshouse. The qualities of light within a glasshouse and a polytunnel are quite different. The plastic cover of a polytunnel tends to disperse the light. Although there is some reduction in the amount of light, as compared with a glasshouse, this diffuse light is actually quite beneficial and gives a very nice growing environment. We are very fortunate to be able to erect both glasshouses and a polytunnel, so avoid any difficult decisions. Before the project commenced, we had a small glasshouse, which remained more or less where it was and now serves for propagation during the spring and as the ‘chilli house’ during the summer, as they are one of my favourite things to grow and this 10′ by 6′ glasshouse will be full with more than forty chilli plants.

With a south facing wall there is a quandary as to where to place the glasshouses for best effect. In the walled kitchen gardens, the benefits of the warmth of the south wall were often exploited further by constructing peach houses and vineries against them. In other cases, however, the wall was reserved for tender fruits, such as peaches and some fine pears, and the glasshouses would be freestanding in front. Although losing the benefit of the south facing wall, these glasshouses were typically heated with hot water pipes, so perhaps did not always need to be placed against the wall. It is a marginal decision, but as we have no intention of heating our glasshouses, the decision was made to mount them against the wall.

One problem encountered in the approach was that a typical leanto construction requires a very high rear wall in order to provide a useful height at the front of the glasshouse. In days gone by, unequal span, typically ¾ span, constructions were common, and lessened the height required of the rear wall whilst still providing a high ridge and large volume of air. Not only are these uncommon today, a 2 meter high wall is still not sufficiently tall. We resorted, therefore, to modifying two nominally 20′ by 8’6″ freestanding glasshouses to mount against the wall. This worked out very well indeed, and the spare parts from the rear elevation of each glasshouse are available to make the ‘lights’ for six coldframes.

Layout of beds

The traditional vegetable garden was dominated by large beds. Vegetables would be sown or planted out in rows, with just enough space between rows to walk, for weeding and harvesting. The seemingly increasingly popular alternative is to use a narrow bed that can be worked from the paths, although this is also, in fact, a very old way of laying out the productive garden. Most often these are referred to as raised beds, but often somewhat inappropriately as the purpose is not always to raise the soil level, but merely to avoid walking on it.

I must admit that I like the look of a traditional bed, planted up with long straight orderly rows. The old walled kitchen gardens would often be laid out with two crossing paths to form four such main beds. Deep borders would also be seen around the walls. To my mind, such a layout is very attractive. The disadvantage with large beds is the need for annual digging. Regular walking on the soil, along with the winter rains, results in considerable compaction, which is relieved by deep cultivation. Not only is the digging hard work, it is not necessarily good for the soil, disturbing the natural structure of the soil and its organisms.

Narrow beds, typically 4′ or 1.2M wide, can be worked on for the most part without stepping on the soil, and can be managed in a ‘no-dig’ manner. I have found, though, that some cultivation is still needed, for removal of deep rooted weeds, and for harvesting crops such as leeks and potatoes, which are deep in the soil. Nonetheless, there is certainly far less work involved in managing a system of narrow beds, once established. Rather than annual digging, manures and composts are applied, and worms relied upon to process it and aerate the soil.

Whether narrow beds are also raised beds is another matter. Typically, the beds will be edged with something to separate soil from the paths, although this is not essential. The soil level will tend to build up somewhat as the amount of organic matter is improved through the addition of manures and composts. However, this is not really a raised bed. When raised properly, there are additional advantages of allowing one to correct for insufficient depth of top soil or other soil related problems, warming up more quickly in the spring, better drainage in areas of heavy clay, and so on. However, there are also some drawbacks. If raised substantially, the typical free draining soil needs more frequent watering. The loose soil is not ideal for brassicas which prefer to be planted in firm ground. A layout with many narrow beds also has many paths to be maintained. If these are not kept weed free, weeds readily encroach on the beds. Since there is no longer any need to walk between rows of crops, planting is generally done at a higher density. Indeed, this is necessary to make up for the amount of space taken by the paths. This does, however, require the maintenance of a good fertile soil to support the increased density of growth. Whilst most vegetables will grow happily in a four foot wide bed, some could benefit from a larger area, such as winter squash and Jerusalem artichokes, and some others do not seem to fit efficiently into the space. Larger beds may also be more flexible when it comes to planning a suitable crop rotation.

In the end, the possibility of reducing the workload, was considered a key factor, and so we laid out our main beds in 4′ wide sections. To complement these, though, the borders along the fencing were left in 3 large areas. Two of these are used for overflow planting, which offers some flexibility and gives plenty of space for crops such as winter squash, and the third provides an area for perennial plants such as rhubarb, Jerusalem artichokes, and horseradish. Asparagus, another perennial vegetable, was allocated a narrow bed of its own. The overall plot is not rectangular, it narrows on both sides towards the south. As I tend to organise everything in an orderly manner, these angles are inconvenient, so the borders were laid out in wedge shapes, taking out the angles of the plot, and leaving the interior rectangular.

Some time was spent considering the crop rotation. This is potentially important for managing both disease and soil fertility. Pests and diseases are often specific to particular plant families. Many can reside in the soil for many years, and planting the same crops in the same place can allow a harmful levels of pests and diseases to build up. By grouping crops of the same family together, and rotating them through other parts of the garden each year, pests and diseases specific to that family lose their host plants. Some may well still remain, but in low numbers, and hopefully never building up to really harmful levels. The plan allowed for 12 beds, each of 4′ by 14′, giving a main growing area of 672 square feet, or just over 62 square metres. This is a handy number of beds as it divides nicely into the common rotations of three and four years. It became apparent, however, that the usual rotations described in gardening books did not divide readily the crops we wanted to grow.

In terms of soil fertility, the typical three year rotation is based on the sequence of legumes followed by brassicas followed by roots. Manures or composts are ideally added prior to planting legumes, and lime ideally added prior to planting brassicas; lime and manure should not be added at the same time. Such a rotation supports the feeding habits of the different crops. First, legumes fix atmospheric nitrogen in nodules on their roots, through the action of certain bacteria, not only providing their own source of nitrogen, but improving the soil condition for the brassicas that follow. These are the heaviest feeders and benefit from the extra nitrogen as well as the previously incorporated manure. Root crops, such as carrots and parsnips, on the other hand, tend to fork in rich soils and do not like freshly manured ground, and therefore follow on perfectly from the brassicas. There are, however, many crops that do not fall into those three categories. The four year rotation adds another bed into the cycle, often for potatoes. Other crops are then added ad hoc.

Initially, I had considered a four year rotation, replicated three times over the twelve beds, and, indeed, attempted this in the first season. However, the crops we wanted to grow just did not divide well into this and I have since moved to a six year rotation, which, along with the use of the borders, works out perfectly for us. The rotation is as follows: legumes – brassicas – roots – onions – potatoes – leeks / cucurbits and sweetcorn. This rotation is replicated twice over the twelve beds. In one set of six, the latter bed is used for leeks, whilst in the other, it is used for cucurbits and sweetcorn. This changes between the two sets each year. The balance now between the crop types is almost perfect for our use; the best rotation plan will, of course, vary depending on what balance of crops one wishes to grow. Saladings are grown wherever there happens to be space, and the borders offer overflow planting and space for winter squash.


The kitchen garden is not just concerned with vegetables, but also fruit. Whilst the orchard is the traditional place for the production of large quantities of top fruit, the traditional walled garden would provide a great deal of fruit, largely grown against the walls to take advantage of the protection and more favourable microclimate that they offer. Even without the benefit of walls, fruit can be readily trained in restricted forms such as espaliers, cordons and fans, so as to take up little ground space. They can be used to great effect to divide up a plot, and low growing stepovers make pleasant edging. Well trained, they can be very attractive, and I find the training involved quite interesting. They can be trained freestanding against a support of posts and wires, or along boundary fencing. At some point, we plan to plant an orchard on the remaining land beyond the kitchen garden. There is, to my mind, something very special about a traditional orchard, and we are very fortunate to have enough space to make a reasonable sized example. However, I am still keen to grown fruit in restricted forms in the kitchen garden.

When planning the position of fruit trees within the kitchen garden, various factors need to be considered, such as the impact on other crops, in terms of shade, nutrient and water usage, and selecting the ideal position for each type of fruit. In particular, if there are any south facing walls or other warm spots, these should be reserved for fruits that really benefit from such locations, such as peaches, apricots, figs, and certain pears. In our case, the south wall is largely taken up by glasshouses. One glasshouse is used for two fan trained peaches and the other for two fig trees. Ideally, more height would be available, but it is just about enough to train the trees. Between the glasshouses, a large pot contains yet another fig. For me, these are the prize fruits that I want to benefit from the heat. Each greenhouse also houses a vine, to be trained along the entire length, and as I tend to choose old fashioned varieties they will need the extra heat to develop good dessert grapes.

After the south wall, in our layout we have a large run of east facing fence. This aspect is not ideal, and neither is it sufficient for the fruit we want to grow. Therefore, two runs of cordon trained fruit are incorporated as borders along and between the vegetable beds, supported by posts and wires. The warmer run, which is within the area protected by the walls, is used for pears, which benefit most from the improved conditions. The other run is used for apples. Both apples and pears grow well as cordons, provided they are spur bearing varieties. The east facing fence in then used for fan trained plums and gages.

Soft fruit, and others such as cherries, benefit greatly from protection from birds, which would otherwise devastate the harvest. The ideal solution is a fruit cage, although ad hoc netting at the right time of year would suffice if there are only a few plants to cover. As we are very fond of soft fruit, a large fruit cage is incorporated into the design. In addition to the usual plantings of raspberries, strawberries, hybrids, blueberries, gooseberries, currants, and so on, the boundaries of the fruit cage allow fan trained cherries to be grown. Various frameworks are available with which to construct fruit cages, but as we already have east, south and west fences, it seems sensible to use these as three sides of the fruit cage.


I have limited knowledge of decorative plants. Although I am very fond of flowers, especially those of the cottage garden, and other decorative plants, I do not have entirely the same interest in studying their cultivation. Nonetheless, flowers do have some place in the kitchen garden. Not only are certain blooms edible, they are often useful as companion plants, discouraging or distracting pests. Beyond that, they do, of course, brighten up the plot. Traditionally, the kitchen garden would likely include certain flowering plants, often utilising one or more glasshouses, which might be heated to cater for the more exotic plants that were in fashion at the time. More practical, however, is the cutting garden: an area dedicated to growing cut flowers. I have little time for the obsessive production of annual flowering plants, especially bedding plants, although I am happy to buy a handful from the garden centre when we want to brighten up the garden. I am, though, interested in developing a cutting garden stocked with perennial flowers and, perhaps, some decorative foliage. This is the least developed part of the kitchen garden; so far it resides on the plan only, and its physical site is currently occupied by a huge mound of stones sieved out of the vegetable beds and other assorted scrap waiting to be cleared. I will update this page with more on the cutting garden once some progress has been made.


Breaking ground

Before digging commenced, there were a few small trees to remove. These did not seem to grow at any great rate and, when we dug around their roots to remove them, we found that the roots did not penetrate deeply; indeed, many ran along just below the turf. The ground has not been cultivated before and is very compacted. Coupled with an excess of stones, it seems that the roots simply could not develop properly. This is useful to observe for when we get around to planting the orchard; it will be necessary to really loosen the soil around the planting holes first.

After clearing the site, the main constructions – south facing wall, fencing, dwarf walls for glasshouses, coldframes, etc. – were marked out, along with the main vegetable beds. Cultivation could then begin. As this was previously uncultivated land, the soil needed loosening to a good depth. It is also important not to swap subsoil for topsoil, with the subsoil being somewhat inert, lacking the organic matter and organisms that make soil live and fertile. The most appropriate method was considered to be ‘bastard trenching’. This is very often confused with double digging, but, in my opinion, is quite different. Double digging involves the removal and turning over of two full spits of soil, whereas bastard trenching involves the removal of one spit of soil and the forking up of the next spit. It is, therefore, considerably easier to achieve, and less disruptive of the soil. We discovered quickly that there was a general excess of stoney matter. To begin with, we simply picked over barrow loads of soil to remove the worst. We later decided that the top spit of soil would benefit from sieving. This is a huge undertaking and to help speed things up, a large sieve was built that fit across the four foot beds or could be securely placed on a wheel barrow. We found that two people working on this could make good progress, with one digging the compacted soil out and loading the sieve and the other riddling with gloved hands. Nonetheless, this was a long, slow job.

Cut turf, provided it was free of perennial weed roots, was turned upside down in the bottom of the trench, and a generous amount of manure added. Further manure and composted green waste was added on top of the beds.

The beds were not specifically intended to be raised beds, but the soil merely contained within timber frames. However, the freshly loosened soil, with the additional of large amounts of organic matter, did serve to raise the level some inches above the top of the surround turf. An economical construction for the beds was made using 6” treated gravel boards for the sides and 2” square posts dug one foot into the ground. Gravel boards are a little rough, and only about 1” in thickness. It would be even better to use 2” thick boards, and perhaps 8” tall, but these are, naturally, much more expensive. Reasonable results can be achieved with gravel boards, although there is a tendency for the weight of soil to push the boards out of position.

Beds were planted up as we went along, ensuring a good harvest even in the first season. The paths, however, were weeded a couple of times, but did not get covered until the next year. This caused some difficulty; not only did they become overgrown where not regularly walked upon, they offered a ready source of weeds from which the beds would become infested, some through seed and others through runners. These could have been mowed, but I could not bring myself to buy a mower just for the purpose of the vegetable beds. Also, we felt it would be better in the long run for the paths to be covered. Whilst hard surfaces, made from self binding aggregates, would be ideal, an easier and perfectly serviceable alternative is gravel. To help prevent weeds from growing through the gravel, weed suppressing membrane was laid between the beds and pegged down. A few inches of 20mm gravel was then spread on top. Not only is this mix of angular stone one of the lowest cost options, it is also nice under foot and no problem for the wheelbarrow. Pea shingle, being small and round, tends to move too much under foot.

We still have an ongoing battle with weeds. Despite removing as much root as possible during the sieving process, and ongoing hand weeding, there are still many perennial weeds showing up, particularly thistles, nettles and creeping buttercup. We also had a large amount of deadly nightshade appear one year, even though we had not noticed it on the plot before. Being very poisonous indeed, we were quite thorough in pulling this, yet it has not yet been conquered. The soil also has a fair amount of annual weed seed, which often germinate just after cultivation. Just raking the soil over will sometimes disturb buried weed seeds and cause a blanket of new weed growth to cover the beds very quickly. Over time, however, this problem will diminish. Perennial weeds should eventually all but disappear, and we expect the extent of annual weed to drop somewhat, also, although weeding will never go away entirely. The borders will always present a challenge, as the weeds growing on the uncultivated lands on the other side of the fence encroach on the now much more rich and fertile cultivated beds.

Although we could grow green manures at the end of the season, there is a certain amount of effort then involved in digging them into the soil. Instead, for those beds for which enrichment is suitable within the crop rotation, we have opted to spread compost and/or manures over the beds once crops have been removed, and then to cover with further weed suppressing membrane. This keeps the beds in good condition for the next season. We have found with the borders in particular that it is worth removing this membrane a couple of weeks prior to planting, and raking over the soil to disturb any weed seeds. These then germinate and can be hoed off before sowing seed or planting out. We found to our great disappointment several rows of early crops such as carrots, beetroots, and spring onions, became swamped by a carpet of weed growth before anything could be done about it. The weeds grew more quickly than the intended crops and there was no way to readily remove them without losing the crops. This area had to be hoed clean and replaced with alternative plants. We found that in the particularly weed troubled borders that they are, at least for the time being, better suited to the planting out of young seedlings rather than direct sowing. This is not particularly problematic, as these are ‘overflow’ areas anyway, and there are plenty of other areas for direct sowing. After a few more seasons, I imagine that these borders will become quite clean as well.


Heavy construction

A kitchen garden can be developed without the need for heavy construction work. However, our site is inundated with rabbits and frequently visited by deer. Thus, we were forced to provide a certain amount of crop protection. To keep out the rabbits a fence must be dug ideally two feet into the ground. To keep out the deer, it must be rather tall. The RHS suggests 6′, but other sites suggest as much as 8′. We opted for a 2 meter – approximately 6’6″ – fence of heavy duty weld mesh. Although mesh is not entirely necessary to exclude deer, we decided to utilise the fence to also provide three sides of a large fruit cage.

After the first season of growing, the garden plan was changed somewhat as we realised that the twelve main vegetable beds and the three borders provided plenty of growing space. A space originally considered to provide four large traditional vegetable beds was instead enclosed with a low fence for keeping chickens. At this point, excluding foxes became a concern, as I had read that they are quite able climbers. We decided to add electric cables to the outer fence, three along the top supported on 45 degree overhanging metal brackets, and one at ground level to discourage digging. The physical barrier alone is almost certainly sufficient.

Along with the trench for the fencing, footings were dug for the south facing wall, and also to provide dwarf walls for two glasshouses and coldframes in front. The wall was constructed of 6” blockwork, with four piers on the rear for added strength. A return at each end of the wall also increases the strength of the construction. The blockwork was topped with capping stones, then rendered and painted. To maximise the use of the wall, dwarf walls were constructed of unrendered blockwork, to raise the eaves of the glasshouses to just below the height of the wall. This also provides the benefit of comfortable working height right up to the eaves and a large volume of air. From the front of the dwarf walls, six coldframes were also constructed.

The fencing was supported by 4” timber posts, concreted in place. Only a small amount of concrete was used so as not to cause too much difficulty when the time comes for repairs. Corner posts were braced in each plane, and the weld mesh clipped to straining wires.

Along with the development of the kitchen garden, the garage and outbuildings were renovated. The four attached outbuildings provide a potting shed, two storage rooms, and a forcing shed. Two courses of blockwork in the forcing shed provide for three narrow beds. These are suitable for cultivating mushrooms, forcing crops such as witloof chicory, and blanching others.


Glasshouses, coldframes, and polytunnel

When we embarked upon the construction of the kitchen garden, we already had a glasshouse roughly in place. Unfortunately, to fit in with the plan, it needed to be moved a couple of feet to better align with the rest of the garden. We therefore took the opportunity to build dwarf walls and raise the level of the glasshouse by two blocks and also add a small raised bed along one side. Raising the level of the glasshouse makes it much more comfortable to work in, as they tend to have somewhat low eaves. This original glasshouse is used, in conjunction with the polytunnel, for propagation in the spring. Thereafter, it is used exclusively for growing chillies, with space for 44 plants in 10 litre pots.

Two new glasshouses were to be fitted against the south facing wall. As we could not find anything suitable for this purpose, we decided to modify two free standing glasshouses from Elite Greenhouses, anchoring them to dwarf walls and to wooden supports bolted to the top of the wall. The result was very sturdy indeed. A further benefit was that the left over glass and glazing bars from the rear elevations were free to be used to produce ‘lights’ – the glazed hinged lids – for six blockwork coldframes. Many coldframes that are now offered for sale seem rather small compared with those of old. Ours are dictated by the glazing, resulting in good sized frames of approximately 6’ by 4’. The lights are still missing, but the frames are being used for growing melons.

Having secured the framework for the glasshouse, the glazing was fitted to the roof only, to avoid damaging the elevations whilst excavating the glasshouse beds. The beds were laid out, using 6” gravel boards and 2” timber posts. A narrow border was laid along the wall to accommodate fan trained fruit – peaches in one glasshouse and figs in the other – with a 2’ wide path and then a 4’6″ wide border. Along the front of each glasshouse a row of 19 canes is inserted into the ground and supported at the top by a wire attached to the roof glazing bars. The canes support in each glasshouse, on alternate years, cordon tomatoes or cucumbers and melons. The rest of the border is planted, on alternate years, with sweet peppers or aubergines, and often with summer squash at the far ends.

The soil was dug out to a good depth, and the beds forked over thoroughly. Manure and old turf were placed in the bottom, then the soil sieved back. The beds were further improved with the addition of more manure and composted green waste. With the beds prepared, the remaining glazing was fitted. Adequate ventilation of covered growing spaces is of great importance. We therefore incorporated many more vents than standard, with eight roof vents per 20’ glasshouse, along with two louver vents. Automatic openers were fitted to all vents. These are simple mechanical devices that rely on the expansion of wax within a cylinder, which operates a piston to open the vent, and can be adjusted to start opening at a given temperature, providing excellent control and preventing the glasshouse from overheating.

Whilst the glasshouses provide for fruits, tomatoes and cucumbers, a polytunnel was also added to allow various vegetable crops to be grown, as well as a wider range of tomatoes and sweet peppers. For freestanding structures, a polytunnel is much more economical, although, again, it is best if the tunnel offers good ventilation. We opted for a robust looking model from Haygrove, which offers full length ventilation as well as a partly ventilated door, with vents covered with insect mesh.

As we selected a tunnel of 4M wide by 12M long – approximately 13’ x 39’ – we considered two main options for the beds: a single central path with beds 5’6” wide, or a narrow border with a 4’ wide central bed. We felt the latter would be more manageable, despite losing more space to the paths. The border is used to grow, for example, tomatoes, sweet peppers and overwintered onions and garlic. The central bed, which is approximately 30’ long, is used for various crops throughout the year as needed.


Fruit cage and other fruit

Fruit within the kitchen garden is grown in the fruit cage, the polytunnel, trained along the walls and fences and on a system of posts and wires. For fruit trees, these are all trained in restricted forms, principally to minimise the space taken, but also because I am rather fond of these forms. A traditional sort of orchard and nuttery, with a small ‘truffiere’, is planned for the remaining free land at the end of the plot, and already has a walnut and sweet chestnut.

The fruit cage provides a large netted area, formed on three sides by the enclosure, with a roof and further side. The main beneficiaries of this protection are the soft fruits, which are otherwise very susceptible to damage by birds. The fenced sides are not wasted, however, with fan trained sweet cherries on the south facing elevation, a sour cherry facing east and a mirabelle facing west. Along the north facing elevation, blackberries and miscellaneous members of the Rubus genus are trained.

The structure of the fruit cage is dictated by a matrix of 4” posts, which support the roof and the outer fencing. These are placed at approximately 8’ intervals and the cage is 4 posts or roughly 24’ wide, and 8 posts or roughly 56’ long, with the longer axis aligned east to west. On the inside of the cage, six beds are formed, running north to south around the available pairs of posts. These are approximately 6’ x 9’6”. This provides enough space to grow a double row of cane fruit down the central axis, supported by wires secured to the posts, along with a row of strawberries in front of each row. Four beds are taken up with raspberries, with varieties selected by fruiting season; two each of early, mid, and late, summer fruiting raspberries and two autumn fruiting. The two remaining beds are used for hybrid berries, one per side, as these need plenty of space. A 2’ path is provide between beds, and around their perimeter, and the remaining space forms a border from path to fence. The west and east elevations follow the same slope as the rest of the garden boundary and the corresponding borders are therefore wedge shaped.

The entrance is located on the north elevation. The border to the left of the entrance is used for gooseberries, trained as double cordons so as to allow a number of varieties to be grown in a relatively small space. The border to the right of the entrance is used for blueberries. As the soil is on the acidic side, these can be planted in the ground. Further preparations which acidic manures and composts, such as composted bracken, help provide a suitable environment for these acid loving plants. The border along the south elevation is planted with black, red, pink and white currants. A jostaberry, a cross between a blackcurrant and a gooseberry, is planted at the north end of the east border. Some space remains on the south, east and west borders for further planting at a later date.

As much as possible of the fruit cage is covered with weed suppressing membrane to reduce maintenance. This includes all four borders and two of the beds – those where the hybrid berries are planted. Raspberry canes tend to come up in a bit of a mess so the reaspberry beds were not suited to using membrane. Openings were cut for planting the larger plants, such as blueberries and currants, whilst smaller round holes were melted through the membrane by heating the rim of an old tin using a blowlamp. This makes a neat hole with the benefit of sealed edges. The membrane is otherwise rather prone to fraying where cut.

Within the polytunnel, three 100 litre pots were sunk, two in the main bed and one at the far end in the border. These are used to somewhat constrain the roots from three small trees – one peach and two apricots, one of which will be fan trained, whilst those in the main bed are maintained as miniature standards.

Being mounted against a south facing wall, the glasshouses offer the ideal place to grow choice peaches and figs; two peaches are being fan trained in one greenhouse and two figs in the other. The roots of figs must be constrained to limit vegetative growth and encourage fruiting. The traditional method is to dig a pit a couple of feet deep, line with slabs, and fill the bottom 6 inches or more with rubble. We opted for a simpler approach, sinking two 100 litre pots into the border, and filling the bottom few inches with coarse gravel.

Each glasshouse also contains one grapevine, which will be trained first vertically up the south facing wall, then along the top of the wall in both directions, with spurs running up the rear roof, secured to wires fastened to the glazing bars. In this way, the grapes will receive good light but will not cause much shading of the other fruit. To make full use of the border, a row of strawberry plants is planted in front of the trees.

Two rows of cordon trained top fruit are incorporated into the garden plan as dividers between the vegetable beds. These are supported on a system of posts and wires. The wires are connected to tension bolts, and the posts are braced at each end for strength. Timbers also run along the top of the posts for further stability. The cordons are planted at 45 degrees, which helps improve bud break along the entire length of trunk as well as offering a longer length of fruiting wood and assisting in the control of vigour. The west facing fence is utilised for plums and gages, fan trained against wires secured to the fence posts. Approximately 12’ width is allowed for each tree.