I wanted a Forest Garden before I discovered Permaculture, and before I found out about Peak Oil. Back in the 80’s (1980’s not 1880’s) I had read a book called ‘Smallfarming’ by Malcolm Blackie. He was a New Zealander, and had included a section on Agroforestry. It made a lot of sense to me, well before I understood the permaculture principles of succession and stacking. About 2006 I bought ‘How to Make a Forest Garden’, by Patrick Whitefield. I found the plant tables and information useful, but was really disappointed that all of the examples/layouts that he included were for really small spaces. With one or two acres to plant, it didn’t seem at all relevant to me, and I put the book away and sulked a bit. I did my Permaculture Design Course with Patrick in 2008, and he seemed to do his best to dissuade us all from Forest Gardening, or at least did a great job of playing Devils’ advocate. Before and after the course I read the ‘Earth Care Manual’, and later re-read ‘How to Make a Forest Garden’. Second time around I was able to take the smaller designs, find a pattern to modify, and then use that to design my own Forest Garden. The book was the same, it was me that had changed. One of the lessons that I took from this was that as I learnt more, I was able to gain more from the books, articles and texts that I read. I now go back through books in my library before starting a new design, and I normally discover facts and ideas that I had missed, or only partially understood, first time around.
This design was the second element of my project to be designed and implemented. The process behind the decision to create a Forest Garden, and the placement of it, is described in the overall design of the project. In order to better understand that, it would be worth reading the Wolds Woodland Farming Project design first. One of the key reasons for creating a Forest Garden was to reduce my reliance on annual vegetables. This is an example of the permaculture principle every important function should be provided by more than one element. It is also an example of the use of succession, which is a design tool, gradually supplementing/reducing/replacing annual vegetable production.
The design uses my own design process, details of which can be found HERE.
The Forest Garden is located in the long block in the North east Section of the aerial photograph above, and is shown as 2 in the key
Establish Aim and Objectives
The Aim of the Forest Garden is to create an acre of productive food forest.
- Provide food
- Supply forage for my bees
- Give wood for fuel
- Provide a good habitat for wildlife
- Improve the health of the soil
- Give Shelter to other Elements within the design
- Look, sound, and smell wonderful
- Provide a foraging area for poultry
Most of the survey information for this design is recorded in the Overall project design, for which I provided a link earlier. Exceptions are noted below.
This area has the most pronounced slope of the whole property, as shown in the diagram below.
The picture at the top of the page was taken from the top of the slope looking down at the house, and gives a sense of scale. The picture below is the same view in Winter, and therefore has less vegetation to obscure the view.
The slopes face slightly West of South. The Southerly aspect increases the intensity of sunlight that falls on the land here. Photosynthesis increases with temperature and as this space gets full sun in the late morning and afternoon, when it is generally warmer, it should help to increase plant growth.
At the time that this design was undertaken the swales and pond had already been dug. The diagram below shows their location.
The pictures below show the swales during construction, and the pond.
This swale bank was augmented with wood, in order to help it retain moisture, and feed plants. This permaculture technique is known as Hugelkultur, and is promoted by Sepp Holzer in his books. It was the only one that I tried this with, and I want to see if it leads to a better establishment of plants on the swale bank. If it does, I can use the technique again. This is an example of the principle observe and interact.
There were about nine existing trees in this space, Hawthorn and Elder, all along the boundary between this space and the coppice area. This suggested that there used to be a hedge here. The hawthorn trees have substantial trunks, and would therefore appear to have been there for a long time. I had planted a small block of ash, and willow, in the South West corner of this field, back in 2000/2001. These can be seen in the aerial photographs above.
The shape of the ground, and the hedges to the North and East, create a sheltered microclimate, providing protection from the damaging cold winds that we get blowing from continental Europe, over the North Sea The slope is steep enough to allow frost to roll away downhill. This makes it ideal for fruit growing. Combined with the additional sunlight mentioned earlier, the space is almost begging to be used for fruit, nuts, grapevines, or something similar.
Forest Garden Information
Having read Patrick Whitefields book, and done plenty of internet research, I thought that I was pretty well set up to design the garden. However I benefited hugely from the release of Martin Crawford’s book, Creating a Forest Garden. There was enough depth in the book to really help with the design process, but without being over complicated. I also read both volumes of Edible Forest Gardens. At the time these seemed too complicated, and detailed. Some of what I read did make it into the design process, particularly when planning for wildlife habitat. Again, having gone back to both volumes later, I found their design process brilliant, and used elements of it in my Chicken Scavenging Design, which can be read HERE.
There was of plenty of information to process, but there were some that were going to be critical to getting the project to work. I identified the following as key to the success of this, and probably most other Forest Gardens.
The biggest mistake that was quoted in the books above was planting trees too close together/too many for the space available. Getting this right was going to be difficult for me to achieve as I also wanted to include a significant yield of firewood from the Forest Garden, and so if I planted too few trees, that wouldn’t be possible. Planting more densely and thinning was an option, but would mean that I couldn’t afford to miss the thinning out without potentially reducing the yield of food, which was the top priority for this space.
Establishment of Trees
Trees, and especially fruit trees, suffer if planted directly into vigorous grassland, as the roots of both compete for water and nutrients in the same part of the soil profile . The methods of establishment normally used are:
- Application of herbicide prior to planting, and for the first couple of years afterwards.
- Rotovation/ploughing, and the establishment of a more benign ground layer prior to planting. This is often carried out in combination with the application of herbicide first.
- The use of plastic based mulches, either in strips, or as mats around individual trees.
None of these suited me. As a beekeeper I will not use anything that might harm them, and that includes all herbicides. I was hoping to plant a lot of trees, and at about £1 each for mulch mats, that was going to be a considerable additional expense. Martin Crawford uses a wide strip of plastic mulch material with creeping ground cover plants behind. Each time that the strip is moved, the creeping plants have already sent roots under it, ready to exploit the light once the strip has gone. He also raises plants to put into the clear space left. This is repeated each year. I liked this method, but with the site exposed to the South West (and the prevailing winds), I had visions of my plastic mulch turning into a large kite, and wrapping itself around something important further downwind. Whilst there are a few properties locally that could do with a bit of ‘gift wrapping’, I felt that it wouldn’t do much for good relations with my neighbours. Using the pattern of this method, but changing the detail seemed possible.
The successful establishment of the fruit trees would benefit from shelter from the wind. This was provided for the cold Winter winds by the hedges to the North and East of the property, and the shape of the land itself. My neighbour’s donkey was doing a reasonable job of making the Eastern hedge more permeable, and if I wanted to harvest firewood from the hedges, I would need to provide additional shelter. There was less shelter from the South Westerly prevailing winds. The hedges to the South and West of the property, were some distance away, as can be seen from the Aerial photographs above. The coppice area will eventually provide shelter to the Forest Garden, but it would be a few years before the trees in that space were tall enough to do so. Providing additional shelter along the western edge would achieve that, and with good species selection, increase the yields of firewood, bee forage, food, and habitat.
Choice and Establishment of the Ground layer
This is perhaps linked with the point above, although my case is slightly different to most. Most Forest Gardens are smaller than mine, so the lower layers (shrub, ground, etc.) are normally made up of food plants, to increase their total yield. With such a large space, there was little need to maximise food production in these lower layers, which would allow me to utilise them to help meet my other objectives. The choice of plants was likely to influence how they needed to be established. My previous designs had identified late season bee forage as a priority, so concentrating on plants that flowered July to October would be beneficial.
Summary of analysis
The design needed to solve the problems above. Planting a lot of trees without overcrowding, suppressing grass growth without chemicals or mulch, and choosing and establishing plants that flowered later in the year. Not much to ask………….
In addition to the list of functions listed in the aims and objectives above, the analysis had identified the provision of shelter as a need. It also provided an opportunity to meet my need for fuel, and increase pollination with the right mix of species.
I also decided to add an element of timber production to the design, to increase the number of outputs from the Forest Garden.
There were already swales and a pond in place to collect rainwater, and infiltrate it into the soil. This is an example of the permaculture principle Catch and Store Energy.
In order to provide protection from the wind, I decided to plant shelter belts of trees. By good species selection, these trees would also provide food, fuel, bee forage, assist in pollination, and provide niches for wildlife. This is an example of the use of the principles every element should provide more than one function, and obtain a yield.
Fruit and nut trees would provide the main focus of this area, which was food.
I decided to devote most of the ground and shrub layer plants to bee forage.
Tree Species Selection
I had already decided to use the many of the same trees in both the Coppice area, and the Forest Garden. This included Hazel, Cherry Plum, Cornelian Cherry, and Sweet Chestnut. Doing so would improve the pollination in both areas, and would also allow me to shift my focus from food to fuel, or vice versa, depending on my circumstances. This is an example of the principle apply self-regulation and accept feedback. Here it is designed in at the outset, to allow me to act on that feedback. I wanted to include Mulberry to exploit it’s potential as a source of chicken forage, which was one of my ‘could do’, objectives. All of these would be improved/named varieties to achieve maximum productivity.
Within the shelter belts I planned to use all of the above, interplanted with Italian Alder. This would again promote better pollination, allow me to harvest fuel, provide bee forage, etc. as outlined in the Choose Elements paragraph. I decided at the last minute to add Amelanchier. Initially I had not included it as it flowers at the same time as Oilseed Rape, but having experienced dry spells where the rape failed to produce nectar, including it would ensure that should that happen again, I had bee forage available. This is an example of the principles every important function should be provided by more than one element and Apply self regulation and accept feedback. Amelanchier has exquisite Spring and Autumn foliage colour, and lovely white flowers, so would enhance the appearance of the Forest Garden, which was one of my objectives.
Other than the timber trees, all of the others provided bee forage, but only Sweet Chestnut met my need for late flowering plants. I therefore added False Acacia (flowering in June). I had also researched the flowering of Eucalyptus as being late, and that of Spinning Gum (E. perriniana) as August flowering. I had grown some from seed, and chose to plant them here. As Eucalyptus can adversely affect the growth of neighbouring plants, I didn’t want to place it too close to the main food production trees, so they would need to go into the shelter belts/timber trees. As these would have the same species planted within them as the production trees, I would be able to watch for any signs of allelopathy, and use that information for any future plantings. This is an example of the principle observe and interact. The Eucalyptus is fast growing, and a good fuel tree, it also keeps it’s leaves over Winter so adds to the shelter at that time of year. I use Eucalyptus essential oil as part of my varroa treatment, and hope that the bees will gain some benefit just from contact with the flowers, and from the nectar. That’s plenty of functions to justify its place in the planting scheme. The juvenile foliage is also stunning to look at.
I used the same mix of timber trees as for the Coppice and orchard design. These were Oak, Western Red Cedar, Larch, and more Sweet Chestnut. These were also to be interplanted with Italian Alder and willow.
I am having an affair with bamboo. I have a research project looking at it’s performance under my conditions. I am particularly interested in it as a perennial food source that can be harvested fairly early in the year. Structurally it provides an ideal environment for chickens, can help to provide shelter for other plants, and amongst its many other uses can provide materials for construction. The Forest Garden was an ideal environment to monitor (Observe and Interact) how it performed. I decided to plant one timber bamboo (Phyllostachys nigra f. henonis), two that provided timber and tasty shoots (P. vivax, and P. praecox), and two that were primarily food sources (P. nuda, and P. auroesulcata f. spectabilis)
Shrub and Ground layers
I wanted to include Rosa Rugosa as part of the shrub layer. Although it is a source of food for humans, I was more interested in it’s long flowering period (bee forage) and the value of its hips to wildlife. The swale banks gave me a drier than ‘average’ niche to explore, so I decided to plant Sea Buckthorn. Again, although a potential food source, I was looking at its value as bee forage and wildlife food, but with the additional benefit of Nitrogen fixation.
I planted Caryopteris and Perovskia on the swale banks, along with Buddleia, in order to augment late bee forage. All prefer drier conditions, which the bank provided. These were chosen after observing honeybees working them for nectar in late Summer at plant nurseries. Another example of the principle observe and interact.
I wanted to include Eleagnus ebbingei in the planting. It flowers late, fruits early, is evergreen, and fixes nitrogen. Sadly it is described as not liking cold wet conditions, which we have for prolonged periods in the Winter. I therefore decided to try a trial planting of three plants, in relatively sheltered positions, to see if they could cope with my conditions. This is another example of the permaculture principle, observe and interact.
As will be explained in the implement section below, my implementation plan precluded the bulk establishment of a ground layer. I therefore decided not to try and plan it in great detail, but to design and add to it over time. This is an example of the design method design by increments, and of the principle use small and slow solutions. I made four exceptions.
I use the essential oil from mint as the main component of a natural treatment for the varroa mite in my bees. It also provides bee forage in late Summer, one of my key needs. I wanted to grow masses of it to see if the nectar collected might have the same effect as using an essential oil. There is also a possibility that the volatile oils produced by plants like mint may help fight bacterial and fungal infections in trees.
There are two patches of Purple Loosestrife next to the river in my local town. It flowers in late Summer, and whenever I saw it in flower, it was always covered in honeybees. I cannot be sure if it’s a particularly good source of bee forage, or if it was the only source during a dry spell, but I wanted to plant some. This is yet another example of the principle observe and interact. The swale ditches gave me the perfect (wet) conditions to plant it into. Yet another edge to exploit.
I grew Musk and Common Mallow in my vegetable garden, but only the Musk Mallow attracted honeybees, even though both are quoted as being attractive to them. Flowering late, and over a long period, attractive to look at, and edible, it was a logical choice.
You can never have enough mulch, and the wet conditions in the bottom of the swales, coupled with the high nutrient levels there, would provide the ideal environment to grow it. So I grew Common Reed from seed, in order to plant in them.
I also decided to leave some areas of grass, and to leave some of it unmown. All of my neighbours cut their grass frequently, and our patches of long grass, and the piles of mown grass, provide a great habitat for voles. Although we do get some vole damage, I love to watch the barn owls and kestrels hunt over the fields. I could tell you that this is me designing in an area for wildlife (zone 5), which would be true, but it also gives me less grass to mow, which at my age, is good people care.
I am describing the placement here in one chunk, but in reality the decisions were made over a period of time with some made during the overall project design, and others during the implementation. The diagram below shows the initial design, and I will elaborate in the text underneath.
Use of Patterning
During the overall project design I observed that the external hedging resembles the woodland glade pattern, described in ‘How to make a Forest garden’, by Patrick Whitefield, and the pattern for cold climates in Permaculture Two’, and ‘The Designer’s Manual’, both by Bill Mollison. The diagram below shows the pattern in schematic form.
The shelter belt trees shown in the South West corner of the Forest Garden Overview, and the few existing trees along the boundary to the West, suggested that I could replicate the pattern by extending the shelterbelt all along that Western side. This would shelter the Forest Garden from the prevailing South Westerlies. In order to maximise the production of fuel, I decided to make the shelter belt five trees wide. To consist of the trees described in the Components section above, with the fruit/nut trees along the outer edges where they would get full sun for part of the day.
The Timber trees would need to be placed to the North of the main production fruit trees to avoid them casting shade, which would reduce fruit production. They would also provide increased shelter from the damaging North Winds. I decided to add some of the hedging/seedling fruit and nut trees to the Southern edge of the timber trees. Here they would still benefit from full sunlight, and add to the functions of this part of the design.
The Timber trees to the North, Shelter belt to the West, and the existing hedge to the East, formed the glade pattern for the whole Forest Garden. I also decided to add another shelterbelt along the eastern side, to give shelter even when the existing hedge was trimmed, or harvested for fuel.
Having designed the pattern at this level, I decided to continue it through all levels of the design. With the West and East Sides Sheltered, I decided to plant a band of trees directly below each swale. This is not the normal way of planting. Swales originate from Australia, where the climate is much drier, and the old soils are depleted of nutrients. The fruit trees are normally planted on the swale to make use of the extra nutrients and moisture. That was less critical on English clay. Having planted three rows of trees below each swale in the Coppice and orchard, I decided to repeat that in the Forest Garden. Planting in bands like this assists water infiltration from the swale, and increases the yield of timber. By good species selection it would also increase bee forage, and aid in good pollination and fruit set. However the two key yields of this choice are an increase in shelter for the production trees, and an increase in wildlife habitat. In ‘Edible Forest Gardens’ Jacke and Toensmeir suggest that creating a ‘lumpy texture’ creates more ecological niches for wildlife. These additional bands, when viewed in cross section, give a ripple/wave effect to the heights of trees, providing a variety of heights. Another result of these bands is that the woodland glade pattern is repeated below each of the swales, albeit with a wider shape.
Although it would have been possible to continue with a more orthodox planting scheme for the primary food trees from this point, I decided to create individual glades for them using a mixture of the species already listed. This is shown in the Forest Garden Overview diagram above, with the glades denoted by the letters Gl. This means that the woodland glade pattern has been used at all levels of the design, from the overall plan for the project, down to sheltering individual fruit/nut trees. The aim of this was to create a really sheltered space, and beneficial microclimate for the production trees, and for the insects that would be needed to pollinate them. Additional benefits include increased woody material in the soil to promote mycorrhizal growth and increase soil organic matter, more bee forage, more fuel, more edge, and the option to increase food production by grafting productive varieties of fruit and nuts onto cheap hedging trees. It also places the production fruit trees in a setting that matches their successional niche. Most fruit and nut trees are woodland edge trees, and this pattern creates that effect for them.
Oh and I nearly forgot, this is a Forest Garden, with ‘proper’ production fruit and nut trees, and each of these were to be placed within the shelter of its own little glade. Just in case that wasn’t clear
The picture above shows a cider apple tree, sat in it’s own little glade. The scythe mown grass hasbeen heaped around it to make it easier to see the pattern. The brown colour is the ground, newly exposed by my skilful use of a scythe (that’s my story, and I’m sticking to it). It shows the extent of the glade. Around that you can just about make out the support trees/shrubs, and the plastic spiral guards. After taking the picture, much of the grass was raked back to the outer edge to help much the trees, as described in the next section. That’s shown in the picture below.
Use of Edge
The use of swales and patterning gives me a huge volume of edges. These include diverse drainage types, vegetation types, vegetation heights, plant densities, light and shade levels, fertility levels, and many combinations of the above. This should really boost the number of ecological niches, and the productivity of the Forest Garden, and is an example of the use of the principles use and value diversity, and use edges and value the marginal.
As an experienced blogger, much of the work that I undertook for the implementation was posted at the time. Rather than repeat all of that typing, I’m going to provide links to the relevant posts so that you can read them just as I recorded them.
I wanted to plant the shelter trees first, to get a headstart on protection for the production trees, but before then I had a few jobs to do. Firstly I planted out a number of potted plants in September 2010. These included the Eucalyptus, bamboo, and a number of shrubs on the swale bank. This is described in a blog post HERE
Much of the detail of the site preparation is recorded in another blog post the following month. It’s called My New Forest Garden. It contains pictures of the site, and details of my use of a 2 meter ‘A’ frame, to both to set out the lines of trees on contour, and to mark out tree planting spacings, directly below the swales.
Another post describes the site preparation. It also touches on the maintenance method that I chose, which would normally be recorded in the next section. The second post is HERE.
My excitement at the prospect of starting the work, and a list of the first batch of trees is recorded in a blog post entitled Tree planting about to begin in earnest.
I reported on my progress, and changes made to the initial design in a post entitled 2010/2011 Tree Planting. Update One
The following post is an update on the Forest Garden made in July 2011, even though the title says 2012. A lot of the content could be put in the Maintain section below. Forest Garden Update July 2012 (sic.) It contains lots of pictures of the Forest Garden during that first growing season, including some stunning pictures of Musk Mallow flowers with bees covered in pollen inside them.
If you have decided not to read the above posts, a very simple summary is that I planned to split the planting into two. Shelter trees first in 2010/2011, then primary fruit trees and glades in 2011/2012. This was amended with the glades also planted in year one, and an increase in the total number of trees planted.
This is shown in the updated schematic below.
Key to production trees
1 Phllyostachys vivax
2 Phllostachys nigra f. henonis
3 Phllostachys nigra f. henonis
4 Phllostachys nigra f. henonis
5 Phllostachys vivax
6 Cider Apple on M25 Rootstock ‘Stoke Red’
7 Cider Apple on M25 Rootstock ‘Tremlett’s bitter’
8 Cider Apple on M25 Rootstock ‘Stoke Red’
9 Hazel ‘Corabel’
10 Cider Apple on M25 Rootstock ‘Fair maid of Devon’
11 Cider Apple on M25 Rootstock ‘Tremlett’s bitter’
12 Hazel ‘Butler’
13 Hazel ‘Segorbe’
14 Sweet Chestnut ‘Bouche de Betizac’
15 Sweet Chestnut ‘Precoce de Migoule’
16 Cornelian Cherry ‘Gourmet’
17 Cornelian Cherry ‘Jolico’
18 Cherry Plum ‘Gypsy’
19 Sweet Chestnut ‘ Marigoule’
20 Cherry Plum ‘Ruby’
21 Cherry Plum ‘Mirabel de Nancy’
23 Mulberry ‘Illinios Everbearing’
24 Mulberry ‘Black Tabor’
25 Phllostachys aureosulcata f. spectabilis
26 Phllostachys nuda
27 Phllostachys nuda
M Monkey Puzzle
S Specimen Sweet Chestnut (unimproved, but large)
Comfrey and other beneficial guild plants will be planted commencing in 2013.
My maintenance regime is an integral part of the design, aiming to assist in the establishment of trees, and the ground layers. It is also designed to help steer succession towards a specific end, and to accelerate that process. If you were lazy, and didn’t read the links above, you do need to read this one, which explains my maintenance method. My New Forest Garden (Site Preparation). As described in the post, the establishment/maintenance is rooted in permaculture principles. Observation and Interaction for the close spacing, and the grass mulch. The use of Small and Slow solutions for the establishment of the ground cover layer, and the Use of Renewable Resources for the grass mulch, close plantings, and ‘chop and drop’. The use of the scythe also allows me to move fertility from the open areas to the trees. This should help to reduce the vigour of the grasses, which will help to favour wild flowers. It will also help with tree growth.
I love this design. The Forest Garden will not reach it’s full potential for some time, so it is difficult to judge the effectiveness of the design, but I love the way that the woodland glade pattern is manifested right from the macro, through layers, and down to individual bays for production trees. I also love the way that adding the shelterbelts, and the bands of trees below the swales, allows me to pack nearly two thousand trees and shrubs into an acre of land, and still leave adequate space for the main production trees, and open sunny areas. I have yet to hear of anybody else, anywhere in the world, doing it the same way. That’s both exciting, and a little worrying at the same time.
All of the little experiments that I am undertaking in the Forest garden adds an element of research to the list of functions that this space provides. Not only does this have a value to the wider permaculture community, but it keeps me mentally engaged, which is essential for me.
The one thing that I would change is that I didn’t lay out a path scheme before planting. Not doing so means that any future path system will have to wind its way between elements of the Forest garden , and will be less efficient. I had considered laying out a deandratic patterned path system, but got carried away when the trees arrived. Just as for the Coppice and Orchard design, I would seriously consider rotovating and establishing a more benign groundcover, perhaps with the addition of a mulch of hardwood chips around each production tree. Not doing so is going to create more work, but does allow me to experiment with an alternative.
The Implementation and Maintenance
As far as the implementation goes, I now have rabbits living within the Forest Garden. This will leave some of my trees vulnerable, as many were planted without guards, relying on the protection of the perimeter fencing. I guess that after 12 years, there were always going to be gaps/breaks. I have therefore used up all of my remaining spirals on the important trees, leaving the willow and alder unprotected. As both are there in part to support the other trees, if they have to act as a decoy in order for others to be left alone, that’s OK. The likelihood is that they will undermine the swale banks, and that my beautiful bands of Musk Mallow, as shown in the picture at the top of the page, will be destroyed. Of greater concern is the potential for them to destroy the new growth of my bamboo.
Providing habitat for voles has also led to the loss of some trees, almost exclusively Italian Alder. This has been interesting as no other trees have been affected, and any tree in a spiral has been left alone. It’s possible that the piles of grass next to the trees has given the voles a great living and hiding place, from which to gnaw at the base of the trees.
Whilst not initially a function designed in at the outset, the bands and patterning act like a comb, catching windblown leaves, and even dust from adjacent farmland. This is an example of catching and storing energy. Whilst I didn’t plan it here, I have used this pattern in the chicken scavenging area, to fulfil the same function. This action is an example of the application of self regulation, and accepting feedback.
The Design Process and Method
My own design process and method worked well for this design. It does seem well suited to this type of design, and to my way of working.
Tweaks started before the design was fully implemented.
The number of bays was increased, as were the overall number of production fruit/nut trees included. These included some cider apple trees, added after I over ordered cider trees for another part of the design. I reduced the length of the glade ‘arms’ to increase the open space. it also gives me the option of adding more shrubs along those arms, keeping the height down, but the functions up. Some of these shrubs may be soft fruit, but the distance from the house makes it less likely that these will be harvested. Late flowering shrubs, like Chinese Mint Bush, might be better, as they will give a significant yield of nectar late in the season. Wildflowers for late season bee forage is another option.
I was offerred some sensibly priced Monkey Puzzle trees, sold cheaply due to cosmetic damage. These were bought and planted in suitable places. Their slow growth and open canopy mean that they would not cast significant shade upon other trees.
Further tweaks include placing the grass mulch further from the trees to reduce vole damage, and using more woody material around the production trees to speed up the establishment of beneficial fungi.
Update September 2012
I started some preliminary experiments for establishing a productive ground layer. having scythed the grass in August, I sowed clumping rye, mustard, chicory, borage, phacelia, 4 types of clover, sweet clover, yellow rattle, lady’s bedstraw, meadowsweet, and knapweed onto patches of exposed soil. Although grass regrowth is competing with the new seedlings, early signs are that this method will help to establish more flowers in the grass, for very little effort. Continuing to mow and move the grass in late Summer should help this by reducing fertility in the open spaces, making it easier for the less vigorous species to become established. The yellow rattle may weaken the grass if it gets a foothold, which is none of the reasons for including it in the mixture.
I will need to evaluate how much damage is being created by rabbits, which will be best carried out at the end of winter. Imay need tofill in some gaps, and use more tree guards to protect the trees. I may also need to plant more trees in the open areas in order to create more firewood and mulch material. I expect to use cuttings of willow and Lombardy poplar for this, to save money.