The first Food Forest lesson I attended during one of Geoff Lawton’s PDCs gave me a whole new perspective on their creation. Geoff’s experience, understanding and explanation of successional process, in conjunction with his vast knowledge of legume plants, has changed the way that I will now approach their design. This relates in particular to their management through time, with some of the strategies mentioned briefly in my previous food forest article.

I was aware, however, after visiting some very vast 20 year old food forests, that the task of understanding and managing such a system can be very daunting for some. Where do I start? What do I chop back? How much can I chop? These decisions can be made quite easily by someone with experience. However, if the Food Forest model is to be achievable by the masses, as it was originally hoped, we need a way of successfully communicating the management steps to be undertaken.

I believe replicable patterns could be a successful means of doing so. It brings a vast area back to an approachable scale, and depending on the desired forest size, could be repeated accordingly.

One thing which stood out for me during Geoff’s lesson, was his recommendation of the ratio of support species to productive species through time: At the initial implementation, 10% Productive to 90% Support; At the point of Canopy Forest, 90% Productive to 10% Support.

After class, I decided to try to identify a replicable pattern of a climax food forest which contained this ratio of 90% Productive Plants and 10% Support Species. I began with a Long Term Legume Tree as the central node, so that each of the surrounding canopy trees would have access to the nitrogen made available. As luck would have it, by surrounding the Legume with 6 mature canopy trees of the same size, with understorey trees below each of the small gaps in the canopy, the ratio of area worked out to be almost perfect at 90:10. These became the central elements of the design which we must support in time and space.

I then sat down one night and decided to reduce the central elements through time, in a series of steps right back to when they were planted. This was an attempt to find other replicable patterns which would bring the system up through time from 10:90 to 90:10 (Productive Spp : Support Spp). A bit of messy scribbling later, and we have the patterns that you see in the Figures attached.

I believe a series of patterns such as this has the capability to allow:
- Efficient design and placement of appropriate support species for our major elements in both time and space
- Efficient positioning of shorter term productive species.
- Confident placement by anyone of support species in relation to Productive species
- Communication of which plants to manage and when (chop & drop, pruning, removal)
- Understanding of which trees have priority for light space over others, leading to appropriate management.

Due to different climates, soil types, rainfall, microclimates etc, it would be impossible to state exact dates of management and removal. What a pattern such as this allows is the communication of events to look for which trigger certain management interventions. For example, if the short term legume trees begin to compete for light with our desired productive species, chop them back. When our medium term Legume tree begins to compete with our desired canopy tree, it is time to coppice it. When the summer rains begin, chop back the short term legumes, cut to 150mm pieces and mulch the major elements Etc.

For a pattern like this to be helpful, a detailed understanding of the various element’s characteristics are obviously essential. This is especially the case during the design phase.

For example, by having an idea how big a mango tree will be in five years time, we can far more effectively design the space into which it will grow. We might choose a papaya for example, which yields early but becomes fairly unproductive after a few years anyway. Hence, it is no loss to chop it down at this point, eliminating competition with our main element the mango. We obviously wouldn’t place a custard apple in this space, which takes four or five years to start producing a decent amount. Another option would be a coppicable legume tree, which has a life span of around 5-7 years such as a pigeon pea. By the time it would have begun competing with our major element, it has already given five years of mulch, nitrogen, and shelter. The more information we have available to us, the more efficiently our system can be designed, once again in both time and space.

A central permaculture database would be a very valuable resource when designing efficient and productive food forests. This could include: tree height and width at a variety of ages (ie, 5 yrs, 10, 15, 20, 25, etc not just at maturity), years to maturity, years until bearing, fruiting life, life of the tree, shade tolerance, fertiliser needs, yields per hectare at various ages, pictures of root structures. This would enable us to begin talking in more specific and quantifiable terms when designing, rather than just the good old “trust us it works” scenario which seems to still be lingering on.

These are just a few of my thoughts and any feedback, thoughts and improvements would be greatly appreciated.

All the best, Cam Wilson