Lessons from Fazenda Bella’s Agroforestry Systems

By on

On September the 15th (2018) we received an informal visit from a team that work developing food security policies at the Ministry of Social Development at Fazenda Bella, in Brasília. This group’s visit and questions are common in guided visits and courses at the farm. The answers tell a story about the evolution of our systems, practices and approaches in the last few years.

Even though we had been thriving with the production of fruits and vegetables since march 2015, last year we won a prize for best practice in sustainability in rural areas from the state Agriculture Department. The prize was shared with many other producers, but brought more visibility to our guided visits, permaculture, natural building and agroforestry courses at the farm.

In addition to their interest in how fruits and vegetables can be produced in agroforestry systems, the team from the Social Development Ministry was interested in the quality of the produce. They were also interested in these systems’ viability for small farmers, and in their efficiency to restore ecosystem processes. We grouped some of their questions and summarised our answers below.

What is the best orientation in relation to the sun’s aspect?

The answer that often bothers those seeking a simplified and standardised understanding of these systems is that the best orientation depends on a number of variables. These might be, for instance, the kind of production being prioritised, slope orientation, how steep the slope and the property’s degree of latitude.

Although we have not implemented our systems at Fazenda Bella on contour, they were established running east-west and relatively across a mild north-facing slope. Even though this might take more time with pruning and managing to optimise sunlight exposure, we chose this layout to favour the horticulture production, one of the enterprises we wanted to develop from the start.

If our choice was to focus on the production of fruits and timber, we would probably have chosen a north-south orientation, which in our case would maximise the sunlight.

Tree rows spaced 5 metres apart at Fazenda Bella.

What is the ideal distance between tree rows?
Again, there is no ‘cookie cutter’ solution. In the first 3 systems we implemented we established the tree rows 5 metres apart and with 3 rows for horticulture between these lines.

During the first year and half we were establishing these 3 systems we did not have a specific focus on any particular tree species for production. On the contrary our focus was on diversifying the system as much as possible as an strategy to fast track our learning. We also did not design grass rows adjacent to tree rows in order to grow our own mulch. What we learned was that with good growing conditions and with this spacing between tree rows the high strata tree canopies might close up rapidly (in approximately 2 years) hindering the horticulture production and demanding extra pruning work to maintain opening for light. We also realised that the horticulture production demanded more mulch more often than we anticipated.


Tree rows spaced 8 metres apart at Fazenda Bella.

As CEASA (a network of produce distribution and marketing) is on our way to deliveries and other jobs, we made contact with truck drivers that delivered produced and products wrapped in straw. This way our truck that would usually return empty to the farm, ended up returning with free mulch loads. An opportunistic resource in permaculture parlance.

A positive side of this tree row spacing was that although we have not focussed on any particular fruit trees to occupy the high strata, we did focus on citruses (medium strata) and coffee (low) and these species thrived in these systems. In these areas we added some jabuticabas every two coffee bushes in the understory.

Showing a stratified tree system at Fazenda Bella

Already taking into account the lessons from the implementation and management of the first plots, last year we implemented new systems in an area of about 1 acre (4.000m2). This time the tree rows were spaced 8 metres apart in order to keep the horticultural production thriving for longer.

Another choice made was to establish the tree rows from seeds. We used plenty of jack bean and pigeon pea as nitrogen fixers and mulch (green manure). We also decided to focus on one high strata tree for production. In this case, moringa oleifera.

An 8 months old moringa.

The tree rows developed very well from seeds and banana suckers. The horticulture consortiums also thrived between the tree lines and we also destined some areas to the cultivation of annuals such as cassavas, beans, pumpkins and okra. Within approximately 6 months with good production some cows broke the electric fence and feasted on our produce. Most of the annuals had already been harvested, but we lost many of the ‘gourmet’ produce we were growing for our clients. After strengthening the fence, reshaping the beds and replanting the area came rapidly back into production.


How input-dependent are these systems?

Nasturtiums and lamb’s-ear (Stachys byzantina) produced for restaurant salads at Fazenda Bella

When implementing these systems, the ideal situation is to face the plants and natural succession as tools to improve the energy flow (maximising photosynthetic processes), the mineral and nutrient cycles (using ‘chop and drop’ pruning techniques and mulching), the water cycle (through the way trees and their roots interact with the rain and the water tables), and the community dynamics (enriching biodiversity above and below the soil surface). The small farmers’ context, however, not always allow us to develop a process-based agriculture such as the one described above, and we find ourselves having to use inputs.

Eucalyptus pole used for construction at the farm.

Once more we decided not to plant grasses to use as mulch between the tree rows and amongst the produce beds. Instead we established grass rows around the whole area. We also began carting grass clippings from our firebreak that surrounds the agroforestry systems to minimise the high risks of bushfire.

We also began to produce our own organic fertiliser. We have been preparing bokashi and learning how prepare it according to specific needs and to the ingredients available.

Is it really necessary to work with the eucalyptus?
Note: This question was (and often is) made due to the fact that the eucalyptus is an exotic species in Brazil and its use in monoculture is known for degrading water tables.

Left row showing regrowth of eucaliptus pollarded at 6m high. Right row showing younger trees ‘skirted’ above bananas.

Once again the answer is contextual. In ideal conditions the eucalyptus grandis, the prevailing specie in these systems, may grow up to 4.5 metres per year. This tree is originally from Australia, a country with ancient soils, and where the indigenous population used fire to manage the landscape for thousands of years. Because it evolved in such conditions the eucalyptus grandis works well with degraded soils in Brazil, producing organic matter to improve and cover the soil.


Pruning of eucalyptus at 6 metres high. All organic matter is fed back into the system.
Natural building wall with eucalyptus poles.

At Fazenda Bella the apical cut of these tress is done when they reach about 6 metres high. It is common to be able to prune these trees 2 to 3 times every year after this first cut. We chose this height in order to use the trees as construction poles in the farm’s natural buildings. However, to keep pruning the trees at this height demands specific training and more time and energy spent in maintenance. For these reasons it is common to see the apical pruning of the eucalyptus in these agroforestry systems done at 2 to 2.5 metres high only. Another variable that weighs in the choice of species to compose these systems it the price and availability of the eucalyptus seedlings. The small farmer can easily find seedling of this exotic (in Brazil) for R$0.25 cents.

Detail showing regrowth after pollarding the eucalyptus and the ‘skirting’ (pruning of lower branches) to guide the system’s stratification.

The mutamba tree (Guazuma ulmifolia), also know as East Indian Elm or Bay Cedar, is endemic to the Central and South Americas and an excellent option of high strata tree for this kind of agroforestry systems. The mutamba tree grows almost as fast as the eucalyptus; produces a fruit from which a sweet nectar can be extracted; it is a bee forage; its bark, leaves and fruit have medicinal properties; it can be used as cattle and other livestock fodder; and it produces a better timber. The price of the seedling, however, is around R$15 reais and it is not as easily found. At Fazenda Bella, we are investing in seed harvesting and the production of seedlings to be used in the establishment of new areas.

Upcoming Agroforestry Courses


Forestry in Practice – With nearly 30 years of experience and literally thousands of farm plans under his belt, Darren J. Doherty comes to the Northern Rivers (NSW, Australia) to teach the course “Forestry in Practice“. The course equips farmers to holistically integrate trees in their landscapes and enterprises. Darren, who also works closely with livestock producers, has been an adamant promoter of tree integration for all farm enterprises. The course will runon the 9th and 10th of February, 2019, at the Holos Regenerative Design learning site in Brunswick Heads, Northern New South Wales.