Sunday, 15 April 2018

Week 2 - The Polyculture Project

It's been a glorious spring week in Shipka!  This week we've continued planting in Ataraxia, our new perennial polyculture trial garden, have been foraging in the mountains for wild garlic and continued preparing the beds in the market garden for the annual crops.

You can read about what we got up to in week one here and  here's what we've been up to this week.

The Perennial Polyculture Trial Garden

We started to plant out the biomass beds in Ataraxia and added some more productive and support plants into the polyculture beds.

Ataraxia - The Perennial Polyculture Trial Garden

We've selected a range of pioneer plants to grow in the biomass beds and will be studying the suitability of each species for use as mulch plants. The ideal plant should grow well on poor soils, have relatively low water demand, be fast growing and tolerant of regular pruning.

We are experimenting with three categories of plants in the biomass beds.  

Nitrogen Fixing Trees and Shrubs ​​ - Paulownia tomentosa - Robinia pseudoacacia - Alnus cordata - Elaeagnus umbellata  
C4 perennial grasses -  Miscanthus x giganteus - Aruno donax 
Fast growing  perennials - Symphytum x uplandicum ' Bocking 14' - Morus alba - ' Vratza 24'

The pioneer plants will be in blocks of 1.5 m wide  x 8 m long  as shown in the below image. 

The biomass plants are planted at 50 cm between plants, 3 rows per bed. This season the plants will establish good root systems and settle into the beds and next season we will begin various pruning regimes, weigh the biomass and measure regrowth rates of each species.  

Alex measuring out the spacing between the Paulownia tomentosa.

Angela planting out Narcissus poeticus - Poet's Narcissus as a bulbous layer in the polyculture beds.

This beautiful spring ephemeral attracts bees and other pollinators and is the species considered to be one of the first daffodils to be cultivated. It's likely to have adorned the gardens and window sills of the Romans, Greeks and perhaps the Egyptians and Sumerians before them.

We also added the shrub layer to the productive polyculture beds which consisted of Gooseberry between the Hazelnut in bed B and Loganberry between the Hazelnut in Bed D    

Here is an overview of the trial beds showing the biomass beds and the productive beds

The Market Garden

We picked up a few trucks full of spoiled straw from a local farmer and a truck load of composted farmyard manure for the market garden.   

The bed preparation work flow is as follows; sprinkle 100 -120 g of ash per m length of bed, fork over the beds to relieve compaction, spot weed the beds chop and dropping all plants back to the surface (apart from those with  rhizomes that should be removed), rake back surface mulch, deposit 20 L of compost, replace the surface mulch, wait for a heavy rain to saturate the soil and then top up the surface mulch.     

Mountain Hike and Wild Garlic Pesto

We took a hike up in the mountains to a spot where Wild Garlic -  Allium ursinum can be found hugging the forest floor and harvested some to make a wild garlic pesto from the leaves. 

This time of year the forest floor, still fully bathed in spring light from the lack of leaves on the Beech trees, is full with flowering spring ephemerals and other forest herbs. These plants play a vital role in the cycling of nutrients by fixing minerals into biomass during the dormant season thereby preventing them from leaching out of the soil profile during the winter rains and snow melt. As the deciduous plants emerge from dormancy and fill with leaves the ephemeral plants having completed their reproductive cycles for the season release their herbaceous tissue back to the forest floor and hibernate in bulbous form until the next season. This resource sharing in time and space is a vital part of effective polyculture design that we try to integrate into our landscape designs for gardens and farms. 

Some Spring Ephemerals and woodland herbs from the surrounding forests 

Wild Garlic Pesto 

What you need - Wild garlic (a few bags full!) lemons, walnuts, olive oil, salt, food processor or blender

What to do - Wash the garlic and put it inside the blender, pour in the oil, squeeze in some lemon juice, sprinkle some salt and walnuts, crumbly goats cheese  and blend it until it is a paste. Taste it and add whatever of the ingredients you feel it needs more of (which definitely won't be the garlic!) Spoon it into jars and put in the refrigerator. Eat it with toast, pasta, cheese and meats.

Here are some photos from around the gardens:

Pyrus communis 'Early Boliaka 'planted 2 years ago and flowering for the first time, hopefully we'll get some August Pear 

Gingko biloba emerging from winter dormancy  

Prunus spinosa - Sloe  under Juglans regia - Persian Walnut . Lots of these shrubs forming thickets on the edge of the native living hedges that surround the garden. They do not seem at all inhibited by the Walnut. 

If you are interested in what we are doing at the project and would like to learn more about regenerative landscape design we have a seven day course coming up in June. For more info click on the poster below.

Appreciate the work that we are doing? If so please consider donating to our Polyculture Project


Saturday, 14 April 2018

Week 1 - The Polyculture Project

It's been a great first week on the polyculture study, we're very happy to welcome Alex, Angela, Malcolm and Victoria to the project.

We started the week taking the soil samples in the new trial garden Ataraxia. The soil analysis is used to assess the Advance Planting Preparation Trial.

The Soil tests we use are from Northern Rivers, an excellent tool for assessing soils  

Advance planting preparation is basically the addition of organic matter into the planting zone 6 - 12 months before planting to improve soil conditions for the incoming plants. This can be in the form of mulches that suppress existing growth and decompose in situ or in the form of green manures that replace the existing growth and improve the soil. The aim of the trial is to discover methods that are inexpensive, time efficient, least disruptive to the existing wildlife and that provide the optimal conditions for the incoming plants.

The perennial polyculture trial beds -  6 beds - 1.5m wide and 25m  long with 50cm internal pathways and 1m wide periphery pathways

You can read more about this trail including the different bed preparation methods we used here.

We'll be publishing the results of these trials once we get the mineral analysis results back from the lab in the next few weeks.
Soil samples bagged and tagged

After taking the soil samples, we we're ready to starting planting out the beds. We marked out where the plants should be with stakes and loosened up the soil with the broadfork before digging the planting holes. The broadfork is an excellent tool for this job, the soils in this garden are very stony and difficult to dig and using the broadfork before digging the holes made the job much easier. 

Thanks again to Krasimir from Gligans Handmade Broadforks for donating this tool to the project. Krasimir makes these tools by hand and they are great quality and super useful. Check out his your video of his tools in action  and check their facebook page here

Malcolm on the broadfork, loosening the soil before the planting holes are dug 

We started to plant out the perennial polycultures in Ataraxia. The below bed is composed of hazelnut, wild garlic, walking onion, snowdrop, fumewort, wild strawberry and we'll be planting asparagus and Gooseberries in the coming weeks. This is the green manure bed and we planted straight into the cover. The green manure cover consists of Onobrychis viciifolia - Sainfoin , Trifolium repens - White Clover and a range of volunteer native plants yet to be identified.  

Angela planting Galanthus sp. and Corydalis sp. spring bulbs  

The Perennial Polycultures that we've been planting in the new garden -  more on this here

Alex and Victoria planting Cornus mas - Cornelian Cherry

We are growing three Bulgarian cultivars of Cornellian Cherry in the gardens. 

We also have these cultivars available from the nursery 

Cornelian Cherry - Cornus mas
Back at the house Dylan and Archie are potting up herbs in the nursery

I found this juvenile Slow Worm - Anguis fragilis under the plant pots in the nursery. Great to see these around the young plants as their main diet is slugs.

Over to the market garden at the end of the week to start preparing the annual beds for this season's crops. We removed some of last year's hardy vegetables, added 100 - 120 g of ash per m length, broad forked it over, pulled weeds and left them on the surface (removed the rhizomatous ones) and added 20 L of composted farmyard manure per m length of bed.

We leave our annual vegetable beds to naturally vegetate during the dormant season and do not disturb them until it is time to plant/sow the next season crops.

Annual ground covers such as the above Lamium purpureum, known as red dead-nettle and Myosotis spp. from the family Boraginaceae produce thick patches throughout the gardens providing an excellent source of nectar and pollen for early pollenisers and a great source of mulch.

Wood ash is applied to the raised beds, a hand trowel full (100-120g) per m length of bed.

Purple Kale, Parsnips and Broccoli left over from last season. Given the relatively mild winter we had this year the Kale and Broccoli have been producing all winter.

Many of the fruit trees are in blossom now, the spring is in full swing 

If you are interested in what we are doing at the project and would like to learn more about regenerative landscape design we have a seven day course coming up in June. For more info click on the poster below.  

Appreciate the work that we are doing? If so please consider donating to our Polyculture Project


Saturday, 7 April 2018

Invertebrate Biodiversity Survey Results - Polyculture Study June/July 2017

June/July 2017 results

By Christopher Kirby-Lambert – 03/04/2018

For May 2017 Results see here

A sample of invertebrates from the gardens
1. Survey Timing and Conditions
Field survey was carried out on the 29th of June (Perennial polyculture trial garden) and the 2nd of July (Market Garden). Both plots were surveyed in the morning between approximately 09.00 and 11.00. Conditions were warm, sunny and calm on the 29th of June and warm sunny and breezy on the 2nd of July.

2. Survey Plots
2.1. Market Garden

The Market Garden is a well-established and relatively complex plot supporting a range of habitats. Roughly 2/3 of the plot area is given over to agriculture. Half of this is polyculture beds used to grow a variety of vegetable crops. The other half is a forest garden. The market garden has been divided into three sub-plots for the purposes of sampling:
Sub-plot 1 - Forest Garden: Young planted fruit trees over a mix of native grass and herb species similar to that found in the Permaculture beds. There were some more mature trees providing shade in places but much of the sub-plot was open. During the June/July survey visit this plot supported a range of flowering plants providing abundant nectar sources for pollinators.
Sub-plot 2 - Polyculture beds: Rows of cultivated ground for a variety of vegetables. The plots supported a range of vegetable species as well as non-food plants, many of which were flowering and provided ample nectar sources. There are numerous wooden growth supports present in the plots which were noticeably being utilised by dead-wood nesting bees and wasps. Between plot rows there are grassy pathways with fringes of native flora. A pond is present at the edge of the area and is surrounded by a patch of coarse grassland with a significant herb component and numerous flowering plants. Overall, the flora in the non-cultivated areas is closest to that observed in woodland glades and rides nearby.
Sub-plot 3 - Scrub:  An area of relatively dense scrub and small trees containing a mix of species but dominated by Prunus sp. and Malus sp.. There is also a large Poplar (Populus sp) shading much of the area. The understory, where scrub was not so dense as to shade it out, was composed of grasses and woodland herbs. There is a damp area due to run-off from the neighbouring road that has been planted with Yellow Flag (Iris pseudacorus) and reed (Phragmites australis), although this is still in its infancy. Deadwood is fairly limited in this sub-plot.

2.2. Perennial polyculture trial garden

Photo by Christopher Kirby-Lambert

A recently acquired plot of land composed of coarse, unimproved grassland with scattered Rose (Rosa sp.) bushes surrounded by a well established hedge dominated by Elm (Ulmus sp.) and Hawthorn (Cratageus sp.). A large mature Walnut is present in the corner of the plot. Previous management was by goat grazing. Since acquisition of the plot a reasonably sized pond has been dug, but not yet filled, and several polyculture beds have been created. This plot was divided into two sub-plots for the purposes of sampling:

Sub-plot 4 – Grassland: This is currently a homogenous area supporting coarse unimproved grassland with a relatively limited herb component. Scattered low rose (Rosa sp.) scrub was present across the sub-plot. The plot showed a much greater abundance of nectar sources during the June/July survey and several species of flowering plant occurred relatively densely throughout the sub-plot.
Sub-plot 5 – Hedgerow: The sub-plot was surrounded on three sides by a large, mature, hedge. The principal tree species present were Elm (Ulmus sp.) and Hawthorn (Cratageus sp.). There were also significant components of smaller shrubs, predominantly bramble (Rubus sp.), blackthorn (Prunus spinosa) and rose (Rosa sp.). Although deadwood was present it was generally sparse and of small diameter.
Sub-plot 6 – New polyculture plot: Newly dug polyculture beds which, at the time of the June/July survey were almost entirely unvegetated. The tracks between beds as well as their margins did however support a range of ruderal plant species, many of which were flowering. The edges of the sub-plot, where it bordered sub-plot 4 were also flower rich and floristically diverse.
3. Target Groups
  1. Coleoptera (beetles): This encompasses a vast range of species with widely differing ecologies, however, the basic ecological niches of many groups are well established and they have been relatively well studied in Europe. In addition the group as a whole is well known to the surveyor and many species are likely to be identifiable to species. In practice the most frequently encountered beetle group in the survey plots by a wide margin were leaf beetles (Chrysomelidae). These species are invariably phytophagous (plant-eating), feeding on living plant tissue. They are often specific to certain plant species or families and so are generally ecologically informative. Weevils (Curculionoidea) were also collected in some numbers and have similar ecological niches. Other groups of beetles collected include saproxylic species that depend on deadwood resources for larval development and tend to feed on pollen from flowers and predatory species.
  2. Hemiptera; Heteroptera (true bugs): Another group that has been relatively well studied in Europe and is relatively well known to the surveyor. The Heteroptera include a wide range of families with varying ecologies. Most of those encountered were predominantly phytophagous and ranged from generalists to specialists on particular plant species.
  3. Hymenoptera; Aculeata: The aculeates include bees (Apoidea), ants (Formicidae) and a number of wasp families, all groups that are well known to the surveyor. Ants are ground or tree nesting and mostly predatory (although some European species will eat seeds). Bees feed on nectar (as adults) and pollen (as larvae) so are entirely dependent on flower resources for food. They nest variously in dead wood, bare ground and soil, and moss (many species are cuckoos, stealing the nests of the host species). Wasps are predatory and nest in a similar range of habitats as bees.
  4. Diptera (true flies); larger Brachycera and Syrphidae: Select groups of flies were recorded. These groups were limited to those that are known to the surveyor and may be ecologically informative in the current survey.
4. Survey Methodology
The field survey methodology utilises a number of widely used collection techniques that, in combination, collect species from most groups present on a site. The use of these techniques is dictated by the habitats present on each plot and roughly follows Natural England’s Common Standards Monitoring (CSM) guidelines1. These guidelines are intended for use in monitoring the quality of Sites of Special Scientific Interest (SSSI’s) in the UK. See Table 1 for a breakdown of the survey methods used on each sub-plot. The total survey time devoted to each sub-plot was the same.

4.1. Field survey techniques
Sweep netting
A lightweight folding circular aluminium frame 40 centimetres in diameter was fitted with a net bag supplied for sweep-netting by GB Nets and attached to an extending lightweight aluminium handle. Net strokes were reasonably rapid, and penetrated as far into the vegetation as possible without the stroke being seriously slowed by its resistance. A maximum of fifty sweeps (counted as single strokes of the net) was taken before examining the catch. The sample was initially examined in the net, noting or capturing large, fast-moving or readily identified species. The remaining net contents were then emptied onto a white tray, and the material in the tray examined for smaller and slower animals. 10 minutes of survey time was devoted to sweep-netting per sub-plot.

Targeted netting
Large, active, species, especially those prone to visiting flowers, resting on leaves, or with regularly visited and recognisable nests, are often most effectively recorded by netting individual animals. This is particularly effective for solitary bees and wasps, but also for some groups of flies. When utilised 30 minutes survey time per sub-plot was devoted exclusively to it. On those occasions, the net used for sweep-netting was employed

Vegetation beating
Samples were taken from tree and shrub foliage, ivy, and dense, tall herbaceous vegetation by holding a net under the foliage and tapping the branches or stems above sharply several times with a stout stick. The sweep net currently in use was most often employed for this purpose. For high vegetation and larger branches, a net with a lightweight folding frame 55 centimetres in diameter and a long bag was also used. This net has the advantage that substantial amounts of foliage can be inserted, or a substantial length of tall vegetation placed next to the net, before sampling. Material was initially examined in the net, then emptied onto a white tray for further sorting. When utilised 30 minutes survey time per sub-plot was devoted exclusively to it.

Active search
Features of significance to invertebrates which are not sampled, or not necessarily adequately sampled, by sweeping, beating or suction sampling were investigated by close examination and hand searching. Attention was particularly paid to: accumulations of plant litter; dead wood; the ground beneath wood, stones and other debris; fungal fruiting bodies; tree trunks; the undersides of plant rosettes; and bare wet ground. When utilised 30 minutes survey time per sub-plot was devoted exclusively to it.

Table 1. Distribution of sampling effort across plots. X denotes that the technique was used in the sub-plot in question.

Market Garden
Perennial Polyculture Trial Garden

Sub-plot 1
Sub-plot 2
Sub-plot 3
Sub-plot 4
Sub-plot 5
Sub-plot 6
Sweep-netting (10 mins)
Targeted netting (30 mins)


Beating (30 mins)



Active search (30 mins)
Total survey time
70 mins
70 mins
70 mins
70 mins
70 mins
70 mins

4.2. Lab methodology

Voucher specimens of all encountered species in the aforementioned target groups were collected and retained. Specimens were collected using a pooter, or individual tubes in the case of larger individuals, before being asphyxiated using ethyl acetate. Collection of specimens was limited to the minimum required to ensure a reasonable chance of identification. Specimens were stored in plastic boxes layered between tissue paper and frozen to preserve them until lab examination could occur.

All collected specimens were examined in the lab using a 7-45x magnification binocular microscope. Attempts at identification have been made using a combination of relevant reference books, scientific papers and web-based resources. Whilst the aim of the identification process is to reach a confident species level determination, in practice this has not yet been possible for many species. Most species have been determined to at least genus level, and identifiably different species within a genus have been differentiated numerically, e.g. Bombus sp. 1 and Bombus sp. 2, when species level determination has yet to be achieved. This methodology means that there is the potential to underestimate diversity by missing cryptic species that are near identical in outward appearance to others in the same genus. The identification process will be an ongoing endeavour and specimens will be re-visited periodically in an attempt to identify, with a reasonable level of confidence, the exact species present on the plots. All collected specimens are either glued onto card or directly pinned and labelled for future reference.

See June /July 2017 Sheet in bottom left corner  

6. Notes on June-July results

The June-July survey recorded a total of 237 species in 45 families from the targeted groups.

The most species rich of the target groups were the Coleoptera with 107 recorded species in 22 families. The Chrysomelidae (leaf beetles) were again the most diverse group of beetles recorded with a total of 20 species. Other relatively species rich groups were the Curculionidae (true weevils) with 14 species; the Coccinellidae (ladybirds) with 14 species and the Cerambycidae (longhorn beetles) with 11 species. The Curculionidae feed on living plant tissue, the Coccinellidae are predominantly predatory whilst the Cerambycidae are saproxylic (dead wood feeders) as larvae and feed on nectar and pollen as adults. All of the remaining families of beetle were represented by 10 or fewer species. These families are a mix of phytophagous (live plant), saproxylic (dead wood) and carnivorous feeders.

The aculeate Hymenoptera were the second most speciose group with a total of 69 recorded species, a noticeable increase on the number recorded in May. The most species rich family during the June/July visit were the Halictidae (furrow bees), with 14 recorded species. Species from the families Colletidae (9 spp.), Megachilidae (9 spp.), Andrenidae (8 spp.), Crabronidae (8 spp.) and Formicidae (8 spp.) make up the majority of the remaining Hymenopteran recorded during the June/July survey. The furrow bees (Halictidae) are all ground nesting species. The Colletidae recorded were all members of the genus Hylaeus which includes burrow and deadwood/plant nesting species. The mason bees (Megachilidae) nest in either dead wood or walls whilst Mining bees (Andrenidae) are entirely ground nesting, generally in open, exposed soil or sand. The Apidae were also still relatively diverse (7 spp.) and composed of a mix of ground nesting and cavity nesting species. All the remaining families were represented by 1-3 species.

Only 26 species of Heteropteran bug were recorded during the June/July survey. shield bugs (Pentatomidae), 10 spp., were the most species rich groups of bugs recorded. The Pentatomidae predominantly feed on sap from plant stems The ground bugs (Lygaeidae), 4 spp.,. The Coreidae (4 spp.), Rhopalidae (3 spp.), Nabidae (2 spp.), Scutelleridae (2 spp.) and Alydidae (1 spp.) were all significantly less diverse but included species that were abundant on both plots.

6.1 Plot 1 - Market Garden
Sub-plot 1 – Forest Garden
  • Highest species richness of any sub-plot (75 spp.)
  • Reasonably diverse Coleoptera (26 spp.)
  • High diversity of Hymenoptera (26 spp.)
  • Diverse flora and structure - rough grassland, tall herbs, young trees, shrubs
  • Reasonable diversity of nectar feeding Hymenoptera (pollinators) due to abundant nectar sources
Sub-plot 2 – Polyculture beds
  • Lower species richness than Forest Garden (67 spp.)
  • More intensive food production than forest garden means that structural and floristic diversity is lower
  • Both planted and wild plants in polyculture beds provide a good nectar source when in flower
  • Reasonably diverse Coleoptera (24 spp.)
  • Abundant nectar feeding bees (24 spp.) which act as pollinators
  • Wooden growth supports provide nesting sites for wood nesting bees and wasps, particularly mason bees in the genus Osmia
Sub-plot 3 – Scrub
  • Predominantly homogenous young shrubs and trees with one or two more mature trees
  • Low species richness (35 spp.)
  • Dominated by Coleoptera (17 spp.) but few (5 spp.) saproxylic species present

6.2 Plot 2 - Experimental Polyculture
Sub-plot 4 – Grassland
  • Relatively simple in structure, being predominantly rough, previously goat-grazed, grassland with scattered low shrubby bushes
  • High species richness (78 spp.) 
  • Numerous patches of flowers attract nectar feeding bees (17 spp.)
  • Most of the remaining species are phytophagous (22 spp.), feeding on either leaves or seeds
Sub-plot 5 – Hedgerow
  • Lowest target group species richness of sampled sub-plots (33 spp.)
  • Dominated by Coleoptera (18 spp.), of which half (9 spp.) are saproxylic for at least part of their life cycle
  • Mature hedgerows with mature elm and blackthorn predominant
  • Some deadwood present providing habitat for saproxylic species
Sub-plot 6 – New Polyculture plot

  • Highest diversity of sample sub-plots (91 spp.)
  • Most diverse group are the Coleoptera (49 spp.)
  • Very diverse Aculeate Hymenopteran fauna (30 spp.), including 22 spp. Of pollinating bee
  • Abundant ruderal plants in disturbed ground
  • Floristically diverse marginal vegetation attracted many nectar and pollen feeders

If you appreciate the work that we are doing please consider donating to our Polyculture Project. If everyone that visits our blog and website were to donate just €1 we would fund our three year research project in no time.