Enhancement of amenity woodland field layers in Milton Keynes

Joanna Francis Alan Morton 01/04/2001
Anyone who lives in or visits Milton Keynes will be familiar with the grid road network that intersects the city. Along and within this transport system are extensive plantations of trees and shrubs that were planted during the early years of the city's development to filter and frame views, creating a visually stimulating experience and a network of wildlife corridors. A visit to these plantations in spring may surprise the majority of woodland habitat managers. Those expecting rank weed species to dominate the vegetation of young, developing plantations, will find instead a wealth of woodland field layer species growing vigorously beneath the trees. Indeed, some 125 ha of the city is plantations have now been enhanced by the introduction of native woodland species including Bluebell Hyacinthoides non-scripta, Red Campion Silene dioica, Ramsons Allium ursinum, Primrose Primula vulgaris, Pignut Conopodium majus, Woodruff Galium odoratum and Hedge Bedstraw Galium mollugo.
This article describes the experimental introduction of woodland field layer species into the young plantations. This experiment has been the focus of extensive research commissioned by the Development Corporation and, more recently, Milton Keynes Parks Trust, which has been responsible for management of the plantations since 1992. Between 1987-2000 this research has enabled the development of techniques for the successful introduction of woodland field layer species into young plantations and established secondary woodlands.
Natural development of woodland field layers  
The dramatic reduction of Britain's semi-natural woodlands during the 19th and 20th centuries has led recently to attempts to create new woodland habitats. In general, only trees and shrubs are planted, leaving the associated woodland fauna and flora to colonise naturally. Plantations that are isolated from existing semi-natural habitat are difficult to colonise and remain species-poor for decades, even centuries (Peterken & Game 1984). With only a limited ability to disperse and establish across intensively farmed land, many woodland plant species have been presented with insurmountable barriers to colonisation.
Accordingly, field layer communities in new woods are often poor and dominated by aggressive, shade-tolerant weed species that persist after canopy closure, such as Common Nettle Urtica dioica, Cow Parsley Anthriscus sylvestris, Bramble Rubus fruticosus and Ivy Hedera helix. The general lack of diversity in the vegetation is reflected in turn in the woodland fauna. Both barriers to colonisation and proliferation of shade-tolerant weeds are most severe in towns and on former agricultural land, where most amenity woods have been planted. Inevitably, ecologists have been drawn to the conclusion that, in addition to the planting of trees and shrubs, field layer species themselves would "have to be put there" (Buckley & Knight 1989), if the aim was to create a more complete woodland community.

Milton Keynes "City of trees"
During the development of Milton Keynes new town millions of trees and shrubs were planted. To promote rapid tree establishment, a weed-free policy was practised for a few years after planting and applications of fertiliser were made. By the early 1980s, when the plantations were approximately 10 years old, trees had established and the herbicide regime had stopped. Tree canopies include a broad mixture of native and exotic broad-leaved species and these, with groups of shrubs, provided a range of shade and leaf litter levels. Soils are clay loams or sandy clay loams, with neutral to slightly acidic pH, low organic matter and moderate phosphate levels.
Once the plantations came out of weed control, natural colonists began to reach the sites, but no woodland field layer species were present. The Development Corporation decided to take an experimental approach to woodland species introduction in the plantations in an attempt to increase both species diversity and plantation amenity value. After the success of a small-scale pilot study in 1981, the large-scale experiment was set up in 1987 to introduce seeds and plants of woodland field layer species into sites covering a total of 2.7ha.
The urban plantations are subject to continuing management by the Parks Trust. Due to the high density of the original planting, thinning begins early and is generally quite heavy. Every four to five years up to 70% of the trees are thinned and groups of shrubs are coppiced. Contractors remove the majority of the timber and chip the brash on site, raking out the piles of wood chips to an approximate depth of 10cm. Whilst as far as possible, the chips are targeted away from areas where woodland field layer species have been introduced, in practice chipping layers occasionally build up on and adjacent to the enhanced sites.
The large-scale introduction experiment in 1987
In 1987 the Centre for Ecology & Hydrology (then the Institute of Terrestrial Ecology), Monks Wood, set up the main experimental plots at six young plantations within the town. At each site 40 plots were marked out to include areas for restricted and main seed mixtures and plant introductions. In spring, woodland field layer species were sown by hand onto the surface of the plots, and in 1988 the young plants were introduced. No further aftercare was given to any of the introduced species. Species mixtures and densities are given in Table 1 and further details are given in Francis, Morton & Boorman (1992).

Table 1.
Field layer introductions made in 1987 and 1988

Species introduced as seeds in 1987
Main Seed Mixture
Sown at Low/High density, approx.
2/20kg ha-1
  • Garlic Mustard Alliaria petiolata
  • Giant Bellflower Campanula latifolia
  • Nettle-leaved Bellflower Campanula trachelium
  • Pignut Conopodium majus
  • Foxglove Digitalis purpurea
  • Hedge Bedstraw Galium mollugo
  • Woodruff Galium odoratum
  • Wood Avens Geum urbanum
  • Bluebell Hyacinthoides non-scripta
  • Hairy St. Johnís-wort Hypericum hirsutum
  • Oxlip Primula elatior
  • Red Campion Silene dioica
  • Stone Parsley Sison amomum
  • Betony Stachys officinalis
  • Hedge Woundwort Stachys sylvatica
  • Upright Hedge-parsley Torilis japonica
  • Wood Sage Teucrium scorodonia
  • Sweet Violet Viola odorata
Restricted Seed Mixture: Sown at approx.
20/200 seeds m-2
  • Ramsons Allium ursinum
  • Lords-and-Ladies Arum maculatumWater Avens Geum rivale
  • Primrose Primula vulgaris
  • Greater Stitchwort Stellaria holostea
Species introduced as plants 1988

Main Seed Mixture
Sown at Low/High density, approx.
2/20kg ha-1
  • Bugle Ajuga reptans
  • Columbine Aquilegia vulgaris
  • Garlic Mustard Alliaria petiolata
  • Giant Bellflower Campanula latifolia
  • Nettle-leaved Bellflower Campanula trachelium
  • Cuckoo-flower Cardamine pratensis
  • Enchanterís Nightshade Circaea lutetiana
  • Foxglove Digitalis purpurea
  • Hedge Bedstraw Galium mollugo
  • Wood Avens Geum urbanum
  • Bluebell Hyacinthoides non-scripta
  • Hairy St. Johnís-wort Hypericum hirsutum
  • Yellow Archangel Lamiastrum galeobdolon
  • Oxlip Primula elatior
  • Primrose Primula vulgaris
  • Red Campion Silene dioica
  • Betony Stachys officinalis
  • Hedge Woundwort Stachys sylvatica
  • Upright Hedge-parsley Torilis japonica
  • Wood Sage Teucrium scorodonia

Both seed and plant mixtures included species predominantly of local woodlands. The inclusion of some species not found locally possibly reflects the generally restricted availability of woodland species in quantity at that time, rather than their specific selection.

Survival and establishment
A multitude of very small seedlings appeared in the sown areas during 1987 and 1988 and the introduced transplants began to establish. Between 1990-92 the sown and planted plots were rigorously monitored and patterns of species establishment were soon apparent.

All of the eighteen sown species, except Narrow-leaved Bellflower Campanula latifolia, were recorded within the sown plots. In 1990 an average of 93 introduced plants was found in each m2, although this density varied greatly. Cover of the introduced species was also variable: generally it was highest in light areas and low in deep shade. By 1992, the woodland marginal species had high cover values and had already produced new populations of seedlings and young plants. In contrast, the true shade-tolerant species were mostly growing slowly, with species such as Bluebell still in the seedling phase.

In 1990 all of the species introduced as plants were present, though densities varied between species. Most species established quickly and many soon flowered, set seed and spread within the planting blocks. Again, different rates of establishment were identified: woodland marginal species produced very high seedling numbers, whilst the shade- tolerant species set less seed and subsequent germination rates were low. Within the latter group, primroses planted at high density produced significantly more seedlings than those in low density blocks. A third group of shade-tolerant species was distinguished by vegetative spread; even at low planting densities these species, such as Bugle, were very successful.

Long-term development: recording progress by 2000
The initial establishment was generally successful, but whether introduced species would persist in the longer term, during the widely varying conditions resulting from ongoing plantation management, was unknown. Regular field visits between 1994 and 1999 confirmed that the woodland species were surviving, but developments in terms of plant densities, cover values, or population structures were not recorded. In 2000, in the fourteenth field season since the initial introductions, the Parks Trust commissioned further replicated monitoring of both the original experiments and plant population expansion.

For the 2000 survey (Francis, 2000) three of the original enhanced sites were selected to represent open, semi-shaded and shaded conditions and three similar unenhanced sites were chosen to enable comparisons between the developing field layers. As in the previous surveys, field layer vegetation was sampled using random quadrats and percentage cover data were recorded. Characteristics of the non-seeded areas of the plantations were also monitored. Population structures for two contrasting species, Red Campion, a woodland marginal and Bluebell, a deep shade-tolerant species, were determined in detail.

Long-term survival and persistence of introduced woodland field layer species
Perhaps the most significant result was that 96% of the introduced species that survived in 1988 were recorded throughout the plantations in 2000. Only one species, Foxglove Digitalis purpurea, was no longer present at any of the experimental sites. Foxglove is a biennial that produces millions of seeds, but it requires over 90% light for germination. It is possible that plantations have become too dark, though dormant seeds may remain in the soil seed bank.

The majority of the woodland field layer species have not merely survived during the years since their introduction, but they have established and now form populations that include seedlings, young and mature plants. Very few, including Wood Sage Teucrium scorodonia and Nettle-leaved Bellflower Campanula trachelium, have been lost from one or two sites at which they were previously recorded. This may be, in part, due to increasing shade levels during the establishment phase. Significantly, the species that have not survived are those that are rare in or absent from local ancient woodlands, so justifying the need for careful species choice.
Measuring the enhancement of the field layer
Comparisons of enhanced and unenhanced plantations show that the 1987/88 introductions have had a marked effect on field layer diversity as measured in 2000. Enhanced sites had 39% more species that untreated sites, and 54% more shade- tolerant species. No significant differences in species richness were found between other components of the vegetation, including naturally colonising forbs, grasses and tree seedlings. The sown and planted species had not merely replaced those that would colonise naturally, but had added to them.

The most significant increase in species richness was achieved in plantations with very dense canopies, notably sites characterised by groups of poplars (Populus spp.) or Horse- Chestnut (Aesculus hippocastaneum), and a few stands (managed by the local authority) that have not been routinely thinned. Beneath the dense canopies shade levels are very high, leaf litter is deep and persistent and the diversity of natural colonists is typically very low. In such sites, introductions have more than doubled the number of species. Where natural vegetation cover is often lower than 10%, there is little direct competition to the introduction of true woodland species. After enhancement, over 70% of the existing field layer cover can be attributed to introduced species.

Spread into new areas
Only four species had spread significantly by 2000. Although most species have successfully established, subsequent dispersal and spread, even within favourable sites, is a slow process. This well illustrates the importance of Rackham's "inertia of vegetation" (Rackham 1980).

Only Wood Avens, a species well equipped for dispersal, has spread in all of the experimental sites and has a recorded frequency of over 98%. Red Campion and Garlic Mustard Alliaria petiolata both produce a great deal of seed with high germination rates. Where conditions are suitable, light for Red Campion and moderately shady for the Garlic Mustard, they have spread out of the original sowing plots.

The fourth species that has spread is Bluebell; a rather unexpected result for a species that is an indicator of semi-natural ancient woodlands in the Midlands and is only rarely recorded in recent woodlands. Bluebell has smooth, heavy seeds and only poor dispersal. In the young plantations its spread has probably been facilitated by specific site conditions: the slopes of noise attenuation mounds that allow seeds to roll away from the parent plants into new areas, accompanied by deep shade and leaf litter layers that are ideal for seedling growth.

Those planted species characterised by vegetative growth had all spread out of the original blocks by 2000. In 1988 the extent of the planting blocks containing the young plants was 4.8m2. Bugle now has a mean patch size of 7.2m2, and that for Yellow Archangel is 11.3m2. These species and Enchanter's Nightshade all flowered well in 2000 but no seedlings were recorded within or around the patches and spread has been predominantly vegetative.
Rates of establishment
Different establishment rates have been identified for the woodland marginal and deep shade-tolerant species. The slow-growing woodland core species have made particularly good progress since the last full survey in 1992, whilst many woodland marginals have decreased in cover.

At the beginning of the 1990s the populations of introduced field layer species in the young plantations were dominated by seedlings with only low numbers of mature plants.

In general, as the field layer species have matured, albeit at different rates, plants have flowered and produced their own seed forming second and third generations, and more. The 2000 survey has demonstrated that the populations are no longer dominated by seedlings and are therefore more sustainable, with a balance being achieved between different ages states. Now, high seedling mortality is far less critical as perennial parent plants can persist vegetatively through unfavourable conditions until the plantation environment is again suitable for reproduction.

Populations of woodland marginal species, such as Red Campion, were initially dominated by seedlings, but by 1992 immature and adult plants formed about 40% of the population and still do so. Between 1992 and 2000, the percentage cover of the woodland marginals has decreased and these species are now more restricted to more open areas.
In contrast, the shade-tolerant species, such as Bluebell, began with populations that consisted solely of seedling plants. By 1992, during a much slower establishment phase, only immature plants were recorded. By 2000, the population structure had completely changed, as during the intervening period plants reached maturity, flowered, set seed and formed new populations. Now, immature and adult plants form about 60% of the total population. For this group, including Woodruff, Pignut and Sweet Violet Viola odorata, percentage cover has increased as species have consolidated their positions and continued to establish and spread.

Indeed, Bluebell is one of the introduced species that has been most successful and that now characterises these amenity plantations. In early summer the swards of blue normally associated with mature, semi-natural woodland habitat are prevalent beneath the enhanced plantation canopies. The average population density in shady areas with deep leaf litter is currently over 600 plants per m2, with about 50 mature plants.

Effects of management
The anticipated problems associated with ongoing plantation management have not materialised. In fact, fluctuations in light and wood chip levels appear to benefit most of the introduced woodland species. It was expected that establishing seedlings and young plants would be vulnerable to competition from an influx of weed species in the increased light levels after canopy thinning. However, the increased light is accompanied by a layer of wood chips, a combination that seems to promote continued establishment, particularly for the deep shade-tolerant species that are adapted for growing in deep leaf litter. As they decompose the wood chip layers provide an excellent substrate for seed germination, enable seedlings and young plants to grow well and benefit from the high light levels, maintain soil moisture in dry spells, and, most importantly, reduce the colonisation of competing weed species. The marginal species, however, do not have these adaptations and seedling numbers are dramatically reduced in areas covered by deep wood chips. Most seedlings of these species are recorded where leaf litter or wood chips are very sparse with patches of bare earth in between them. Mulch has been used very successfully for the introduction of woodland field layer species in other schemes (Cohn & Packham 1993).

The plantation management continues to maintain the all-important site heterogeneity and provides a range of conditions for species introduction. This heterogeneity is vital for the successful development of a community of introduced woodland field layer species into new areas. It allows the use of species mixtures that contain plants with different requirements; each will find a suitable niche in which to establish.

Field layer community development
A general comparison of the NVC types in the seeded and unseeded plantations indicates that they are very similar and are best described as W24 Bramble-Yorkshire Fog (Rubus fruticosu-Holcus lanatus) underscrub. Because of the field layer introductions, there is a possible, but very weak, trend in the seeded areas towards NVC type W8a, a sub- community of the Ash-Field Maple-Dog's Mercury (Fraxinus excelsior-Acer campestre- Mercurialis perennis) woodland. This community type is a feature of the three local ancient semi-natural woodlands that are actively managed and regularly surveyed by the Parks Trust.

Any further development towards a W8 community type is, however, very unlikely, because most of the characteristic species of the W8a sub-community, e.g. Ground-ivy Glechoma hederacea, Dog's Mercury and Wood Anemone Anemone nemorosa, as well as a host of others, are missing from the seeded and planted field layers. Whilst a few may colonise naturally in time, most have very poor dispersal and propagation and would have to be introduced as young plants. Many have much more specialised requirements than the generalist species used for amenity enhancement and are not usually available in quantity from commercial sources. Their much more restricted distributions in the British flora should be maintained and their introduction would rarely be recommended.

Recommendations for field layer introductions
From the earliest stages of this work the introduced species fell into three distinct groups, that require different approaches to introduction (Table 2). The groups are based on the regenerative strategies of the field layer species and reflect their behaviour in natural woodland conditions. Additional species have been included in the groups as a result of further experimental introductions in woods in Berkshire (Francis 1993).
Table 2.
Species groups for the introduction of woodland field layer species into established plantations

  Light-demanding woodland marginal/ hedgerow species Deep shade-tolerant woodland core species
High germination rates
Quick to reach flowering maturity
High seed numbers
Low germination rates
long seedling phase;
Moderately high seed numbers
Predominantly vegetative
Spread by stolons/runners; or with very low germination rates
Introduction method
Low/medium rates:
3kg ha-1
High rates:
10kg ha-1
Low rates:
4-9 plants m-2
Examples for amenity

Alliaria petiolata
Arum maculatum
Brachypodium sylvaticum
Digitalis purpurea
Festuca gigantea
Galium mollugo
Geum urbanum
Hypericum hirsutum
Silene dioica
Stachys officinalis
Stachys sylvatica
Stellaria holostea
Torilis japonica

Allium ursinum

Conopodium majus

Fragaria vesca

Galium odoratum

Hyacinthoides non-scripta

Primula vulgaris

Viola odorata

Ajuga reptans plantations
Anemone nemorosa
Carex sylvatica
Circaea lutetiana
Euphorbia amygdaloides
Glechoma hedercea
Lamiastrum galeobdolon
Oxalis acetosella
Teucrium scorodonia
Veronica montana S
Viola riviniana
(Primula vulgaris)

Most woodland field layer species fit into these groups very readily. A few fall into more than one category, e.g. Primrose. In ideal conditions where there is a light plantation edge, but only a low level of competing vegetation, primroses can be successfully introduced using a high sowing rate. However, seed is very expensive and so if there are budget limitations and/or there are significant amounts of competing vegetation, then plug plants at high density of approximately 9 plants m-2 should be used.

Species to be introduced as seeds are best sown in late winter/early spring, as some require vernalisation to break dormancy. Plants may be introduced once the ground has warmed up but whilst it is still moist so that no artificial watering is necessary. It is best to select a range of common species that grow in local semi-natural woodland on the same soil type as the site to be enhanced. Within this context, choose a mixture of species with wide "ecological amplitudes" that will have varying requirements for light/ shade, moisture and pH. Sometimes, with appropriate permission, limited amounts of certain species may be harvested as seed from local woodlands; but plants may not be gathered legally from the wild. Alternatively, seed and plants should be sourced from a reputable supplier, insisting on the correct provenance to ensure the conservation of local genetic characteristics.

In Milton Keynes the field layer introductions are made once light levels beneath tree canopies are between 10% and 40% of daylight in summer. It is recommended that seeds and plants are targeted into areas where conditions are most suitable for their growth and establishment. This may be within a group of dense shrubs where leaf litter is accumulating, competition is low and shade levels are high. Such areas may form only 10%, say, of a plantation, but species introduced into these areas will thrive and spread into the rest of the site as conditions become suitable. Targeting specific areas in this way reduces costs to £60-100 per ha, and enables field layer enhancement to become a viable proposition. Field layer introductions can be made prior to the increase of shade (Anderson 1996), but such an approach may entail difficult and costly longer-term control of competing vegetation.

Long-term prospects
It is unlikely that the new plantations will become directly comparable with local ancient woodlands where woodland species diversity is so high. The introduction project has, however, been revealing and exciting because today the field layers in the amenity woodlands are perceived as being naturalistic and are extremely attractive, with benefits for other wildlife species. The experiment has demonstrated that woodland field layer species, including some ancient woodland indicator species, can be introduced successfully as seeds and/or plants into established plantations where they can form viable reproducing populations. The introduction methods are complementary to normal forestry practices and do not require any special treatments. The plantations have been directly enhanced by increasing inherent species richness and now provide an attractive amenity throughout the city. In time, these enhanced habitats may be colonised by more woodland plant and animal species and this is the focus of current research.

Part of this work was funded by an NERC CASE studentship held at Imperial College and within the Forest Research Co-ordinating Committee (FRCC) Special Topic Scheme on farm woodlands, with additional work funded by Milton Keynes Parks Trust. We are grateful to Mike Street and Rai Darke at MKPT and George Peterken for comments and advice.

References Anderson, P. 1996. The wrong trees and what about the shrubs? Enact 4: 20-22 Buckley, G. P. & Knight, D. G. 1989. The feasibility of woodland reconstruction. In Biological Habitat Reconstruction (ed. G P Buckley), 171-188. Belhaven Press, London Cohn, E. V. J. & Packham, J. R. 1993. The introduction and manipulation of woodland field layers: seeds, plants, timing and economics. Arboricultural Journal 17: 69-83. Francis, J. L., Morton, A. J. & Boorman, L. A. 1992. The establishment of ground flora species in recently planted woodland. Aspects of Applied Biology 29: 171-178 Francis, J. L. 1993. The introduction of woodland field layer species into secondary woodlands. PhD thesis, University of London. Francis, J. L. 2001. The woodland field layer introduction trials, Report 2000. Ecological Studies in Milton Keynes 127 Peterken, G P & Game, M. 1984. Historical factors affecting the number and distribution of vascular plant species in the woodlands of central Lincolnshire. J. Ecol. 72: 155-182

Rackham, O. 1980 Ancient Woodland. Edward Arnold, London

This article was originally published in April 2001 as:
Francis, J. and Morton, A. Enhancement of amenity-woodland field layers in Milton Keynes. In British Wildlife Vol. 12 (4), April 2001.

Joanna L. Francis and Alan Morton
Joanna Francis is a freelance ecologist based in south-west England. Alan Morton is a Lecturer in Ecology in the Biology Department at Imperial College, London. Flora locale in association with the Milton Keynes Parks Trust and Joanna Francis holds occasional training days, showcasing the woodland ground flora experiments, usually in early May. See the training section on www.floralocale.org for details.

This article has been reproduced with kind permission from the authors and the Milton Keynes Parks Trust. It was originally published in British Wildlife (Francis and Morton, 2001). Photos have been supplied by the Milton Keynes Parks Trust.  


Page last modified on Friday 11 of May, 2012 18:23:36 BST