Biological Flora of the British Isles: Pteridium aquilinum (L.) Kuhn


  • R. H. MARRS,

    1. Applied Vegetation Dynamics Laboratory, School of Biological Sciences, Liverpool L69 7ZB, UK, and The Botany School, University of Cambridge, Downing Street, Cambridge CB2 3EA, UK
    Search for more papers by this author
  • A. S. WATT

    1. Applied Vegetation Dynamics Laboratory, School of Biological Sciences, Liverpool L69 7ZB, UK, and The Botany School, University of Cambridge, Downing Street, Cambridge CB2 3EA, UK
    Search for more papers by this author

  • *

    Abbreviated references are used for many standard works: see Journal of Ecology (1975), 63, 335–344. ‘A.S.W.’ refers to the unpublished material held in the A. S. Watt bracken reference and manuscript collection at the University of Liverpool. Nomenclature of vascular plants follows Flora Europaea and, where different, Stace (1997).

R.H. Marrs (tel. +44 151 7955172; fax +44 151 7955171; e-mail


  • 1This account reviews information on all aspects of the biology of bracken Pteridium (mainly aquilinum ssp. aquilinum) that are relevant to understanding its ecological characteristics and behaviour. The main topics are presented within the standard framework of the Biological Flora of the British Isles: distribution, habitat, communities, responses to biotic factors, responses to environment, structure and physiology, phenology, reproductive characters, herbivores and disease, history, and conservation.
  • 2Pteridium is a complex genus comprising a number of species, subspecies and varieties. The treatment here is based on a recent revision that incorporates both morphological and molecular data, and is related to its geographical distribution.
  • 3Pteridium is thought to be a woodland genus, but it can grow in the open. It is cosmopolitan and occurs on all continents except Antarctica. It responds to human disturbance and is often found in open spaces after forest clearance and cultivation. In some situations it can be a troublesome weed, causing problems for land managers. Moreover, its abundance and distribution in Britain are predicted to increase as a result of global climate change.
  • 4Pteridium aquilinum ssp. aquilinum, the most common taxon in the British Isles, occurs in many plant communities, and it is apparently limited by frost and waterlogging. Its abundance has probably increased in the relatively recent past as a result of changing land management, and this increase impinges on plant communities with a high conservation interest. The changed land management reflects changing use of agricultural land and also a reduction in the use of Pteridium as a resource. Accordingly, in many places Pteridium is viewed as a weed and management is needed to control it and restore more desirable vegetation. These management techniques are summarized.

Dennstaedtiaceae (Hypolepidaceae). Pteridium aquilinum (L.) Kuhn (Pteris aquilina L., Eupteris aquilina (L.) Newm.), bracken, eagle fern. Fern with extensive, repeatedly branched, underground rhizome system; rhizomes variably tomentose but with no scales. Fronds large (15)30–180(440) cm, usually tri-pinnate but variable, ranging from two- to four-pinnate, although where present the fourth-order subdivision often only partially demarcated (Thomson 1990). Fronds solitary, pubescent when young, dying in autumn in northern Europe. The stipe about as long as blade, up to 1 cm in diameter; it is dark and tomentose at base, half-cylindrical and glabrous above (Fl. Br. Isl.). Blade shape more or less deltoid in outline, and pinnae lanceolate or oblong, the margin revolute, veins free; sori marginal and mostly continuous, with the sporangia borne between the outer indusium, the modified margin and an inner indusium (Tryon 1941). Variable number of sori produced, absent in many years. Spores brown and very finely spinulose, tetrahedral-globose and produced in July and August in northern temperate conditions (Tryon 1941). Called eagle fern because the distribution of vascular bundles looks like an eagle in cross section (or an upturned oak tree) (Thomson 2004).

Pteridium has a curious combination of relatively primitive external characters and relatively advanced internal ones (Bower 1928). Within the Pteridophyta, Pteridium is considered relatively primitive on the basis of: presence of a vestigial indusium, initial basipetal succession of sporangia, hairs but not scales on the rhizome, equal dichotomy of the axis in its early development and open venation. However, the highly developed vascular system with true vessels through the stipe and lamina are indications of an advanced structure. This vascular system is presumably one of the reasons why Pteridium can grow to such a large plant, and be such a pernicious weed.

(a) taxonomy and evolution

Taxonomic overview

Pteridium is an isolated and well circumscribed, cosmopolitan genus; however, the subgeneric taxonomy has proved complex, difficult and controversial, for at least three reasons (Tryon 1941; Thomson 2000a, 2005). First, there is considerable phenotypic plasticity, with evidence of environmental influences on characters used in taxonomic descriptions; second, there are many intermediates between the described taxa; and, finally, there are few good characters for separating taxa below the subspecies level (Thomson 2005). The first comprehensive attempt (Tryon 1941) to describe Pteridium used a series of morphological characters, including: rhizome hairs, hairiness of the rachis and frond undersurface, hair distribution, angles of pinnae to the mid-nerve, shape of the ultimate segments, including their division and mode of attachment, the relative sizes of the inner and outer indusia, and the sequence of unrolling of the frond. Tryon's review considered Pteridium as a single species (P. aquilinum) with two subspecies (sspp. aquilinum and caudatum), containing 12 varieties between them. Since then there have been many reviews (Page 1976, 1994), with new species and subspecies both described (Page 1989, 1994; Page & Mill 1995a,b) and refuted (Ashcroft & Sheffield 1999; Karlsson 2000).

Recently a comprehensive review of the genus Pteridium using a combination of morphology (usually basal pinnae and basal pinnules), choloroplast genome sequencing and Arbitrarily Primed Polymerase Chain reactions has been completed (Thomson & Alonso-Amelot 2002; Thomson 2004; Thomson et al. 2005). This large study has to a large extent confirmed Tryon's (1941) conclusions, but has standardized objectively the subspecific and varietal rankings, and tidied up the nomenclature. An important result was that morphotypes within biogeographical regions (North America, Europe, Asia and Africa) were in general more similar genomically than they were to phenotypically similar forms from outside their geographical region (Table 1).

Table 1.  Summary of the current taxonomic treatment of Pteridium and approximate geographical distribution (after Thomson et al. 2005). The unconfirmed potential variety reflects specimens that appear very close to the subspecies but require further research
SpeciesSub-speciesUnconfirmed potential varietyPloidyBroad geographical distribution
  • *

    Strictly this taxon has two valid names P. feei and P. aquilinum var. feei. However, recent molecular work (J.A., Thomson, unpubl. data) suggests that this taxon is close to the North American P. aquilinum sub-species and is likely to be re-designated as shown provisionally here.

  • Pot. var. fulvum variety has been found on steep east-facing banks of the River Tummel, Perthshire Scotland.

  • Pot. var. osmundaceum has been recorded from the south and eastern end of Loch Tummel, Perthshire, and elsewhere in the Grampian Region, Scotland.

P. caudatum  2n = 208Central and South America
P. esculentumarachnoideum 2n = 104Central and South America
esculentum 2n = 104New Zealand and Australia
P. semihastatum  2n = 208Asia/Australasia
P. aquilinumlatiusculum 2n = 104North America
lanuginosumpubescens2n = 104North America
pseudocaudatum 2n = 104North America
*feei 2n = 104Central America
japonicum 2n = 104North-east Asia
decompositum 2n = 104Hawai’i
wightianum 2n = 104South-east Asia
aquilinumfulvum2n = 104West Europe–North Africa
pinetorumosmundaceum2n = 104North and Central Europe
capense 2n = 104Sub-Saharan Africa
centrali-africanum 2n = 104Sub-Saharan Africa

Thomson's revision recognizes four taxa at species level; two diploid species (P. aquilinum), a northern hemisphere lineage consisting of 11 subspecies, and a predominantly southern hemisphere group (P. esculentum) with two subspecies; and two allotetraploids, P. caudatum of Central and South America and P. semihastatum of south-east Asia and Australia (Section VI, Thomson & Alonso-Amelot 2002). The re-assessment confirms Tryon's separation of the genus into northern and southern hemisphere taxa, but separates them at species level. The southern hemisphere species (Tryon's P. caudatum) has been renamed P. esculentum, as P. caudatum had been shown to be tetraploid genomic hybrids of a northern P. aquilinum subspecies and P. esculentum ssp. arachnoideum. Tryon's northern hemisphere pan-boreal var. latiusculum has now been divided into three subspecies; ssp. latiusculum sensu stricto of North America, ssp. pinetorum of Northern Europe and ssp. japonicum of north-east Asia (Thomson 2004). Two African subspecies were identified: sspp. centrali-africanum and capense; both are closely related to each other and to European sspp. aquilinum and pinetorum. The Hawai’ian endemic taxon was retained as ssp. decompositum.

The allotetraploid P. semihastatum overlaps in range over south-east Asia and northern Australia with its diploid southern hemisphere progenitor (P. esculentum ssp. esculentum) but it is also broadly coextensive with its northern progenitor P. aquilinum ssp. wightianum. For both allotetraploid species, there is considerable heterogeneity, suggesting both species had multiple origins (Thomson 2000a,b). As both also replace one, or both, of their diploid progenitors in certain situations, this implies a selective advantage (Thomson & Alonso-Amelot 2002). These authors also speculate that hybridization between southern and northern lineages might have provided species evolved in temperate climates that had an enhanced ability to colonize more tropical conditions.

British taxa

Within Britain, the majority of Pteridium is P. aquilinum ssp. aquilinum but other taxa have been recorded. Page & Mill (1995a,b) described some of these plants as species or subspecies (e.g. P. aquilinum ssp. atlanticum and ssp. fulvum; P. pinetorum and P. pinetorum ssp. osmundaceum). There has been considerable dispute over ssp. atlanticum, a plant found on Arran, North Ayrshire (Ashcroft & Sheffield 1999), and in the recent revision P. pinetorum is reduced to a subspecies of P. aquilinum; this taxon has been reported from a single locality near Aviemore, Highland Region, where it formed a lower more open canopy than ssp. aquilinum. Both ssp. fulvum and ssp. osmundaceum are now categorized as potential varieties within P. aquilinum ssp. aquilinum and P. aquilinum ssp. pinetorum, respectively; both have been recorded in Scotland (Table 1). Additionally, a triploid plant has been recorded from Clocaenog Forest, North Wales (Sheffield et al. 1993). Irrespective of the taxonomic controversy, within the British Isles there is evidence of the variation and speciation that are found elsewhere.

(b) biogeography

The current world-wide geographical range of the various taxa noted above (Table 2) is clearly bound up with evolutionary processes. The current taxonomic division at species level reflects a combination of geography and time since separation of Gondwanan and Laurasian lineages, and alloploidy. Below the species level, evolution has apparently occurred through geographical separation, with different sets of subspecies found in North America compared with Europe and north Asia, and distinct subspecies found in sub-Saharan Africa and on the islands of Hawai’i. However, within all three geographical groups (North America, Europe/Asia and South Africa) there are several closely linked subspecies indicating ongoing speciation (Thomson et al. 2005), although as there are range overlaps it is not due to simple allopatric separation. There is some evidence that, in South America, the allotetraploid species (P. caudatum) is found at lower altitudes than the diploid P. esculentum ssp. arachnoideum, although there is some overlap (Thomson & Alonso-Amelot 2002), and in sub-Saharan Africa P. aquilinum ssp. centrali-africanum is locally parapatric with ssp. capense but it tends to be tropical in distribution especially in the drainage basins of the Congo and Zambezi rivers (Thomson et al. 2005). The environmental factors affecting the distribution of other overlapping taxa remain unknown.

Table 2.  Geographical notes on Pteridium world-wide; abstracted from Page (1976) but modified to account for revised taxonomy (Table 1), and supplemented by information from Tryon (1941), Thomson & Alonso-Amelot (2002) and Thomson et al. (2005). n.d. = no data.*, for this taxon nomenclature is under review (Table 1)
SpeciesSub-speciesRangeHabitat notesAltitude (m)
P. caudatum Central and South America, does not occur at high latitudesClearings, pastures, dry hillsides, cleared forests600–2000
P. esculentumarachnoideumCaribbean, Central and South America extending further South than P. caudatum into Paraguay, Uruguay and north ArgentinaClearings, pastures, dry hillsides, cleared forests500–3000
esculentumAustralia, New Zealand, New Caledonia, Polynesia and MicronesiaOpen forests, dry slopes, dry rain-forest edges, sometimes in dense forestTo 1350
P. semihastatum North Australia, Philippines through to norh IndiaThickets, scrubland clearances, open slopes and woodland edges0–2500
P. aqulinumlatiusculumBoreal but extend southwards to Oklahoma and Tennessee, isolated elsewhereSunny, dry slopes, low stature woods and acid, sandy soils0–2700
lanuginosum-pubescens, in Page (1976)American western states from Alaska towards California and Mexico and east to Wyoming, Colorado and western TexasMoist-dry woods, clearings, pastures, thickets and woods0–3250
pseudocaudatumConfined to eastern coastal plain Cape Cod to Florida; inland sparsely across Texas, westward to IllinoisOpen woods, pastures, burnt areas and old fields, usually dry, poor soilsn.d.
*feeiMountain ranges of Mexico, Guatemala and HondurasOpen spacesTo 2800
japonicumNorth-east Asia and JapanTemperate deciduous forests and cool-temperate grasslandsn.d.
decompositumHawai’ian endemicDry open forest, scrub and forest margins300–3000
wightianumHimalayas, east to Taiwan, south to Sri Lanka, and throughout Indonesia to PhilippinesDry hillsides, jungle clearings, abandoned cultivated areas, volcanic craters and grassland700–3300
AfricanaquilinumThroughout Europe and into North Africa, including Azores and CanariesWoods, thickets, pastures, recent burns and is common on dry, open places0–600 and to 1800 Alps, possibly up to 3100 on mountains
pinetorumBoreal, often confused with ssp. aquilinum, north Scandinavia to UralsConiferous and deciduous woodland, coppices and slopes, dry sandy soils, unusually on calcareous soils (Komarov 1968). Common open sub-alpine Abies and Larix forestsn.d.
capenseSub-Saharan Africa. Primarily in south-west AfricaOpen grassland, dry, moderately light woods and virgin forestTo 1400
centrali- africanumSub-Saharan Africa, principally in Congo and Zambezi watershedsAs capense but more tropicalAs above

Given the taxonomic uncertainties, the geographical distribution, habitat preferences and evolution within Pteridium require considerable further study.

(c) general remarks

From a pragmatic viewpoint, we recommend Thomson's revision (summarized Table 1) as the working standard. However, most of the current literature either assumes that Pteridium is either a monotypic genus, or an older name has been used. Accordingly, to avoid confusion, in this review, we use either the generic name Pteridium, or unless otherwise stated, we cite the name used in the original manuscript in quotation marks. However, most research quoted here is for Pteridium aquilinum ssp. aquilinum. Similarly, we suggest the terminology on bracken anatomy outlined in Thomson (1990) be adopted.

I. Geographical and altitudinal distribution

Pteridium is the most widely distributed Pteridophyte, indeed it is one of the most widely distributed vascular plant species. It has a world-wide range, being found on all continents except Antarctica, and it is excluded only from desert regions, high mountains and from some areas of the tropics (Page 1976). Within the British Isles, Pteridium is ubiquitous (Fig. 1) and in Europe it has a widespread distribution (Fig. 2). It is excluded from mountain tops and the extreme north of Fennoscandia, Iceland, Spitzbergen, Greenland, the Faeroes and some parts of southern Europe, for example parts of eastern Spain and the central Balkans (Huntley & Birks 1983). Pteridium was originally a component of open forest communities long before the coming of man and his agriculture, but its range has expanded markedly as a result of man's activities.

Figure 1.

The distribution of Pteridium aquilinum in the British Isles. Each dot represents at least one record in a 10-km square of the British National Grid: (○) pre-1950, (•) 1950 onwards. Mapped by H.R. Arnold, using Dr A. Morton's DMAP software, Biological Records Centre, Centre for Ecology & Hydrology, Monks Wood, mainly from data collected by members of the Botanical Society of the British Isles.

Figure 2.

The European distribution of Pteridium aquilinum (•) on a 50-km square basis; ▴, subsp. brevipes. Reproduced from Atlas Florae Europaeae, vol. 1, by permission of the Committee for the mapping of the Flora of Europe and Societas Biologica Fennica Vanamo.

Pteridium occurs from sea level to over 3000 m, essentially remaining in a temperate climate, with increasing altitude at lower latitudes. The recorded altitudinal limit for Pteridium in the UK is c. 600 m: 585 m on Lochnagar, Highland Region, Scotland (Pearman & Corner 2004) and 610 m on Cairngorm, Highland Region, Scotland (A.S.W.).

II. Habitat

(a) climatic and topographical limitations


There are few parts of the British Isles where temperature permanently excludes Pteridium, the exceptions being above the altitudinal limit and in some low-lying frost pockets such as in East Anglia. The extreme north of Britain is near its geographical limit, e.g. the Island of Foula (Lat. 60°9′, 26 km west of mainland Shetland, altitude c. 61 m OD) where it was nearly exterminated after the severe winter of 1917 (Braid 1937). In 1956, no live Pteridium, but dead remains were found there, and it was thought that it disappeared after the severe winter of 1947 (Messenger & Urquhart 1959). Further, surveys in 1959 and 1969 failed to find it, suggesting that Pteridium has been eradicated (Hawksworth 1969). It can therefore be assumed that Pteridium's latitudinal limit is related to winter frost. No Pteridium is found in the Faroes, Iceland, Spitzbergen and northern Scandinavia (Watt 1950).

Temperature is seen to be more important when considered as number of day degrees. At its altitudinal limit in the Cairngorms light is greater than at sea level in the south, but the number of day degrees is c. 600, which may be close to the lower limit.

Rainfall and humidity

Pteridium is not excluded by rainfall or atmospheric humidity from any part of Britain although the amount and incidence of rain affects frond height and density. Pteridium tends to cover greater areas in the west of Britain, which might be attributed to: (i) the more oceanic climate of the west producing favourable conditions for establishment from spores, (ii) the continental climate of the east reducing Pteridium production because of increased incidence of frost, and (iii) the greater availability of uncultivated land in the west. Stunted fronds have been reported under drought conditions on the Burren, Ireland (Webb 1962).

In Quercus petraea woodland Pteridium receives 3.7% of the total incident rainfall per annum (Carlisle et al. 1967). Potential evaporation (interpolated) varies from an annual average of 380 mm in Easter Ross (Scotland) to 843 mm in the Channel Islands (Hulme & Jenkins 1998).

Exposure to wind

Density and height of Pteridium are both affected by exposure to wind. It is absent from the most exposed places at all altitudes from sea level to altitudinal limits (Leach 1930; Braid 1934; Conway 1949; Barkley 1953), and exposure was considered the general factor limiting Pteridium at altitude (McVean & Ratcliffe 1962). Gillham (1955) correlated its behaviour at three sites differing in exposure. With increasing exposure, there was a reduction in mean frond height from 76–91 cm to 61 cm to 8–30 cm. In the most exposed places, Pteridium patches were confined to hollows and sheltered places only; the hollows provided sufficient shelter for a sparse population (Gillham 1955). Change in exposure is often correlated with other environmental factors, e.g. at Braunton Burrows, Devon, frond height reduced with increasing exposure up a dune slope, but this may also have been affected by the increasing distance of the Pteridium roots from the water table (Willis et al. 1959a).

Wind has both mechanical and physiological impacts. Fronds tend to break at the point of insertion, shortening their productive life. However, with increasing exposure there is a decrease in (i) frond height, (ii) the size of all its parts, (iii) a change in frond form and habit, and (iv) increased xeromorphism (Bright 1928). In severely exposed places, frost and exposure to wind, acting separately or together, probably exclude Pteridium from localities on the west coast of Britain, exposed islands and higher altitudes. Spray interacts with exposure and Pteridium is often brown after September storms on Pembrokeshire cliffs (A.S.W.). Salt-laden gales led to withering of the fronds and there is subsequent degeneration of the rhizomes leading to a grass community (A.S.W.).

Light flux

Pteridium can survive in a range of light flux densities from heavy shade to full sunlight. It thrives when fully exposed to daylight. The duration of daylight is not usually a limiting factor in the north, because day length is greater from March to September than in the south, where fronds remain green for longer. A mean total daily photosynthetically active photon flux of 4.23 mol m−2 day−1 was measured for ‘P. aquilinum var. esculentum’ growing under Pinus radiata D. Don, and this was six times greater than the flux reaching Blechnum discolor (Forst. f) Keys growing under Nothofagus truncata (Col.) Ckn (Hollinger 1987). Hollinger measured the light compensation point at 11.6 ± 1.5 µmol photons m−2 s−1.

The presence or absence of Pteridium, its vigour and its density are all influenced by light. It increases in density and vigour in woodland gaps. In evergreen pinewood, Pteridium is confined to gaps (mean relative light 24%) and absent where the mean is 17% (Summerhayes et al. 1924). In oak and birch woods the light reaching Pteridium varies from 7.3% to 53% (Hopkinson 1927), but individual fronds persist when relative light is as low as 4.1% (Salisbury 1918). Dring (1965) found a reduction of more than 55% in spore production in dense shade.

Slope and aspect

Pteridium is commonly found on slopes (Elgee 1914; Hughes & Aitchinson 1986), e.g. in the Peak District Pteridium is restricted to the steeper slopes (> 20°) and within the north-east and south-west quarters. It was most common on south-facing slopes at higher altitudes within this area (Grime & Lloyd 1972). At lower altitudes aspect is less important (Lloyd 1972).

(b) substratum

Parent material

Pteridium grows in soil varying widely in both physical and chemical composition (Watt 1940), and there is little evidence that mineral composition is a major factor in influencing Pteridium distribution. The general impression of the grazier and the forester is that it grows best on, and is an indication of, deep, well-drained loam or sandy loam, but it is excluded from saline and waterlogged soils. The oft quoted Welsh saying of ‘Newyn dan y grug, Arian dan yr eithin, Aur dan y rhedyn’ (‘Famine under the heather, silver under the gorse, gold under the bracken’) confirms this view (Smith & Taylor 1995).

Pteridium occurs on a wide range of soil profiles and soil types, from sandy soils (> 90% sand with negligible clay) to those with up to 39% clay. In the soil survey of Wales, Pteridium was found growing on six soil groups, confirming its wide tolerance of soil conditions; these were: raw skeletal soils; shallow lithomorphic soils; deep, permeable brown earths; acid podzols; surface water gleys; and man-made soils (Thompson et al. 1986). Pteridium was most common in the Manod association, which is extensive on steep ground (the ‘Ffridd’) where there is a depth of well-aerated soils.

Where the profile appears important, it is usually because of water and aeration. Pteridium grows on shallow soils, but in many dry areas (e.g. East Anglia) or in exposed sites in the humid west it grows better on deep soils because of the greater amount of available water (Watt 1976). Pteridium is more common on acid, nutrient-poor, well-drained, sandy to loamy soils, but this is often because it is excluded from other soil types because of intensive management.

Height and seasonal variation of the water table

Although Pteridium tends to be found on well-drained soils, Pteridium is neither confined to them nor is it absent from waterlogged soil. Pteridium occurs in flushes and gleyed soils, although frond density is often lower than on drier soils illustrating the detrimental effects of waterlogging. Pteridium can persist and become dominant on soils with a periodically high water table at the soil surface, or even where there is running water (Poel 1951). Indeed living rhizomes have been uncovered crossing Welsh mountainside flushes. These apparently contradictory data are reconciled when account is taken of the oxygen concentration of the soil and its incidence in relation to rhizome growth. Thus, in apparently waterlogged situations the entire rhizome morphology can consist of long meandering rhizomes, bearing single or few fronds, lying on the wet mineral surface but underneath a Sphagnum or mor layer. In Felsham Hall Wood, Suffolk, a bracken glade on a gleyed podsol with iron pan that floods through capillary action during winter has been described; Pteridium grows well, sometimes reaching over 4 m by growing in its own litter and humus, which has a maximum depth of c. 23 cm (Rackham 1980). Alternatively, Pteridium rhizomes may be found in wet soils where there are stones or other features which improve drainage (Poel 1951).

In some circumstances Pteridium invasion from adjoining dry soils can be impeded by waterlogging, with communities often showing sharp boundaries (Jeffreys 1917; Poel 1951; Nicholson & Robertson 1958; Willis et al. 1959a; Anderson 1961). The importance of waterlogging, or rather lack of aeration, has been confirmed both in the field and laboratory (Poel 1948, 1951, 1961).

Just how long the rhizome can withstand submersion in oxygen-deficient water is unknown. Poel (1951) examined rhizomes 4 months after flooding without detecting any change. The experimental and the field evidence indicate that a periodically high water table presents a physiological barrier to rhizome penetration. Pteridium can grow in a relatively high water table during the vegetative period, providing there is a continuous supply of aerated water, although the rhizome and root system remain shallow. When the water table falls, there is liability to drought. In field measurements the oxygen diffusion rates found under Pteridium were at the upper end for upland communities (Table 3).

Table 3.  Oxygen diffusion rates in soils under a range of upland plant communities: the diffusion rates (g × 10−8 O2 cm−2 min−1) were measured using a platinum electrode (depth 14.5 cm, surface area 27.02 mm2) (Poel 1960)
Plant communityOxygen diffusion rate (mean ± SE)
Molinia caerulea-Nardus stricta with Deschampsia flexuosa17.31 ± 0.81
Pteridietum (1.3 m tall)15.54 ± 0.63
Pteridietum in Agrostido–Festucetum15.46 ± 0.32
 Agrostis–Festuca–Nardus stricta14.18 ± 1.42
 Calluna–Pteridium14.01 ± 1.08
Pteridium island12.52 ± 0.77
 Calluna12.33 ± 0.39
Agrostido–Festucetum with Pteridium11.65 ± 1.03
Juncetum acutiflori (flush) 9.65 ± 0.34
 Juncus effusus–Deschampsia cespitosa 8.25 ± 0.26
 Caricetum nigrae 6.96 ± 0.47
Mixed Carex nigra, Juncus acutiflorus, Menyanthes trifoliata, Molinia caerulea, Sphagnum sp. 5.39 ± 0.27
Mixed Carex bigelowii, C. rostrata, Juncus acutiflorus, Menyanthes trifoliata, Molinia caerulea, Sphagnum sp. 3.59 ± 0.27

Soil pH

Pteridium can grow over a wide pH range. Frond height is not a function of pH although at the extremes fronds are few and small. Laboratory experiments with young sporophytes confirm tolerance over a wide range of acidity (Conway 1949; Schwabe 1953). In the field Pteridium is found on soils with a pH ranging from 2.8 to 8.6 (Willis et al. 1959a,b; Rackham 1980; Koedam et al. 1992), although most bracken is found in moderately acidic soils. In a survey of 200 stands the modal pH was 5.5 (Salisbury 1925). In another survey of 31 stands the mean pH of the A(Ea) horizon was 4.5 (A.S.W.). Rackham (1980) recorded perhaps the tallest Pteridium (> 4 m) in the UK in a glade at Felsham Wood, Suffolk, where the Pteridium-dominated soil profile had a mor humus (pH 3.0–3.7) overlying a mineral layer (pH 3.0–4.2).

Similarly, vigour is not related to the soil Ca content (Conway & Stephens 1957). Pteridium's classification as a calcifuge is not justified as it has been reported on calcareous soils and limestone pavement (A.S.W.; Elgee 1914; Webb 1962); in France Pteridium has been found in a soil with 60% CaCO3 and pH 8.33 (de Litardière 1933). In Breckland it grows on shallow soils with a CaCO3 content up to 17% (0–15 cm depth), with its rhizomes growing into a layer containing about 50% CaCO3 (Watt 1940).


The organic matter content in the A horizon below Pteridium also varies over a wide range (< 1–92%, A.S.W.). Part of this variability may be due to the absence of effective mixing agents in most Pteridium communities. At death, the frond contains a substantial amount of cations and there is evidence that these are leached out quickly. Presumably they are either transferred to the soil system, thus reducing acidity, or leached (Galtress 2001). Burning litter temporarily increases the pH of the soil but the released bases are soon lost (Donnelly 2003). However, in comparisons of recently invaded and dense Pteridium stands, there has been little to support the idea that Pteridium litter is an effective contributor to soil organic matter or an antidote to acidity, i.e. there was no detectable change in pH, exchangeable Ca and total exchangeable bases (Watt 1940).

Worms and other fauna

Pteridium is most often found on podzols where there is a sharp transition between the organic (F and H) layers and the mineral soil, which emphasizes the paucity or absence of mixing agents such as earthworms. The pill millipede Glomeris marginata has, however, been shown to be an important consumer of Pteridium in Denmark (Nielsen, cited by Elton 1966), and as earthworms are common on rendzinas and brown earths, which also support Pteridium, presumably they are involved in its decomposition.

Other fauna may also play a part in decomposition and nutrient turnover, but this has not been evaluated: moles, mice and shrews have been recorded from Pteridium patches at Lakenheath Warren, Breckland, Suffolk (A.S.W.).

Rate of decay and humus incorporation

There is much variation in the type and thickness of Pteridium litter, and its rate of decay. At the end of winter, or in early spring when fronds are emerging, litter thickness varies from negligible to c. 60 cm. During this period the relative proportions of L and F layer vary, but often the H layer is negligible or absent, and the F layer rests directly on the mineral soil. The decay rate depends on soil texture, moisture content and humidity.

Where Pteridium fronds are tall and sturdy they may break between 38 cm and 61 cm; the laminae form a canopy at that level, erect with oblique petioles of varying ages beneath, and an F layer accumulates on the surface. Generally Pteridium decay is much slower than the rate of addition, with up to 11 years required for complete decay (Frankland 1966a,b, 1976). However, in the west of the UK fronds can disappear within 12 months; on Mull and Skye in tall Pteridium (1.5 m), mull humus was found at the soil surface amid an undergrowth of Hyacinthoides non-scripta and Oxalis acetosella, indicating that earthworms are active.

Decomposition is affected by local variations in altitude, exposure, woodland vs. non-woodland situations and whether the litter is mixed with that from other plants. Elfyn Hughes (pers. comm. 1972) observed variable breakdown rates of dead Pteridium fronds within a very limited area alongside the Synchnant Pass (near Conwy in North Wales). There, the fronds of the previous years were accumulating, whereas within 100 m Pteridium litter was absent.

Humification can be a slow process. Decaying petioles and rhizomes must affect the soil organic content but the process is slow and patches of soil have been observed as being unaffected after 35 years’ invasion (A.S.W.). No rigorous study has been made of rhizome decay. Observations suggest that the tissue first becomes brown in the centre near the vascular bundles but the outer black ‘rind’ of the rhizome can persist for at least 33 years, and probably longer (A.S.W.).

The fauna involved in decomposition have not been determined, but the following groups of organisms were recorded from Pteridium litter at Lakenheath Warren, Suffolk, in 1970: Annelida (Enchytraeid worms), Nematoda, Acari (49 spp.), Diptera (2 spp.), Protura (all I.N. Healey, pers. comm.), Coleoptera (14 spp.; W. Block, pers. comm.), Collembola (36 spp.; A. Macfadyen & I.N. Healey, pers. comm.), Aranaeae (22 spp.; E. Duffey, pers. comm.), Opiliones (4 spp., Elton 1966), Myriapoda (> 2 spp.), and Isopoda and Pseudoscorpionida (1 sp. each). Many of the Acari and Collembola can feed on fungal mycelium in the litter and some eat Pteridium spores, but the part played by other members of the soil fauna is unknown.

III. Communities

The distribution and cover of Pteridium can range from rare to complete dominance. It is rare in a wide range of communities where it is limited by moisture, aeration, frost and exposure to wind. Pteridium patchiness is obvious at the landscape scale, and this is related primarily to land use, variability in terrain, type of grazing animals and land-use history. There is also within-stand patchiness, which can in some places be related to the growth phases of Pteridium (Sections V, VI, Watt 1976).

Pteridium occurs over a considerable area of land in Britain, with an estimate of 17 073 km2 (7.3%) containing some Pteridium (Pakeman et al. 1995). This overall total was split between four categories: dense stands in the open (4762 km2, 2%); sparse stands in the open (4278 km2, 1.8%); dense stands in woodland (2261 km2, 1%) and sparse stands in woodland (5771 km2, 2.5%). There also appears to be a considerable flux with both gains and losses in Pteridium-dominated land over quite short time intervals (Pakeman et al. 1995). The changes between 1984 and 1990 showed approximately equal fluxes between dense Pteridium communities and both managed and moorland grass land. However, there was a considerable change from dense Pteridium in 1984 to dwarf shrub heath in 1990 (Table 4; Pakeman et al. 1995). These fluxes were tentatively ascribed to the effects of changed land management; where Pteridium was reduced, it was probably because of increased control measures or tree planting; increases could have resulted from fresh invasion or recovery after a previous successful control treatment. Furthermore, Pteridium was present in 122 000 km of linear features, which included hedgerows and roadside verges. These linear features could provide a focus for invasion into adjoining land should climate or land use change (Pakeman et al. 1995).

Table 4.  The main fluxes of land cover into and out of Pteridium dominance between 1984 and 1990 in the UK (from Pakeman et al. 1995)
1984 land cover of areas surveyed as Pteridium-dominated in 1990 (km2)1990 land cover of areas surveyed as Pteridium-dominated in 1984 (km2)
Managed grasslandDense PteridiumMoorland grasslandDwarf-shrub heathPteridium cover in 1984
Managed grassland  400   
Dense Pteridium40032001005004200
Moorland grassland  100   
Dwarf-shrub heath  100   
Pteridium cover in 1990 3800   

Nevertheless, Pteridium occurs in a wide range of communities; for brevity we consider Pteridium only within UK communities, although a link from these to Corine biotope classes provides a wider European perspective (Table 5).

Table 5.  The National Vegetation Classification (NVC) communities and subcommunities in which Pteridium aquilinum has been recorded; I–V are constancy classes (20% intervals) and approximate Domin ranges are presented in parentheses (after Rodwell 1991a,b, 1992, 2000). CORINE biotype types are also presented; these are cross-referenced to the main NVC communities (Hill 1996)
NVC ClassDescriptionMain communitySub-communityCORINE biotope
Woodlands and scrub (W)
W4Betula pubescens–Molinia caerulea woodlandI (1–5)II (1–5)I (2–5)    C44.A1
W7Alnus glutinosa–Fraxinus excelsior–Lysimachia nemorum woodlandI (1–4)  I (1–4)   C44.31
W8Fraxinus excelsior–Acer campestre–Mercurialis perennis woodlandI (1–5)I (1–5)I (2–5)I (3)I (3–5)I (1–4) C41.32
W9Fraxinus excelsior–Sorbus aucuparia–Mercurialis perennis woodlandI (1–9)I (1–6)I (1–9)    C41.31
W10Quercus robur–Pteridium aquilinum–Rubus fruticosus woodlandIV (1–10)IV (1–9)III (2–7)IV (1–10)V (1–10)III (1–8) C41.21
W11Quercus petraea–Betula pubescens–Oxalis acetosella woodlandIV (1–9)III (1–9)IV (1–5)IV (1–9)IV (1–8)  C41.532
W14Fagus sylvatica–Rubus fruticosus woodlandIII (1–6)      C41.121
W15Fagus sylvatica–Deschampsia flexuosa woodlandIII (1–9)I (1–2)II (2–7)V (3–8)V (3–9)  C41.121
W16Quercus sp.–Betula sp.–Deschampsia flexuosa woodlandIV (1–10)IV (1–10)III (2–9)    C41.52
W17Quercus petraea–Betula pubescens–Dicranum majus woodlandIII (1–8)III (1–7)III (1–7)IV (1–8)III (1–8)  C41.532
W18Pinus sylvestris–Hylocomium splendens woodlandI (1–4)I (1–3) I (1–2)II (1–3)II (1–4) C42.51
W19Juniperus communis ssp. communisOxalis acetosella woodlandII (1–5)II (1–5)II (1–5)    C31.88
W21Crataegus monogyna–Hedera helix scrubI (1–8)I (2–8)I (1–2)I (1–3)   C31.812
W22Prunus spinosa–Rubus fruticosus scrubIII (1–6)III (1–6)III (1–4)III (1–4)   C31.811
W23Ulex europaeus–Rubus fruticosus scrubIII (1–7)III (2–5)III (1–7)III (1–4)   C31.85
W25Pteridium aquilinum–Rubus fruticosus scrubV (1–10)V (1–10)V (6–10)    C31.831
Mires (M) and heaths (H)
M15Scirpus cespitosus–Erica tetralix wet heathI (1–4) I (1–4)I (1–3)I (1–4)  C31.11
M27Filipendula ulmaria–Angelica sylvestris mireI (4–5)  I (4–5)   C37.1
H1Calluna vulgaris–Festuca ovina heathI (2–4)I (2) I (4)I (4)I (3) C31.225
H2Calluna vulgaris–Ulex minor heathIII (1–7)II (1–2)IV (2–7)II (2–3)   C31.238
H3Ulex minor–Agrostis curtisii heathI (1–5)I (2–3)I (1–5)I (1–4)   C31.238
H4Ulex gallii–Agrostis curtisii heathI (1–7)I (2–3)II (1–7)I (3–7)I (3–5)  C31.235
H6Erica vagans–Ulex europaeus heathI (1–2)II (1–2)I (1)    C31.234
H8Calluna vulgaris–Ulex gallii heathII (1–7)II (2–5)I (3)II (1–7) II (3) C31.235
H9Calluna vulgaris–Deschampsia flexuosa heathI (1–7)II (2–5)I (1–4)I (4–7)I (3–5)  C31.2254
H10Calluna vulgaris–Erica cinerea heathI (1–4)I (1–4) I (1–3)I (1–3)  C31.2256
H12Calluna vulgaris–Vaccinium myrtillus heathI (1–6)II (1–6)I (1)I (1)   C31.2256
H21Calluna vulgaris–Vaccinium myrtillus–Sphagnum capillifolium heathII (1–3)III (1–3)     C31.2128
Grasslands and montane communities (MG–mesotrophic grasslands, CG–calcicolous grassland, U–calcifugous grasslands and montane communities)
MG1Arrhenatherum elatius grassland; Arrhenatherum elatioris (Br-Bl 1919)I (2–6)I (3–6)I (2–5)I (4–5)I (3)  C38.22
MG5Cynosurus cristatus–Centaurea nigra grassland; Centaureo-Cynosuretum cristati Braun-Blanquet & Tüxen 1952)P. aquilinum noted only in the description of sub-community MG5c–doesn't appear in data tableC38.112
CG10Festuca ovina–Agrostis capillaris–Thymus praecox grasslandI (1–4)I (1–4)I (1–3)I (1–3)  I (1)C34.321
U1Festuca ovina–Agrostis capillaris–Rumex acetosella grasslandI (1–4)I (2–3)I (3–4)I (4)I (3)I (2–3) C35.22
U2Deschampsia flexuosa grasslandII (1–8)III (1–8)     C35.13
U3Agrostis curtisii grasslandII (1–5)      C35.1
U4Festuca ovina–Agrostis capillaris–Galium saxatile grasslandI (1–8)I (1–6)I (1–4)I (8)I (1–4)II (1–4) C35.12
U20Pteridium aquilinum–Galium saxatile communityV (4–10)V (4–10)V (4–10)V (4–10)   C31.86
Maritime communities (SD–sand dune communities, MC–maritime cliff communities) and vegetation of open habitats (OV)
SD9Ammophilia arenaria–Arrhenatherum elatius dune grasslandI (1–6)I (3–6)I (1–3)    C16.2211
MC12Festuca rubra–Hyacinthoides non-scripta maritime bluebell communityII (1–5)II (1–5)I (2–5)    C18.2
OV2Briza minor–Silene gallica communityIII (1–3)      C82.3
OV25Urtica dioica–Cirsium arvense communityI (5–6)II (5–6)     C87.2
OV27Epilobium angustifolium communityI (1–8)II (3–5)II (1–8)II (5–8)I (5)  C87.2

(a) pteridium communities

Pteridium occurs as a component species in woodland, heath, grassland, maritime habitats and in vegetation of open habitats (Rodwell 1991a,b, 1992, 2000). It occurs in 42 communities and 153 subcommunities: 62 woodland, 47 heaths and mires, 30 grassland and 14 maritime or open habitats (Table 5). In most communities it is an occasional species and forms a small proportion of the cover. In a number of woodland and scrub communities it is an important species in the field layer, and can be locally prominent on some grasslands and heaths. Pteridium occurs in grassland communities across the pH spectrum from calcifugous grasslands and montane communities, through mesotrophic to calcareous grasslands. Notably, it is absent from mire communities. However, it is constant (constancy = V) in only two communities: the bracken-heath bedstraw Pteridium aquilinum–Galium saxatile and the bracken–bramble Pteridium aquilinum–Rubus fruticosus agg. communities.

The Pteridium–Galium community (U20) is usually found in areas of heathland or grassland, and is fairly uniform across the whole country. It is relatively species-poor (Rodwell 1992), and both calcifuge and shade-tolerant species may be present. In the Sheffield region, the mean number of species found beneath a canopy of bracken ranged from 4 to 6 m−2 (Grime et al. 1988). With increasing dominance of the fern, the flora is progressively impoverished. Where bracken is disturbed, or where it is degenerating, more species occur. Where this community is found in woodland, it has often appeared after previous disturbance, especially in acidic, nutrient-poor woods.

The Pteridium–Rubus scrub community (W25) is usually found on nutrient-rich woodland sites. This community forms only a small proportion of bracken-dominated land in Britain, but is more floristically diverse than the Pteridium–Galium (U20) community. It has the general appearance of a Quercus robur–Pteridium aquilinum–Rubus fruticosus agg. woodland without an intact tree cover, but usually with many vernal species still present. The occurrence of these vernal plants, such as Anemone nemorosa, which has poor dispersal (Grime et al. 1988), suggests that woodland has been replaced. Where this community lacks a strong vernal element it is associated with sites of relatively young woodland or severely disturbed sites.

In addition to these communities, Pteridium is abundant or common (constancy = IV or III) in a range of communities, e.g.: (i) woodland –Quercus woodland (W10, W11, W16, W17) and Fagus woodland (W14, W15), (ii) two scrub communities –Prunus spinosa–Rubus fruticosus agg. (W22) and Ulex europaeus–Rubus fruticosus agg. (W23), (iii) two heath communities with Pteridium common in subcommunities – the Calluna vulgaris–Ulex minor heath (H2) and the Calluna vulgaris–Vaccinium myrtillus–Sphagnum capillifollium heath (H21), (iv) grasslands –Pteridium is common in the Deschampsia flexuosa grassland (U2), and (vi) it is common in the Briza minor–Silene gallica open vegetation class (OV2).

Where Pteridium is present in woodland, it sometimes achieves dominance beneath the tree canopy (Rodwell 1991a), and when this occurs it hinders the re-establishment of woody species. Often it varies with each coppice or forest cycle, expanding after felling, and becoming gradually suppressed as the canopy develops. As Pteridium expands and contracts there is a reduction and a subsequent increase in species diversity in the field layer (Rodwell 1991a).

Pteridium communities can harbour a number of notable plant species, e.g. Melittis melissophylum and Orchis mascula early in the season in south-west England (A.S.W.), Galium parisiense (Grassland A, Lakenheath Warren, Suffolk, and Watt 1940), Hymenophyllum wilsonii in a Betula–Vaccinium myrtillus community on Skye, Highland Region (Birks 1973), Calamagrostis canescens and Phragmites communis on Wangford Fen, Suffolk, Carex paniculata with Pteridium in tussocks in Tuddenham Fen, Suffolk, Eriophorum angustifolium and Sphagnum fimbriatum on Flitwick moor, Bedfordshire, with Iris pseudacorus and Cladium mariscus on the Norfolk Broads (G. Crompton, pers. comm.) and intermingled with Phragmites australis at Slapton Ley, Devon (M.C.F. Proctor, pers. comm.). Sometimes species usually considered as woodland plants survive in bracken after woodland removal, e.g. Ceratocapnos claviculata, Colchicum autumnale and Polygonatum multiflorum (R.J. Hornby, pers. comm.), all described as local by Clapham et al. (1987). Others such as Cornus suecica (very local in England) and Trientalis europaea (local in Britain), are found in Pteridium on the North York Moors (S. Rees, pers. comm.). Trientalis europaea is also found in Aberdeenshire (R.J. Pakeman, pers. comm.). In the New Forest, Pteridium gives essential protection and shade to Gladiolus illyricus (R.J. Hornby, pers. comm.).

(b) community interactions

Pteridium is often a secondary invader of abandoned farmland. Indeed, the presence of Pteridium on islands in Connemara has been used as supporting evidence of past grazing activity (Webb & Glanville 1962). Elsewhere, in west Scotland Pteridium invades and suppresses Agrostis and Festuca, and Calluna-dominated moorland, usually as a result of excessive burning and/or heavy grazing. On grassland, however, it usually does not become completely dominant, as some grasses survive under the canopy. Where high-quality, prescribed Calluna burning is carried out, there is no conspicuous spread of Pteridium (Ninnes 1995). However, Pteridium will invade if the Calluna becomes degraded; this can occur if Calluna colonization after fire is delayed by heavy grazing, or if there is excessive accidental or poor-quality burning. Overgrazing by sheep may also lead to a permanent replacement of Calluna by Pteridium (Miles 1979).

IV. Response to biotic factors

(a) ecological status in british vegetation

Tansley regarded Pteridium as ‘a plant of woodland origin, of moderate shade, and it often marks the sites of woods which have been destroyed: but when it is freed from control in the open it often becomes a pestilent weed’ (Tansley, Br. Isl.). Rackham (1986) questioned the validity of this generalization citing the presence of Pteridium in Breckland, an area that had not carried woodland for centuries. However, palynological evidence indicates that woodland was present in Breckland in the past (Godw. Hist.). In woodland conditions, Pteridium performance would be expected to be limited by low spore production as a result of shade, and both dispersal range and conditions for the establishment of both gametophyte and young sporophyte would be unfavourable. In the climax woodland of pre-Neolithic vegetation, Pteridium could be expected in open-canopied woods, wind-throws, burns and canopy gaps during the woodland regeneration cycle. Together with the rather specialized conditions needed for the successful establishment and survival of the gametophyte and young sporophyte, this would account for the rarity of the spores in the pollen diagrams of the pre-Neolithic. As the area of woodland has progressively reduced and human interference increased, spore production in the open would have increased; there would be wider spore dispersal and an extension of the establishment niches for the gametophyte (Godwin 1956).

In the open, Pteridium is excluded from land under continuous cultivation; the few hectads from which it is absent in Great Britain (Fig. 1) are mostly under cultivation. Where cultivation is abandoned Pteridium re-invades. Today, most Pteridium is found therefore in woodland, scrub, waste heath and hill pasture. Among the several explanations offered for the apparent spread of Pteridium within the last 200 years or so is a change in the way in which the land is managed. Previously, Pteridium was used for a variety of purposes within the rural economy (Section X); this management has declined and Pteridium has spread as a result.

Recent success of Pteridium has been attributed to a range of factors, including its competitive powers, life form, vegetative reproduction and virtual immunity to grazing. It is arguable that this success owes as much to the reduction of the competitive power of its rivals through the activities of herbivores and especially grazing by domestic stock.

Pteridium occupies a special place in British vegetation: like all ferns it requires moisture for sexual reproduction, indeed it appears to have a greater requirement for moisture at this stage than other fern species. However, it is different from most other British ferns in that the sporophyte flourishes outside woodland, so much so, that it has become a pest species.

(b) the relation of dominant pteridium to itself and other species

From studies of morphology across Pteridium fronts, Watt (1945, 1947) demonstrated cyclic changes between four distinct phases and grass heath at Lakenheath Warren, Suffolk (Fig. 3a). It is important to note that the grass heath phase of this cycle is not free of Pteridium (Watt 1947). In an aerial photographic study of the distribution of plant communities at Lakenheath, evidence was found to support this cycle (Marrs & Hicks 1986): there was degeneration to grass heath with Pteridium densities < 1 frond m−2 over an area of at least 90 ha. This cyclic regeneration pattern has not been reported for many other sites, so it may not be a general phenomenon. Watt (1969, 1970, 1971, 1976) suggested that the increasing litter depth was one of the main factors contributing to the degeneration. At this site there had been little grazing for at least 20 years and, coupled with poor incorporation of dead fronds into the mineral soil, a sharp transition from decomposing litter to the mineral soil developed. When deep litter (84 cm) developed, 72% of the total rhizome was present in the litter layer, and there was a general rise in the level of the younger rhizomes in the upper L layer (65%) as compared with the lower F (only 19%). The roots also showed a marked reduction in the litter layer immediately above the mineral soil, from 12.4 to 4.2 m per 70 cm3. In other words, the Pteridium in this one spot had, through time, moved from the mineral soil to the litter of its own provision, and from a medium of relatively high to that of low nutrient status (Fig. 3b).

Figure 3.

Diagrammatic representation of (a) the Pteridium regeneration cycle (Marrs & Hicks 1986), and (b) associated changes in the litter and in the behaviour of the short- and long-shoots through this cycle in a podzol at Lakenheath Warren (Suffolk); black = live; white = dead (Watt 1945, 1947).

As the rhizomes develop in Pteridium's own accumulating litter, the rhizomes and buds grow in a less favourable medium, where they may be exposed to frost and drought (Section V). Fronds become smaller, litter production falls and eventually the rate of gain becomes less than the rate of loss (A.S.W.). In time the dense canopy opens up and is no longer capable of excluding young woody plants by shade. The duration required for this process is unknown, but is likely to be decades rather than years. Grazing of the colonizing shrubs may prevent this process.

Competitive ability of Pteridium

Pteridium can persist in a wide range of environmental conditions (Section II) and has many of the classic attributes of a competitive species, i.e. (i) the ability to store large resources in rhizomes that are protected from all but the most severe frosts, (ii) rapid frond production in spring, (iii) development of a dense canopy and litter layer, and (iv) it is generally unaffected by grazing, insect herbivory and pathogens.

The interactions between Pteridium and other species depend on their inherent characteristics, and the differentiating effect of the environment on their relative competitive powers and management, but also on time and particularly the effect of Pteridium on itself. In slightly damper habitats, Pteridium and Molinia caerulea are co-dominant to the exclusion of almost everything else. Molinia caerulea is apparently one of the few plants that can successfully compete with Pteridium, but this may reflect soil waterlogging/low aeration rather than the sole effect of M. caerulea (Summerhayes et al. 1924).


Pteridium can interfere with the germination, establishment and growth of other species through shading and smothering (Watt 1919; Humphrey & Swaine 1997; den Ouden 2000) or the build up of a deep litter layer (Jeffreys 1917; Sydes & Grime 1981; Lowday & Marrs 1992b; den Ouden & Vogels 1997). Pteridium can also produce large amounts of potentially toxic compounds (Section VI) in soil (Whitehead 1964), with many of these chemicals being involved in anti-herbivore and anti-pathogen defence. This has led to the suggestion that some of these compounds could induce allelopathic interference on other species (Bohm & Tryon 1967; Glass & Bohm 1969; Gliessman & Muller 1972, 1978; Nava et al. 1987; Dolling et al. 1994). A large number of studies has been done, but the results have been contradictory, varying with climatic zone of origin and the nature of Pteridium material tested (Table 6). Gliessman (1976) has attempted to resolve the problem of the high variability in the phytotoxic potential of Pteridium toxins by hypothesizing that toxin release is geared to the period of germination of co-occurring plants. However, den Ouden's (2000) review (Table 6) suggested that the variability in response obtained was more likely to be as a result of the experimental methods employed. Thus, whilst allelopathy cannot be ignored as a potential mechanism for Pteridium interfering with other species, direct proof is lacking.

Table 6.  Overview of results from studies investigating phytotoxic potential of Pteridium on other species (from den Ouden 2000). Results from a series of experiments are reported based on climatic regime, the processes measured (germination, radicle elongation and seedling growth), the test media in which seeds or seedlings were exposed to bracken substrate (extracts or leachates of plant material and Pteridium substrate mixed into soil), and the different components of Pteridium that were used in the experiments (green and senescent fronds, and dead fronds or litter). Experimental results are presented as percentage of individual tests that showed significant (P < 0.05) inhibition of plant performance with respect to the controls. The total number of individual tests is given in parentheses
Test mediumPteridium component testedPlant development processTotal % (n)Reference
Germination % (n)Radicle elongation % (n)Seedling growth % (n)
  1. References: 1, Brown (1967); 2, del Moral & Cates (1971); 3, den Ouden (1995); 4, Dolling et al. (1994); 5, Everson & Breen (1983); 6, Gliessman & Muller (1972); 7, Gliessman & Muller (1978); 8, Gliessman (1976); 9, Horsley (1977); 10, Martin & Sparke (1982); 11, Nava et al. (1987); 12, Stewart (1975); 13, Taylor & Thomson (1990); 14, Tolhurst & Turvey (1992).

Temperate climates
 ExtractGreen fronds60 (5) 60 (5) 33 (3) 54 (13)1, 2, 3, 4, 9
Dead/litter 0 (8) 56 (9) 29 (17)2, 3, 4, 8, 12
 LeachateGreen fronds 0 (3)  0 (3)  0 (6)10
Dead/litter 0 (3)  0 (3)  0 (6)10
 Mixed in soilDead/litter 25 (5) 33 (5)12
Total Temperate 16 (19) 40 (20) 25 (8) 29 (47) 
Mediterranean/subtropical climates
 ExtractGreen fronds 0 (13) 14 (14)  7 (27)6, 8, 11, 14
Senescent  0 (2)  0 (2)5, 6
Dead/litter 75 (4) 75 (4)5, 6
 LeachateGreen fronds 63 (8) 63 (8)13
Senescent100 (2)100 (2)13
Dead/litter 71 (7) 71 (7)6, 7
 Mixed in soilGreen fronds14 (7)  0 (7)100 (4) 28 (18)11
Dead/litter100 (1)100 (1)6
Total mediterranean/subtropics  5 (20) 39 (44)100 (5) 33 (69) 

Inhibition of phytopathogenic fungi and bacteria has also been described (Nava et al. 1987); an aqueous extract of the fronds inhibited both fungi and a gram +ve bacterium; methanolic and ethanolic extracts were less effective against the fungi.

Relationship with Calluna

The distributions of both Pteridium and Calluna occur over a wide range of edaphic tolerances with an appreciable overlap in habitat. The classic work between Pteridium and Calluna at Lakenheath Warren (Suffolk) showed that Pteridium and Calluna occupy different proportions of land in space and time (Watt 1947, 1955). In the podzol area, rabbits (Oryctolagus cuniculus L.) were abundant until 1941 when they were significantly reduced on this site (Watt 1960). In 1933 an experiment was established of 12 plots (1.48 m2) with scattered fronds and closely grazed grass heath, but with no Calluna. The area was burnt in the autumn of 1947 and again in June 1959. By 1971 Calluna had invaded and occupied 37% of the plot area with Pteridium fronds growing through the Calluna bushes quite freely.

Calluna also follows a regeneration cycle and its competitive ability diminishes in the mature and degenerate phases (Watt 1947, 1955). This can be demonstrated from data on the distribution of fronds in circular Calluna bushes of different diameters. Frond density decreased from the periphery of the bush toward the centre (Table 7a), even though none of the bushes had reached the degenerate phase in the centre. Rhizome characteristics led to a similar conclusion (Table 7b), with much lower amounts of rhizomes and fewer buds under Calluna compared to Pteridium communities outside. When rhizomes from a patch dominated by Pteridium over grass heath were compared with those from a 1-m diameter Calluna bush, there were similar lengths of long-shoots, but a lower length of short-shoots and bud numbers (Table 8), illustrating the competitive impact of Calluna in restricting Pteridium.

Table 7.  Effect of Calluna on (a) distribution of Pteridium fronds in different sized Calluna bushes and (b) rhizome properties in different communities from Pteridium-dominated into Calluna bushes of increasing diameter (6 August 1954, A.S.W.). All data were originally expressed in Imperial measures, but have been recalculated
(a) Radius class (cm)Diameter of Calluna bushes (cm)
90 cm n = 4120 cm n = 3150 cm n = 3215 cm n = 1
0–15 6.900 0
15–30 0
30–4613.02.81.4 8
46–61 3.94.9 0
61–76  5.4 0
76–91    7.5
91–107   10.5
(b) CommunityLength of rhizome (m m−2)Number of terminal budsNumber of new fronds
Long-shootsIntermediate shootsShort-shoots
In dead Calluna centre194 60 28 5.5 0.5
In Calluna30127421522 5
In Calluna165232 6914 6
In periphery Calluna36534417121 4
Outside Calluna in Pteridium patch3633994543110
Outside Calluna in Pteridium patch29532932522 6
Table 8.  A comparison of Pteridium rhizome length and bud morphology in a Pteridium-dominated community and in an adjacent Calluna patch (24 March 1954, A.S.W.). All data were originally expressed in Imperial measures (inches and number of buds per 4 square feet), but have been recalculated
MeasurePteridiumPteridium in Calluna patch
Rhizome length (m m−2)
 Long-shoots  8.9  9.2
 Intermediate  7.8  8.6
 Unclassified  0.7  0.9
 Short-shoots  8.0  5.7
Numbers (number m−2)
 Terminal buds169124
 Lateral buds108 46
 Main rhizomes 30 24
 New fronds 51 19

Watt (1955) then showed that Calluna and Pteridium growing along a front could maintain their respective positions as a result of interdigitation, where Pteridium invaded into pioneer and degenerate Calluna and was repulsed by building and mature Calluna. Watt's study plots were revisited 35 years later and it was shown that in the absence of management to maintain Calluna, this system had broken down (Marrs & Hicks 1986). However, a managed variant using burning to maintain vigorous Calluna, which keeps Pteridium in check, has been proposed (Section III; Ninnes 1995).

Relationship with Deschampsia flexuosa

There is anecdotal evidence that Deschampsia flexuosa can, like Calluna, also compete with Pteridium in some circumstances, especially when D. flexuosa increases after grazing is stopped, as described at Lakenheath Warren, Suffolk (Watt 1976).

Where Pteridium is controlled D. flexuosa is a common component of the colonizing vegetation, often forming mixed grass-heath communities with Rumex acetosella and Campylopus introflexus and sometimes becoming the dominant species (Pakeman et al. 2005). However, these authors suggested that this vegetation might be a transient type, as it could either revert back to Pteridium domination or change further into Calluna-dominated vegetation. In long-term studies of Pteridium control in the Peak District (Derbyshire), D. flexuosa dominance was linked to the removal of sheep grazing; where this occurs tree species (Quercus spp., Betula spp. and Sorbus aucuparia) also invaded (Le Duc et al. in press).

Relationship with trees

As a woodland understorey species, Pteridium is suppressed presumably through a combination of reduced light and moisture. Where clearings are created, Pteridium increases in both cover and height (Harmer et al. 2005). However, in a national woodland survey of Britain, Pteridium was one of the four dominant field-layer species (the others being Rubus fruticosus agg., Hedera helix, and Mercurialis perennis), which together produce a series of overlapping curves that cover the distributions of all other woodland species (Corney 2006). Pearman (2004) has described three of these four species (excepting M. perennis) plus Urtica dioica as ‘native thugs’; he hypothesized that these species could cause more problems for the conservation of native woodland flora than invading alien species.

In other open situations it is reasonable to expect that Pteridium would occur as a late-successional species within a grass/heath/moor→Pteridium→woodland sequence. However, there is evidence that Pteridium can suppress tree colonization (Tolhurst & Turvey 1992; Dolling 1996; Humphrey & Swaine 1997). If this is so, natural colonization by late-successional tree species can occur only under two conditions: either at the start of the succession when Pteridium is also invading and at relatively low vigour – the Tolerance model of Connell & Slatyer (1977); or if the Pteridium is established, then some factor or factors cause the Pteridium to diminish, even if only temporarily – essentially the Inhibition model of Connell & Slatyer followed by a catastrophe (Marrs 1988; Marrs et al. 2000).

There is no doubt that under some circumstances Pteridium inhibits invasion of tree species, either by direct shading or by the physical presence of litter (Miles & Kinnaird 1979a,b; Marrs 1988; Humphrey & Swaine 1997; Marrs et al. 2000). However, Marrs (1987a) found low densities of Betula spp. seedlings (0.005 m−2) invading plots where a dense Pteridium canopy had remained untreated. In a second experiment (Marrs 1987b), Betula seedlings were excluded at frond densities of > 20 m−2 and restricted to 0.25 seedlings m−2 when frond densities were between 11 and 20 m−2. Pinus sylvestris seedlings also persisted where the frond density was 28 m−2, but were restricted to single colonists in plots with frond densities > 20 m−2. Marrs & Hicks (1986) in a study at Lakenheath Warren showed that 5.5 ± 2.1 (SE) Pinus sylvestris seedlings invaded sparse Pteridium (39 ± 2 fronds m−2; mean height 40 ± 1 cm), and about half that number (2.5 ± 1.1 m−2) in dense Pteridium (33 ± 1.3 fronds m−2; height 89 ± 5 cm). Pinus sylvestris saplings (> 0.5 m tall) also showed differences in size-class distributions (small = 0.5–3 m, and tall > 3 m). The sparse Pteridium community had greater densities of small plants (25 vs. 4300 m−2) but lower densities of the tall plants (7 vs. 17 300 m−2) than the dense Pteridium. Marrs & Hicks (1986) suggested that these results depended on previous fluctuations in the Pteridium canopy density. The evidence therefore on balance suggests that tree seedlings are inhibited, to some extent, by dense Pteridium, but perhaps not completely.

If there is a reduction in Pteridium cover, especially where the litter layer has been disturbed, e.g. by animal tracks, there can be rapid tree invasion. In a study at Holme Fen, near Peterborough, 76 Betula spp. and 6 Salix spp. saplings invaded a 2-m wide pathway cut into tall, dense Pteridium (30 fronds m−2; 2.5–3 m tall, with a 0.5 m litter depth), but no saplings colonized the uncut, dense Pteridium areas (Marrs & Pakeman 1995). Where Pteridium has been reduced experimentally, mortality of tree seedlings is reduced and their performance increases (Tolhurst & Turvey 1992; Dolling 1996; Humphrey & Swaine 1997; Marrs et al. 2000).

V. Response to environment

(a) gregariousness

Pteridium occurs across a spectrum from a single isolated frond to large patches covering substantial areas of land. Solitary complete plants may be newly, or recently established young sporophytes, or they may be relicts of a disintegrated community. However, a single frond may be attached to a relatively complex rhizome system. Where a single frond was excavated from abundant Juncus subnodulosusMolinia caerulea community at Cothill, Berkshire (A.S.W.), the underground part consisted of a main rhizome 45 cm long, which carried five short-shoots (three live, two dead); the longest was 20 cm, with 16 nodes and no frond, the next longest was 15 cm with 12 nodes and it carried the only frond (65 cm tall).

In larger areas there is often a considerable patchiness, which can be classified into two types, where the patchiness is: (i) related to habitat variability in space (terrain, soil depth, aeration and exposure) or time (severe/frequent winter/spring frosts, competition with other plants); and (ii) induced by the effect of Pteridium on itself. The former is evident in varied terrain of hilly country where factors such as exposure to wind, soil moisture and soil depth play an important part. The latter can result from the cyclic regeneration process and/or rhizome development in the litter (Section IV). Patchiness also occurs at different scales, Anderson (1961) described two scales; a small-scale patchiness occurred around the position of a previous frond, which could be associated with a ventilating effect through decayed and decaying petiole bases; and a larger-scale patchiness at a larger scale perhaps resulting from the slow growth of competing clones.

Both isolated fronds and plants in large patches may be limited by exposure, frost and soil moisture as well as competition from other plants (Sections II, IV).

(b) performance in various habitats

There is much variation in performance in different habitats in frond density, height and biomass (Table 9). The relationship between these three performance measures is complex, as sites which score highly on one variable may be ranked lower on another (Table 10). The maximum height recorded is 4.4 m but generally mean frond height is < 1.5 m in open conditions. Density ranges from 1 to 80 fronds m−2. Estimates of frond dry matter production show a large range from 3 g m−2 in very sparse stands to 1410 g m−2 (14.1 t ha−1). There are, however, very large variations in space on individual sites (range c. ± 30% around a site mean; Pearsall & Gorham 1956; Watt 1964) and through time. For example, Lawson et al. (1986) measured peak frond biomass in three consecutive years at Lindale in Cumbria at 8.5, 11.2 and 9.4 t ha−1, a 12–15% range around the mean, and in a range of sites in both lowland and upland Britain large fluctuations between years have been noted (Table 9) (Marrs et al. 1998a; Le Duc et al. 2000). Pearsall & Gorham (1956) estimated the minimum frond biomass required to provide complete canopy cover to be 0.8 t ha−1. These data have been supported in montane Venezuela (‘Pteridium aquilinum’ var. ‘caudatum’ and var. ‘arachnoideum’); the maximum frond density was 160 m−2; maximum height was 2.2 m and maximum frond biomass was 1300 g m−2 (Alonso-Amelot & Baechler 1996).

Table 9.  Bracken frond production measured by a range of authors in the UK; mean values or ranges are presented. Watt (1964) data recalculated from Imperial measurements
SiteDensity (fronds m−2)Length (cm)Dry mass (g m−2)Reference
  1. n.d. = not determined.

Beaulieu Heathn.d.n.d. 892Pearsall & Gorham (1956)
Stony Crossn.d.n.d.1408Pearsall & Gorham (1956)
North Bentleyn.d.n.d.1072Pearsall & Gorham (1956)
South Bentleyn.d.n.d. 848Pearsall & Gorham (1956)
Bowland Forestn.d.n.d.1104Pearsall & Gorham (1956)
Lakenheath Grassland A 1.3–1.7   23–40 2.8–9Watt (1964)
Lakenheath Grassland B 2.9–3.0   35–36 15–16Watt (1964)
Lakenheath Grassland C21.5–24.2   63–65170–197Watt (1964)
Lakenheath Grassland D46.5–54.0  124–126874–973Watt (1964)
Lakenheath Grassland E  25–37.1   47–48199–136Watt (1964)
Weeting  18–28  105–180320–720Lowday & Marrs (1992a)
Cavenham  18–30   70–130200–570Marrs et al. (1998a)
Standford PTA, Norfolk  28–42n.d.420–760Pakeman & Marrs (1994)
Mull  28–36126.1–138474–592Paterson et al. (1997a)
Sourhope  27–3687340–425Paterson et al. (1997a)
Lake District  29–36   71–72283–409Paterson et al. (1997a)
Clywd  23–34   72–150534–558Paterson et al. (1997a)
Breckland  27–35   74–79262–480Paterson et al. (1997a)
Devon  15–16  111–115282–362Paterson et al. (1997a)
Lindalen.d.n.d.850–1120Lawson et al. (1986)
Sourhope 158.8 56.5391.3Le Duc et al. (2000)
36.8–80.1 46.9–67.0186–698 
Sourhope 226.5 69.9297.5Le Duc et al. (2000)
22.0–35.6 55.8–97.5205–360 
Peak42.6 91.1521.9Le Duc et al. (2000)
28.8–59.4 58.7–121.7165–888 
Carneddau44.0 74.1537.6Le Duc et al. (2000)
32.2–57.1 61.8–94.6402–972 
Cannock 136.8105.9676.4Le Duc et al. (2000)
23.4–53.1 84.6–129.3383–223 
Cannock 229.7 73.7294.2Le Duc et al. (2000)
14.9–51.3 63.1–98.5104–497 
Table 10.  Ranking of sites in a study by Le Duc et al. (2000) based on different measures of frond performance in the UK
SiteRanking for:
DensityLengthDry mass
Sourhope 1164
Sourhope 2655
Cannock 1411
Cannock 2546

Rhizome biomass also varies within a single stand and from year to year (Table 11); the maximum value reported was 5.1 kg m−2 (51 t ha−1), although many stands have much less than this. The ratio of short:long-shoots is also very variable (Table 11), high values reflecting established, dense stands and low values obtained from near actively invading bracken fronts. Le Duc et al. (2003) in a comparative study of several sites found considerable variation in biomass from year-to-year, with successive reductions in rhizome biomass from 1998 to 2000. They speculated that this might have been caused by recent very wet years when the soil was severely waterlogged (Poel 1961). In the Venezuelean study, the maximum biomass (fronds + rhizomes) measured was 3110 g m−2, of which two-thirds was rhizome (2110 g m−2) (Alonso-Amelot & Baechler 1996).

Table 11.  Estimated values for total rhizome dry mass (M) and dry mass ratio of short-shoots (main frond-bearing fraction) to total rhizomes (R) reported for untreated bracken
Bracken statusM (kg m−2)RReference
  • *

    Estimated from shoot dimensions, assuming constant densities.

  • Interpolated from graph of dry masses.

  • Fresh-weight ratio values reported, thus uniform mass loss on drying assumed.

  • §

    Standard errors for Carneddau 1999 and 2000 not available separately.

Advancing margin 0.15Watt (1940)*
Maximum stand height 0.13 
70 ft (21.5 m) from front 0.16 
Hinterland 0.30 
Uniform bracken (1973)  Williams & Foley (1976)
Annual average1.90.32 
Late April2.00.31 
Monthly average minimum1.60.29 
Monthly average maximum2.10.38 
October average2.10.31 
Dense stands, more fertile sites> 3.0 Daniels (1981)
Advancing margin 0.10Whitehead (1993)
Mature stand 0.30 
Middle of mature stands (four sites)1.0–1.50.18–0.47Pakeman & Marrs (1993)
Middle of mature stand  Pakeman & Marrs (1994)
Uniform area of dense litter (Autumn)  Paterson et al. (1997b)
Bracken-infested heath (late April)  Marrs et al. (1998b)
Bracken-infested moorland (Autumn) (in rank order)  Le Duc et al. (2003)
 Cannock 1 19985.14 ± 0.160.42 ± 0.011 
 Cannock 3 19982.58 ± 0.090.41 ± 0.019 
 Cannock 2 19982.67 ± 0.270.39 ± 0.012 
 Carneddau 19984.58 ± 0.080.36 ± 0.008 
 Carneddau 1999§2.880.35 
 Peak 20001.85 ± 0.040.32 ± 0.014 
 Sourhope 1 19992.89 ± 0.090.30 ± 0.017 
 Carneddau 2000§1.400.27 
 Sourhope 2 19993.00 ± 0.110.24 ± 0.014 

Estimating productivity is extremely difficult, especially for the current rhizome production, and values vary between the pioneer and the stable, mature communities. Pearsall (1959) estimated that total Pteridium production including the rhizome component at one site was 21 t ha−1 year−1, a value among the highest for herbaceous plants in the UK. In Grassland D at Lakenheath, Watt (A.S.W.) measured annual frond and rhizome biomass at 12.4 and 27.3 t ha−1, respectively. Assuming an approximate age of 30 years for the plant, this suggests a net rhizome increase of 911 kg ha−1 year−1, and a net production of 1325 kg ha−1 year−1. These estimates are conservative as they ignore turnover of organs/tissues and respiration.

Pteridium productivity has also been estimated using process-based models, based initially on the work of Williams & Foley (1976). The original model considered Pteridium in three compartments (frond biomass, rhizome biomass and rhizome carbohydrate concentration), within a five-stage annual growth-cycle (Fig. 4; Pakeman et al. 1994; Section VI). The model has provided reasonable predictions of the Williams & Foley data, and against results collected specifically to test it (Pakeman & Marrs 1993, 1994; Paterson et al. 1997a,b). The model was more sensitive to variables affecting frond productivity than rhizome transfers. The model was further developed to produce predictions of bracken productivity across the UK; comparisons were made using country-wide climate data based at sea level and running the model until an equilibrium biomass was obtained with altitudinal corrections (Pakeman & Marrs 1996). The greatest biomass is predicted at 3136 g m−2 for the south-west of England where the growing season is long and uninterrupted by frosts. Biomass was predicted to be restricted in the south-east by low water availability, and in Wales, northern England and Scotland by the short growing season as a result of frosts. The lowest predicted equilibrium biomass was 169 g m−2 for Cairngorm, Scotland. The maximum predicted value is less than that found for the greatest rhizome biomass found in the UK (Table 11); it is likely that this discrepancy results from either imprecise information in the model or lack of precision in the meteorological data needed to run the model. The meteorological data are averaged for 40 km grid squares and do not take account of small-scale microclimatic effects, south-facing slopes, sheltered areas, etc. (Pottier et al. 2005).

Figure 4.

Stages in the yearly growth cycle of Pteridium, showing generalized changes in frond biomass, rhizome tissue biomass and rhizome carbohydrate concentration, and five phases (described in text, Sections V, VI). After Williams & Foley (1976) and Pakeman et al. (1994).

(c) effects of frost, drought and waterlogging


Pteridium is very sensitive to frost; indeed frost often controls its distribution, locally via microtopography, altitudinally and latitudinally (Section II). Both air and ground frost are important constraints; air frosts tend to be non-selective over varied relief in local terrain but are less common and more sporadic than ground frosts. Autumn or early frosts may also cause premature senescence and death of the frond. In process-based simulation models, frost and low temperatures are the factors that limit the growing season (Pakeman et al. 1994).

Frost, especially as an air frost, affects meristematic tissues of both rhizome and frond, and even mature parts of the frond. Winter frost is particularly harmful, particularly when there is no snow or litter protection. In a closely grazed grassland area on Lakenheath Warren in 1940 severe winter frost killed 79% of frond buds on the rhizome and 44% of rhizome apices. Under a thickness of 6 cm of Pteridium mor, a single rhizome apex remained alive. This frost reduced frond number in the following season by 63% compared to the mean value for the previous 12 years. Moreover, the replacement fronds were derived from deeper rhizomes than usual and petioles were c. 50% shorter than pre-frost levels. It took 14 years for frond density to return to pre-frosted levels (Watt 1950).

Symptoms of spring frost vary, depending on the intensity, incidence and the state of frond development. If the frond is still succulent the effect is lethal. The apices, pinnae and pinnules of older fronds are susceptible and the shape of the crippled fronds varies, much depending on the degree of unrolling of the pinnae at the time of the frost. Fronds are often misshapen or have a contorted appearance with brown, necrotic tissue. The symptoms are often mistakenly attributed to a pathogen but they may, however, provide a focus for facultative organisms to invade and spread into live tissue. Late frosts can be particularly damaging. For example an air frost killed vast areas of Pteridium 30–40 cm tall as late as 29 June in 1954 in Breckland, Suffolk (A.S.W.). In this area, frost is particularly important as it can occur in any month, and young fronds have been killed by ground frost even in August (A.S.W.).


Pteridium is sensitive to low water supply. Drought may cause fronds emerging from below the soil surface to die. Emergent fronds may either be killed and shrivel back into the soil, or continue to grow up under the canopy as delicate shade forms, which may or may not survive depending on available moisture. Once established, Pteridium is remarkably tolerant of dry conditions, because of its thick cuticle, rapid stomatal response (Section VI) and the rigidity of the pinnae. Drought damage is rare, especially where the rhizomes and roots are relatively deep. However, this does occur, for example, during a drought in 1935 at Grassland E (podzol) on Lakenheath Warren where Pteridium was invading the closely grazed grassland. The rainfall over an 8-week period was distributed unevenly, with 20 mm in the first 2 weeks, none for 4 weeks and 3 mm in the last 2 weeks. On the 25 August 1935, there was a striking contrast between the fresh green of the sparse invading fronds and the yellow of the dead adult fronds of the mature phase (A.S.W.).


Waterlogging is a major constraint on Pteridium (Section II). The behaviour of the roots and rhizomes differs in aerated and non-aerated soils (Poel 1951, 1960, 1961; Anderson 1961; Jarvis 1964; Watt 1964, 1979). In aerated soils the roots were vigorous, profusely branched and golden in colour. In the unaerated treatment the roots were short, dark-brown, restricted to the surface of the substratum, and growth was poor. Poel concluded that water was not inhibiting but that oxygen diffusion rate, or a chemically reduced product of waterlogged conditions (CH4, H2S, Fe++), could be limiting (Poel 1961).

VI. Structure and physiology

(a) morphology

Pteridium may be visualized as a travelling geophyte (Watt 1940) having a main rhizome axis travelling horizontally in the soil and dying away behind as it moves forward.

The rhizome

The young sporophyte has a short, radially symmetrical, erect axis bearing 0–10 small, simple, spirally arranged leaves with indentations (Fig. 5j–l). Apical growth of the leaves ceases and two new buds arising from near the margin of the apical meristem produce dorsi-ventral shoots, which are negatively phototropic (Fig. 6). After penetrating the soil these two initial buds produce further branches that radiate outwards, each branch giving off alternate shoots, forming, in the adult community, a complex, much-branched underground shoot/stem system. This young rhizome can produce adult pinnate leaves at the end of the first season (Dasanayake 1960).

Figure 5.

Generalized life cycle of Pteridium. (a) Lower surface of a fertile leaf (here a pinnulet) showing the single continuous marginal sorus (coenosorus). (b) Mature sporangium (above) and sporangium after release of spores (below). (c) Spore, showing trilete mark ‘raphe’ and spore wall. (d) Germinated spore with developing prothallus with rhizoids, the earlier filamentous protonema is not illustrated. (e–h) Cross-fertilization is normal in Pteridium, archegonia and antheridia are usually formed simultaneously on one prothallus. (e, f) Smaller young prothallus with antheridia. (g, h) Larger, older prothallus bearing archegonia. (i) Spermatozoids released from the antheridia move towards the mature egg cell. (j) Developing sporeling (sporophyte growing on gametophyte). Normally only a single zygote develops from each prothallus. The first leaf of the sporeling is bipinnate; successive leaves are more complexly divided (illustrated, Gottlieb 1958). The prothallus degenerates while the growing sporeling withdraws material from it and becomes increasingly independent. The rhizome of the sporeling then develops (see Fig. 6). (k) The adult sporophyte does not become fertile (l) before 3–4 years of age (Conway 1957; Dasanayake 1960). From an original by Drs F. J. Rumsey and E. Sheffield, reproduced by Thomson (1990), and redrawn by K. Lancaster, University of Liverpool.

Figure 6.

Development of the rhizome in the sporeling of Pteridium in generalized form (see Gottlieb 1958; Dasanayake 1960; Gottlieb & Steeves 1961): (a) prior to formation of leaf 7–9, the shoot axis is directed upwards, phyllotaxy is spiral, and a single root arises beneath each leaf (Dasanayake 1960); (b) the shoot apex bifurcates; (c) each apex bends laterally, turns downwards, and then (d) the stem bifurcates again and penetrates the soil to form the rhizomes, and by this time the first rhizome-borne fronds appear as young croziers. From Thomson (1990), redrawn by K. Lancaster, University of Liverpool.

There are two main categories of rhizome shoot (long and short), distinguished only by their relationship to the frond. They are formed via a dichopodium, where one branch is relatively undeveloped and appears lateral to the other (main) branch (Bower 1928). This bifurcation occurs at the end of every shoot. The degree of suppression or underdevelopment varies: in long-shoots more rapid growth carries the abaxial frond some distance from the main stem; in short-shoots the short internodes are the successive alternate main arms of the dichopodium. The suppressed or dormant arm is represented by the bud (actual or potential) at the frond base. A typical rhizome system is illustrated (Fig. 7g–l). Field separation can be done on the basis of internode length: long-shoots with internodes > 15 cm, short-shoots with internodes < 5 cm. However, intermediate types occur (Watt 1940).

Figure 7.

Major morphological features of Pteridium in generalized form, along with recommended terminology: (a) frond lamina (blade); (b) rachis; (c) pinna; (d) stipe; (e) nectary (minor nectaries are usually present at the base of pinnae, and even some pinnules, Page 1982); (f) crozier with hairs; (g) leaf primordium borne on the short-shoot; (h) shoot apex; (i) lateral line; (j) petiolar roots (found in var. esculentum, O’Brien 1963); (k) roots; (l) rhizome; (m–p) pinna; showing (m) pinnule; (n) midrib of pinna; (o) pinnulet; and (p) midrib of pinnule; (q) lower surface of pinnule showing coenosorus continuous around margin; (r–w) section through the margin of pinnulet; showing (r) upper surface; (s) lower surface; (t) mature sporangium; (u) indusium; (v) sporangium after spore discharge; (w) false indusium (segment margin): (y–z) showing (x) frond primordium; (y) abaxial bud and (z) adaxial bud. From Thomson (1990), redrawn by K. Lancaster, University of Liverpool.

The proportion of short- to long-shoots in mature communities is probably a reflection of environmental conditions (Table 12). The factors affecting change from (i) one type of shoot to the other, and (ii) the proportion between the categories in mature communities are related to the vigour of the plant, and the environmental conditions under which it was grown (Watt 1940).

Table 12.  Relative distribution (%) of Pteridium rhizome shoot-types at different sites in East Anglia (A.S.W.)
SitePhase/areaRhizome shoot-type
Blaxall heath (Suffolk)Mature 046.941.411.7
Degenerate phase11.732.955.4 0
Lakenheath Warren (Suffolk)
 Area D33.634.931.5 0.3
Area E50.926.221 1.9
West Tofts (Norfolk)Pinus plantation71.715.3 8.3 4.7
Felsham Hall Wood (Suffolk) 2.8

The rhizome is exceedingly plastic: a long-shoot may become an intermediate one and then a short-shoot, and vice versa. The long-shoot varies in diameter from 0.5 to 3.8 cm (Conway & Stephens 1954) and increases in length by about 10–17 cm annually, and each short-shoot produces one leaf primordium during the growing season. Fronds may be produced by the apical meristems of long-shoots, but the primordia are carried away from the main axis by strong growth of the abaxial bud (Dasanayake 1960; O’Brien 1963), accounting for the earlier belief that the long-shoots were truly leafless (Watt 1940; Webster & Steeves 1958; Thomson 1990).

The length of the rhizome per unit of soil surface varies, usually between 13 and 25 m m−2 (Smith 1928; Watt 1964), with a maximum record of 202 m m−2 (A.S.W.). Rhizome buds make up around 70% of the apices of the rhizome, and they may remain dormant for up to 10 years. The physiology of dormancy is unknown but excavating the rhizome and replanting breaks dormancy. The stem and the leaf tissue develop during winter.

Long- and short-shoots differ in their behaviour: long-shoots tend to grow horizontally (albeit with much variation), whereas short-shoots usually grow erect and obliquely towards the soil surface, running parallel to it just below the surface, or above it in the humus layer or in surface Sphagnum. Intermediate-shoots also run more or less horizontally and behave as long-shoots but usually at a shallower level. There is, however, considerable variability in direction of growth in both rhizome types (Watt 1976).

As the rhizome moves forward, it dies away behind. Why it does so is unknown, but the length of the main axis (overall plant size) is related to the prevailing environmental conditions, especially rhizome depth. Generally, the shallower the rhizome, the smaller the plant. As plants grow at different rates it is not known whether rhizome death is a question of age, the breakdown in resource transport and consequent exhaustion of reserves, or the effect of external factors.

Young rhizomes are covered with white hairs, initially standing perpendicular, and internally there is thin-walled aerenchyma with intercellular spaces allowing diffusion between the air inside and outside the rhizome. With age, the cells in the plane of the epidermis of the rhizome change colour, effectively sealing the internal tissues from the outside, and the hairs are appressed to the rhizome surface as the apex pushes through the soil. The epidermis consists of thin-walled and wrinkled cells below a single line of dark brown cells; the outer and radial walls of which are strongly thickened. These are succeeded by a band of three or four cells deep; these cells are yellowish with very slightly thickened walls and contain a small number of starch grains. As the rhizomes age the hairs become at first golden-brown and then disappear leaving a black rhizome surface.

The ultimate factor restricting the rhizome to any depth may be physical (hard rock with a continuous surface) but rhizomes can penetrate the hard, compact unweathered chalky boulder clay of Breckland (A.S.W.). The limits of rhizome penetration have been reported between 5 and 97 cm along the same rhizome length (Watt 1940); however, in exceptional circumstances it has been recorded at 2.5 m (R.H.M., unpubl. data). On two occasions rhizomes have been seen to emerge from the soil into mor humus and then re-enter the soil layer c. 5 cm distant (A.S.W.).

Dead rhizomes tend to be at a lower depth than live ones (Fig. 3; Watt 1976); indeed there was a linear relationship between depth of origin of the short-shoot and its age (based on number of nodes). The younger short-shoots tend to arise from shallow rhizomes. There is, however, a fairly wide scatter and a number of short-shoots have a deeper origin than expected (Fig. 3). In extreme examples, the whole rhizome can lie wholly within the accumulated Pteridium litter, 5 cm from the surface (A.S.W.), which makes it susceptible to frost, drought, etc. (Sections II, V). The maximum depth recorded for frond differentiation in soil is 38 cm (Conway 1949), although undifferentiated buds are found on deeper short-shoots.

In mature communities new rhizomes run at a shallower depth and bear vigorous fronds, which emerge early compared to older rhizomes, giving them a competitive advantage. It is not clear how much of this differential effect is due to age or to competition from more vigorous fronds from young shoots. This process leads to suppression of the fronds and ultimate death of the old shoot. Various factors may prevent this, e.g. the death of the early emerging fronds from the young shoot from late-spring frost.

Interestingly the rhizomes act as an organ for the uptake of mineral elements (Tyson et al. 1999a,b), although the relative importance of rhizome vs. root uptake is unknown.


Roots arise on all sides of the shoot and grow in all directions, without any apparent regularity. In the radial young sporophyte a single root arises beneath each leaf (Dasanayake 1960) followed later by numerous secondary roots that become longer than the leaves. Roots arise at the rate of about 1 root cm−1 from the invading rhizome of an adult sporophyte running at 18 cm in the A-horizon of a podzol. A few growing straight down are fat (3 mm) at emergence and taper only slightly; they have a number of short stumpy (5–10 mm) horizontal branches, and reach, but do not, penetrate the B-horizon.

In pioneer communities, the roots grow as fast as the rhizome but the descending roots, although they may penetrate the surface of the humus-iron B-horizon, rarely proceed further. Others emerge from the side; they are fat but quickly taper, grow horizontally and then bend vertically upwards and grow to within c. 2 cm of the soil surface. At this stage the root system in the soil is very diffuse, but may reach 130 cm depth (Watt 1976). Roots are more numerous on the older, ascending short-shoots.

In mature communities roots are concentrated at or near the surface with a rapid reduction in length per unit of soil volume with depth. There is, however, much variation in root production and behaviour varying for example, between 2 and 6.7 roots cm−1 on long- and short-shoots, respectively, on vigorous Pteridium at Weeting Heath, Suffolk, but only 0.5 and 1.9 cm−1 in the Sphagnum at Flitwick, Bedfordshire (A.S.W.).


(i) Development: the frond base buds  Beside the short-shoot apical bud there is also, theoretically, a frond bud at the base of each frond. The frond-base buds are morphologically shoot apices that develop either into shoot apices or to short-shoots, bearing fronds. Many of these are not produced, or at least are not visible externally, and those that are produced may remain dormant for up to 18 years (Smith 1928), but most lose viability after a number of years (6 years is more usual). For example, in one excavated plot (0.36 m2) there were 207 buds (including approximately 40 apical buds) with 23–57 differentiated fronds but they only produced 8 adult fronds (A.S.W.). The number of fronds attaining maturity therefore is strikingly less than the number differentiated and this in turn is very much less than the potential production. This bud pool is one reason for the increased density after cutting and the general difficulty in eradicating Pteridium. If all these buds could be induced to produce emergent fronds, cutting as a means or eradication would be much more effective (Lowday et al. 1983; Lowday & Lakhani 1987; Lowday 1987).

(ii) Development of the young sporophyte  The first leaf and the root of the sporeling emerge virtually simultaneously from the underside of the prothallus and are distinct long before the axis is visible (Fig. 5e–h). The sporeling leaves increase in size and complexity from 2 pinnae to 3 pinnae to 5 pinnae, etc. After 7–8 leaves have been produced there is a bifurcation of the axis. Once at the rhizomatous stage the first 2–10 leaves are borne directly on the rhizome.

Initially the sporeling has an erect axis with a radial, three-sided apical cell. After c. 9 leaves, the growth of the radial axis ceases and two branches are formed dorsi-ventrally at the apex and grow down into the soil (Bower 1928). The two new buds arise from the meristematic tissue near the margin of the apical meristems. In ‘var. latiusculum’ three growing seasons are usually required for complete leaf development (Webster & Steeves 1958).

(iii) Form and size of the Pteridium plant  There has been some confusion in the literature over nomenclature of what constitutes a bracken plant. This has been clarified by Le Duc et al. (2003), who defined: (i) a bracken patch as an area of fronds comprising single- or multiple-genets and/or single- or multiple-plants; (ii) a genet has a constant genetic make-up, but may comprise many individuals (ramets or clones), i.e. independent physiological units (Daniels 1985); and (iii) a plant or ramet, which is a single genet and a physiologically independent unit. These definitions concur with those of Birch (2002).

The young sporophyte is at first a circular patch of a single individual, which then spreads centrifugally like other rhizomatous species. In time the older parts die, leaving some younger branches that develop into independent ramets, which in turn repeat the process. All the daughter ramets are a single genet, i.e. the same genetic make-up as the originator. In the degenerate phase many ramets can be found, and they often have irregular or even undefined orientations of their main axis (Watt 1940). Under these circumstances the patch is formed from ramets of a single ramet, but in other circumstances communities could contain individuals of diverse genetic origin. For example, Melville (1965) described the establishment of sporelings from prothalli in a burnt area of rather patchy Pteridium. After 8 years, it was impossible to distinguish these new genets from the original population, indicating that some patches could have some genetic diversity. There is limited knowledge of the detailed genetic make-up of Pteridium patches, but in an enzyme electrophoresis study of seven British populations, genetic heterogeneity and sub-structuring was low in six populations (Wolf et al. 1991). Thus, genetic uniformity is likely but cannot be assumed.

The clear distinction between genet and ramet clears up confusion over plant size. A single genet can cover a considerable distance (390 m, Sheffield et al. 1989; 1015 m, Parks & Werth 1993), but this probably consists of a multitude of individual clones. The size of a ramet may, however, be quite small with estimates ranging from 20 m for a main rhizome to 335 m for an entire ramet complex. Moreover, average ramet size decreases as they age: 61 m for a ramet about to become independent reducing to 12–28 m in the degenerate phase (Watt 1940). Thus, the number of ramets per unit area may vary from a few in the advancing fronts to many hundreds of smaller ones in the degenerate phase. In the degenerate phase of Grassland E at Lakenheath Warren, ramet density was estimated at 27 000 ha−1 (Watt 1940). Large variations are expected in plant density.

Coarse-scale genetic mapping indicated that a population can consist of many small genets, and a few large ones (sometime up to 1 km across). However, fine-scale mapping of the genetic make-up of individual fronds in spatially discrete patches showed extensive intergrowth of plants (Parks & Werth 1993).

Similarly, Pteridium patches can persist on the same site for a considerable period, at least 1500 years in Finland (Oinonen 1967a), although they may consist of many individuals and with continual turnover of these individuals. The genetic make-up of these Finnish patches has not been assessed.

(iv) Frond density and height  Frond density varies greatly in time and space (from 0.1 to 75 m−2). Areas of Pteridium with similar frond height and rhizome depth show considerable differences in frond density. There is no simple relationship between density and height. In unfavourable habitats (e.g. shallow dry soil and exposed situations) density and height are correlated, but in more favourable conditions, intraspecific competition may reduce density. For example, after cutting, Pteridium often has a much greater frond density but all fronds are much shorter (Lowday & Marrs 1992a). This implies that in uncut stands the taller fronds either suppress frond development through a dominance mechanism, or they suppress smaller fronds through competition. Moreover, the majority of fronds derived from young short-shoots have a competitive advantage over fronds from older shoots. Thus, both the density and the height may also be an expression of age only and not of habitat potential.

There is also considerable variation in height, which is affected by exposure to wind, soil depth and water availability. However, the most important factor influencing height is rhizome depth. The mean height measured in 33 populations in ‘rough grazings’ across the UK in 1944 was 63 cm with extremes of 29.2 cm to 145 cm. The maximum height recorded was 4.4 m (A.S.W.).

Fronds have been recorded as failing to reach the surface as a result of either damage by rolling or through physical obstacles such as trodden tracks (Conway 1949). However, these failures are difficult to reconcile with the emergence of fronds from a footpath covered with tarmac (A.S.W.).

(v) Frond form and mode of branching  The frond varies in shape, in dimensions and in proportions of its parts. The frond is long and petiolate: the lamina in short fronds is deltoid, in taller fronds it tapers more gradually at first but later more abruptly to an acute apex consisting of opposite to sub-opposite pairs of bipinnate pinnae, followed by simply pinnate, the last 2–5 cm of rachis bearing doubly pectinate simple segments of diminishing size following a triangulate apex. The rachis is channelled on the upper side. The numbers of pinnate pairs is variable. The fronds are generally arched, the degree being related to shade, but some fronds have an almost tricuspid apex, probably due to a check to extension by drought.

The pinnae and pinnules are short and are acuminate to obtuse (Fig. 7o–p). The pinnules usually vary at right angles to the pinna midrib (preferred terminology to the older and ambiguous term costa; Thomson 1990), although sometimes they are at a more oblique angle. The pinnule midribs are slightly, to densely, pubescent beneath, and less so above, and the penultimate segments are usually pinnatifid. The longest centre segment, or part of a segment, is from three to six, usually about four, times as long as broad and the ultimate segments are normally straight. Most pinnules are glabrous or slightly pubescent, with a lower surface densely pubescent (Fig. 7o–w).

The stipe is relatively long, the vascular bundles numerous, the blade coarse, the segments very numerous, ovate to linear, the sori marginal and almost continuous, the sporangia are borne between the outer indusium, the modified margin of the segment and the inner indusium (Fig. 7r–w).

(vi) Lateral lines  Two lateral lines (Fig. 7l) run through the rhizome and frond; they consist of aerenchyma, open tissue with cell walls less thickened than the adjoining sterome and with stomata in the epidermis. The aerenchyma does not bear hairs, which are found on all other parts of the plant, except mature rhizomes where they have sloughed off. Observation suggests that the width of the line and activity in the central aerenchyma vary inversely with soil aeration.

(vii) Nectaries  Nectaries (Fig. 7e) are found on the abaxial surface of the stipe at the base of each lower pinna or pinna pair (Darwin 1877; Power & Skog 1987). They are dark-brown, smooth somewhat swollen areas, and they usually lack trichomes, the exceptions being near the pinnules near the tip of the frond. They are c. 1–4 mm in diameter and are smooth protuberances raised 0.1–1.0 mm above the surface of the stipe (Power & Skog 1987). They are structured nectaries with an obvious difference between the secretory (nectariferous) tissue compared to the ground parenchyma. The nectariferous tissue consists of a specialized, thin-walled epidermis and secretory parenchyma cells; this tissue is separated from the vascular tissue by large parenchyma cells (Rumpf et al. 1994).

The nectaries secrete liquid containing sugars (glucose, sucrose and fructose) and 20 amino acids, the most common amide being glutamine, at low and variable rates with an overall concentration of 13 mm (Lawton & Heads 1984). In the northern hemisphere secretion was greatest in mid-June (maximum rate = 2.69 µL nectary−1 24 h−1), the period of initial frond expansion (Phases 2–3, Fig. 4), then declining slowly to zero in late August.

The nectaries are visited by ants and other invertebrates. Differences have been found between sites, with four and eight visitor taxa recorded from the UK and New Mexico, USA, respectively (Lawton & Heads 1984). In the UK, three Hymenopteran genera were noted (Formica, Myrmica and Leptothorax), and virtually all ants on a frond feed on a nectary during their visit (Lawton & Heads 1984). Other invertebrates also visit the nectaries (Table 13); some of these species (e.g. mites in the Anystis genus: Acarina) are potentially beneficial, some (robbers) feed on the secretions, and others (robber-herbivores) are bracken-specialists (Lawton 1976, 1982). It is tempting to believe that the association between the extrafloral nectaries and ants is mutalistic, i.e. in return for nectar the ants exclude or depress herbivore populations on Pteridium. However, ant exclusion experiments suggest that although the ants can remove caterpillars and reduce invertebrate populations, the effects on most herbivore populations, and hence Pteridium, are weak (Heads & Lawton 1984, 1985; Heads 1986; Rashbrook et al. 1992).

Table 13.  Adult insects and one species of mite recorded feeding from Pteridium nectaries at Skipwith Common, Yorkshire, between 1971 and 1982. All taxa have been recorded in most years and most are common visitors to the nectaries. ‘Robbers’ refer to species where the adults feed at the nectaries but they are not associated with the plant in any other way. ‘Robber herbivores’ refer to species where the larvae feed on Pteridium (Lawton & Heads 1984)
Potentially beneficial predators and parasitoids (excluding ants)‘Robbers’‘Robber herbivores’
Acari: ProstigmataCollembola: Entomobryidae Entomobrya sp.Diptera: Anthomyiidae Chirosia albifrons
  C. histricina
 C. parvicornis
Coleoptera: CoccinellidaeColeoptera: ScirtidaeHymenoptera: Tenthredinidae
 Adalia decempunctataCyphon padiStrongylogaster lineata
 Coccinella septempunctataC. variabilisAneugmenus sp.
 Propylea quattuordecimpunctata
 Thea vigintiduopunctata
Hymenoptera: VespidaeColeoptera: Elateridae 
 Vespula sp.Athous haemorrhoidalis 
Ctenicera cuprea 
Dalopius marginatus 
Hymenoptera: BraconidaeColeoptera: Malachiidae 
 Unidentified speciesMalachius bipustulatus 
Hymenoptera: PteromalidaeColeoptera: Nitidulidae 
 Unidentified speciesBrachypterus glaber 
Hymenoptera: MiscellaneousColeoptera: Lathridiidae 
 Unidentified taxaDienerella ruficollis 
Coleoptera: Mordellidae 
Anaspis sp. 
Diptera: Nematocera 
Miscellaneous unidentified taxa 
Hymenoptera: Apidae 
Bombus sp. 
Apis mellifera 

(b) mycorrhiza

Mycorrhizas are routinely recorded from the adult sporophyte under field conditions in the northern hemisphere, although infection levels varied considerably (Jones & Sheffield 1988). Mycorrhizas are less well developed in the young sporophyte (Conway & Arbuthnot 1949), and there are records of large specimens without an endophyte. When a mycelium is present it shows several forms. From 20 out of 24 samples in a variety of habitats in France, Boullard (1957), described (i) endophytic mycorrhiza with replete hyphae peripherally in various stages in peripheral cortical cells forming a pseudo-parenchyma, (ii) looser mycorrhiza with terminal spiracle vesicles, and (iii) others branching freely from loose balls, vesicles and arbuscules. Where vesicles are present, they were rich in oil and intact, but empty (collapsed) examples also occur. Mostly large hyphae in loose balls are found in the cortical cells where they form a brown branched, shrivelled shape. At depth, the mycelial trunks become smaller in diameter and branch.

Infection of the root occurs via root hairs; the travelling hyphae carry the ‘infection’ horizontally through the outer and middle cortex, while the digestive zone occurs only in the endo-cortex. Arbuscules and sporangioles were observed in the two or three cell-layers immediately adjacent to the epidermis (Rayner 1927). Experiments on Pteridium in pot culture with soils showed that at two levels of phosphorus (< 8 µP ml−1) growth was much enhanced by endotrophic mycorrhizas; root colonization by the endophyte may be negligible (Cooper 1975). However, a relationship between colonization and soil available phosphorus was not apparent in the field (Jones & Sheffield 1988).

(c) perennation: reproduction

Rate of spread

A distinction should be made between the enlargement of established patches of Pteridium and the creation of new patches. The increase in area is largely the result of the rhizome spread along Pteridium fronts, possibly coupled with an increase in vigour of existing patches. New patches can occur only from spores or rhizome movement (soil transport or animal movement).

The rate of growth along Pteridium fronts varies from negative values to 1.27 m/year (Table 14), although a value for 1.8 m/year has been recorded (A.S.W.). Both cutting and application of the herbicide asulam treatment can reduce encroachment rates, cutting by up to one-third and application of asulam set the front back c. 4 m (Pakeman et al. 2002).

Table 14.  Estimated rates of linear encroachment by Pteridium stands in the UK determined from aerial photography and ground measurement (modified after Pakeman et al. 2002)
SiteEncroachment rate (m year−1) (mean)Time period (years)Source
Aerial photography
 Levisham Moor, North Yorkshire0.3519Pakeman & Hay (1996)
 Ramsley Moor, Derbyshire1.27 5Pakeman & Hay (1996)
 Weeting Heath, Norfolk0–1.426Marrs et al. (1986)
 Cavenham Heath, Suffolk0–0.935Marrs et al. (1986)
Ground measurement
 Braunton Burrows, Devon< 0.3–0.9 (mean = 0.49) 6Willis et al. (1959b)
 Scotland (four heathland sites)−1.0–3.0 2Miller et al. (1989)
 Unspecified, Scotland0–0.9 5Braid (1934)
 Four sites, Durham Coal Measures0.35–1.25 2Jeffreys (1917)
 Lakenheath Warren, Suffolk0.43 8–18Watt (1954)
 Cavenham Heath, Suffolk0.4 3Marrs et al. (1986)
 Levisham Moor, North Yorkshire0.55 6Pakeman et al. (2002)
 Ramsley Moor, Derbyshire0.36 5Pakeman et al. (2002)

Under experimental conditions Conway (1949) measured rhizome growth rates in transplanted, young sporophytes (6 weeks old). After 16 weeks, growth (early September) of two plants was measured; their rhizomes were 20 and 33 cm, respectively, and both had two side branches each with a frond bud. Conway (1949) then showed that two weeks later (late September) one plant had a tip-to-tip span of 90 cm, at least a doubling in size. At the end of the season the fronds were of adult form although none bore spores. Six months later, during winter, there was an increase in the number of rhizome branches, which descended in the soil. In the second season a frond was present by the end of April and was 31 cm high. After 19 months there were 64 fronds in various stages of maturity including 43 still green and adult and some bore sporangia from the end of August. The rhizome overall length was 1.8 m with numerous strong rhizome branches, some 2 cm in diameter.

(d) chromosomes

Chromosome counts of n = 52 (gametophytic), 2n = 104 (sporophytic) characterize most plants (Jarrett et al. 1968; Sheffield et al. 1993) but sporophytic counts of 2n = 52 for Spanish and 2n = 208 for Galapagos specimens have been reported (Jarrett et al. 1968; Löve & Kjellqvist 1972). The Spanish material led to suggestions that the basic chromosome number for Pteridium was x = 26 and the most common specimens (2n = 104) are tetraploid. However, voucher specimens for the Spanish material cannot be located and specimens collected from the same area are 2n = 104 (Sheffield et al. 1989; H.A. McAllister, unpubl. data). Evidence from isozyme analysis demonstrates that sporophytes (2n = 104) behave as simple diploids (Wolf et al. 1987). It is now widely accepted that the basic chromosome number in Pteridium is × = 52 and that the Galapagos specimens represent tetraploids. One record of a triploid 2n = 3x = 156 has been reported (Sheffield et al. 1993). Unusually, for odd-numbered polyploids this plant was fertile, producing both spores and gametophytes.

Thomson (2005) considered two species as allotetraploids (2n = 208): P. caudatum (Central and South America) and P. semihastatum (Southeast Asia and Australia) (Section I). These species have a DNA content of approximately 31 pg compared to the diploid value of c. 16 pg (Thomson & Alonso-Amelot 2002; Thomson 2005).

(e) physiological data

Response to shade

Pteridium fronds show a marked difference between exposed and sheltered leaves (Bright 1928), although lamina of both types may occur on the same frond. Exposed pinnae are more xeromorphic than pinnae from more sheltered locations (essentially similar to sun and shade leaves in other species), with a stiffer, harder and thicker lamina (318 µm compared to 163 µm; Boodle 1904), with thicker, outer epidermal walls (5.7 µm compared to 2.9 µm), and a well-differentiated, nearly continuous hypodermis. Exposed fronds are also darker green than ones from sheltered locations; the veins are more sunken and the subdivision of pinnae is not usually so great.

The numbers of pinnae decrease with exposure, exposed leaves being smaller. The position of the bend of the frond also varies with exposure, with erect fronds in less-exposed conditions, and more inclined fronds in exposed conditions.

Water relations

Pteridium is not found in the driest places and cannot tolerate exposure to extreme drought (Section V). In a comparison of water loss from detached turgid fronds of Dryopteris filix-mas and Pteridium, Dryopteris showed an initial rapid water loss, followed after stomatal closure by a much slower rate of loss. However, after the initial water loss in Pteridium, the stomata closed quickly and thereafter there was almost no loss, the assumption being that the cuticle was an effective barrier to water loss. The number of stomata has also been shown to reduce with increased exposure (Bright 1928); on three ridges at Hindhead Common (Surrey), stomata density (stomatal number cm−2) reduced as follows with increasing exposure: Ridge 1: 327 > 291 > 240; Ridge 2: 291 > 236 ≃ 243; Ridge 3: 298 > 273. This implies that Pteridium is an efficient xerophyte. However, in experiments where temperature was increased, and both drought and nitrogen addition were imposed on Pteridium in competition with Calluna, physiological damage has been reported under drought conditions (Gordon et al. 1999a,b). Water use efficiency (WUE) in Pteridium increased in response to drought or nitrogen addition; Pteridium was the weaker competitor because Calluna depleted water from the Pteridium rooting zone, and Pteridium increased its WUE (Gordon et al. 1999a).

Photosynthesis and respiration

Pteridium has relatively high maximum mean rates of photosynthesis, for example in ambient air maximum rates recorded are between 9 and 11 mol CO2 m−2 s−1, which are two to three times greater than other ferns (Hollinger 1987; Caporn et al. 1999). Net photosynthesis rates increase in response to a 200 µmol mol−1 increase in CO2 concentration, but this did not translate into effects on growth or allocation of dry mass (Whitehead et al. 1997; Caporn et al. 1999). Photosynthesis rate and biomass were increased significantly by added nutrients (N and P), especially early in the season, where assimilation rates could be increased to between 14.6 and 16.6 µmol CO2 m−2 s−1 (Caporn et al. 1999).

Photosynthetic rate is greatest in early morning and reduces during the day as a result of high solar radiation and plant temperature and reduced moisture availability (Steele et al. 1996). Pteridium has a temperature optimum between 10 °C and 25 °C and is inhibited by more than 21% oxygen. The CO2 compensation point is between 30 p.p.m. and 70 p.p.m. CO2. Respiration increases to a maximum at 50 °C, but is reduced by a decrease in water content (Johansson 1923, 1926).

NAR and translocation in Pteridium are essentially the same as in higher plants (Schwabe 1953; Whittle 1964). NAR for the young sporophyte ranged from 0.47 to 0.68 g dm−2 week−1, and was within the range shown by Hordeum spp. Translocation occurs in sieve cells rather than in sieve tubes as there are no companion cells in Pteridium.

Response to nutrients

Pteridium responds to N (especially as NH4-N), P and K additions in a manner similar to angiosperms (Schwabe 1953; Conway & Stephens 1957; Whitehead et al. 1997; Caporn et al. 1999; Gordon et al. 1999a,b). The sporeling grows well in the presence of Ca and cannot therefore be regarded as a calcifuge, although spore germination was reduced on lime-treated medium (Schwabe 1951, 1953). A high N supply increased leaf number but its effect was impeded by P deficiency. N deficiency leads to marked leaf chlorosis, a reduction in growth, and an increase in the proportion of dead leaves. P deficiency is much more serious than K deficiency. P deficiency caused a reduction in leaf size; the leaves became hard and brittle, with brown necrotic areas and an increased hairiness. There was also an increase in the number of buds but a considerable reduction in rhizome mass. Increased K supply accelerated the death rate of the older leaves; often the majority of the lamina was dead, but the tips were alive.

Chlorosis due to Mn deficiency occurs; it causes small, yellow fronds, with poor sporulation and which died prematurely (Hunter 1942). The Mn concentration in a healthy frond was 27 times that of a chlorotic frond, and chlorosis appeared to be intensified by a high tissue Ca content interfering with Fe mobility.

Pteridium can also sequester nutrients very efficiently when they become available; for example after tree harvesting Pteridium rhizomes had greater concentrations of P, K and Mg than in control sites (Lederle & Mroz 1991).

Prothalli (gametophytes) are also sensitive to mineral deficiency (Schwabe 1951). Lack of N reduced growth and the prothallus was chlorotic; P deficiency resulted in severe deficiency symptoms, producing a pale, almost-transparent, greyish-green prothallus, with some male gametangia but no sporelings; K deficiency reduced prothallus size, but the effects were comparatively small compared to N and P deficiencies. Ca deficiency produced no effect, and lack of Mg had only a slight effect with a moderate chlorosis degree and S deficiency had no effect on the prothallus but the early sporelings were stunted and the leaves markedly chlorotic. Spore germination was particularly high on NH4-rich media, but NH4 toxicity was found in the sporophyte under P deficiency (Schwabe 1953).

(f) biochemistry

Carbohydrate and nutrient fluxes between frond and rhizome

In Britain, the carbohydrate and nutrient composition of the fronds and rhizomes varies seasonally; essentially the rhizome operates as a storage organ, at first acting as a source that transfers carbohydrates and nutrients to fronds, but switching to a sink after the fronds achieve maximum size. This is generally described in five phases (Fig. 4, modified by Pakeman et al. 1994; from Williams & Foley 1976). The phases are:

  • 1During the winter all the living biomass is below ground, and rhizome biomass is lost through death or respiration.
  • 2Frond expansion is fuelled initially from rhizome reserves, and then by a combination of rhizome reserves and the photosynthate produced after frond emergence.
  • 3The source–sink relationship between frond and rhizomes then shifts so that some photosynthate continues to be used for frond growth while some is translocated to the rhizome.
  • 4Frond growth ceases, and all photosynthate other than that used for frond maintenance is translocated to the rhizome to replace carbohydrate reserves.
  • 5Carbohydrate stores are replenished and so surplus photosynthate can be used to produce new rhizomes and roots.

The interesting point is the factors that control the changeover between these phases. The start of frond expansion (onset of Phase 2) is probably controlled by increasing soil temperature (Ader 1990; Pitman & Pitman 1990) but it is not clear whether the trigger is a single temperature or an accumulated temperature threshold (Pakeman et al. 1994). The point at which Phase 2 changes to Phase 3, i.e. the point where the frond no longer depends on rhizome reserves, is approximately the point at which the frond is two-thirds its maximum dry mass, a point when the rhizome reserves are at a minimum (A.S.W.). Finally, in the northern hemisphere Phase 1 is reinitiated after frond death, usually as a result of the first autumn frost; early spring frosts cause frond death and there is an additional demand on rhizome reserves to fuel frond expansion. In other parts of the world these fluxes may not be as apparent as Pteridium may grow throughout the year.

Secondary plant compounds

Pteridium can contain a wide range of secondary plant compounds including sesquiterpenoids, ecdysones, cyanogenic glucosides, tannins and phenolic acids (Cooper-Driver 1976). It is presumed that these compounds confer a selective advantage to Pteridium in terms of reducing herbivory, increasing disease resistance, or providing an allelopathic effect (Sections IV, VI). A large number of compounds has been identified (Section VI) including shikimic acid (Evans 1976a), braxin A1, aquilide A, quercetin, cyclohexanecarboxylate, ptaquiloside, pterosins, potelosides, pterianosides, inter alia (van der Hoeven et al. 1983; Ojika et al. 1984; Saito et al. 1987, 1989, 1990; Alonso-Amelot et al. 1992, 2000; Ngomuo & Jones 1996), and their derivatives, including illudane-type and proto illudane sesquiterpene glycosides (Castillo et al. 1997, 1999, 2000).

Pteridium may also contain at least three ecdysteroids, the moulting hormones of insects: ponasterone, ecdysone and ecdysterone (Svatoš & Macek 1994). Ecdysone derivatives were active when injected into locusts but not when fed to them as their chief diet.

These chemical differences have been used in chemotaxonomy, leading Cooper-Driver (1976) to suggest that Pteridium was a monospecific genus. However, Bohm & Tryon (1967) identified varietal differences on the basis of presence/absence of certain organic acids including o-coumaric acid, which was present in ‘P. aquilinum var. caudatum’ but not in ‘var. latiusculum’.


Cyanogenesis, found in many plant taxa, acts as an antiherbivore defence mechanism. Cyanogenesis is a highly variable polymorphic characteristic of Pteridium found in both gametophyte and sporophyte under natural conditions (Cooper-Driver & Swain 1976; Hadfield & Dyer 1988). It does not correlate with morphological characters, but young fronds tend to be more cyanogenic than older ones, a typical feature of anti-herbivore defence strategies (Cooper-Driver & Swain 1976). The condition appears greater in some environments than others, e.g. in deep shade (Schreiner et al. 1984b). A field study confirmed that the amount of sheep and deer grazing was related to the proportion of acyanogenic fronds in the population (Cooper-Driver et al. 1977). Population assessments of insects on cyanogenic and acyanogenic fronds indicated that numbers were significantly lower on the cyanogenic ones, and subsequent feeding tests showed larval growth was slower, and mortality greater (Schreiner et al. 1984a,b).

VII. Phenology

(a) appearance and growth of shoots

In Britain, the new frond is usually first visible on the short-shoot in July; it grows slowly during the following months to emerge above the soil surface in the spring (April/May/early June) of the following year. After emergence, frond growth is at first very rapid, overlapping the end of the ‘light phase’ in deciduous woodlands, and unfurling completely into mature fronds in July. Thereafter, the fronds senesce, but generally they last until October, or occasionally November. However, departures from these general dates are common, with temperature (mainly frost) and drought being the most important controlling variables. Early frond emergence has been recorded in March on the Solway, and in April in the south and west of Britain. Frond emergence is usually sigmoidal, but this relationship can be disturbed by death due to frost.

Fronds from young short-shoots are initiated and develop earlier than those from older shoots, and if they are not killed they form a canopy and inhibit the development of fronds from older shoots. In a dry spring in Breckland, late-developing fronds dry up either before or after emergence: in wet springs, they emerge, become etiolated and chlorotic, and die under the canopy of earlier emerging fronds. In places liable to annual spring frosts, the earlier emerging fronds will be killed and the late emerging fronds from the older shoots are given a greater chance of development and survival (A.S.W.). That fronds from young shoots grow faster and emerge earlier is of great importance in understanding the behaviour of a community. Early emergence predisposes these fronds to spring frost damage.

An example of temporal variation at a stated place is given in Watt (1950) where fronds first emerged on 7 June in 1940, whereas in the preceeding 4 years, they had first appeared in early May. This one-month delay was explained by a sequence of factors hostile to young frond development. Between 1937 and 1939 emergence was early and rapid, but late frosts killed 59% of the fronds and crippled the survivors, which attained a mean height of 32 cm compared to a mean of 54 cm, and a maximum of 68 cm for the period 1934–48. The reduction in frond numbers allowed new fronds to develop. However, the early development and growth of these new fronds was rapid and they were particularly susceptible to severe frosts in the following winter; by March 1940, 79% of fronds and 44% of rhizome apices had been killed. Surviving fronds were either from deeper origin, perhaps developing from the long-shoots, or had developed after the incidence of winter frosts. Drought enhanced the effect of the frosts, between May and June 1940 only 26.9 mm of rain fell. The net result was both the delay in frond emergence and a reduction in frond density to 10 fronds m−2, compared to a mean density of 28 and with a maximum of 58 fronds m−2 (1934–48). Frond emergence and growth may be rapid (Watt 1940), for example in 1937 when spring frost damage was relatively low. However, in other years checks to the growth of fronds were found, usually frost, but drought can also be significant.

The primary factor controlling frond emergence is presumably temperature; in March, invading rhizomes at 30 cm depth were recorded with active apices and roots at 6.2 °C while at lower depth, under thick litter, shoots were still dormant at a temperature of 5.1 °C. Frost is also an important factor; winter frost such as in 1940 delayed emergence. The earlier emergence within woods rather than outside is probably related to the perennially higher temperature during winter rather than the influence of temperature just prior to emergence. Otherwise in the open, current temperature seems important.

Other factors affecting the date of emergence include the productivity of Pteridium in the previous season, the amount of litter protection available, the temperature just prior to and during emergence, the soil temperature, the age of the shoot and the impact of frost damage (if any) to fronds in the previous year. These factors also affect the age composition of the stand.

(b) rhizome and root growth

In Britain, the growth of the rhizomes (long- and short-shoots) is relatively fast between May and October and slow, but steady, during the rest of the year (Watt 1940). Information on root growth and development is lacking.

(c) sporulation

The receptacle can appear as early as mid-May, the sporangial initials in mid-June and spores are mature 6–8 weeks later, but timing varies with the weather. The spores develop at various times; hence spore maturation may cover an extended period. The normal period of sporulation extends from the end of July to late September or even early October with a maximum in August to September (Conway 1957). Air-borne spores caught in a trap at Cardiff over the years 1956–71 confirm the peak in August to September but they occurred between June and November with exceptional captures (once each) in December 1959 and in March 1971. Such ‘out of season’ spore detection noted elsewhere (Peck 1974; Coles 1988) has been attributed to re-flotation and spores arriving from other continents (Caulton et al. 1995). In Edinburgh between 1989 and 1993, peak spore densities were low, never exceeding 2 spores m−3 air, and the total spore count was < 1% of ‘total pollen’ (Caulton et al. 1995). In this study, no spores were detected in July and the season lasted from August for approximately 9 weeks, with the peak count being detected from late August to mid-September depending on season.

In other situations where bracken is a local feature, spore densities are a much greater fraction (25–30%) of the total spore:pollen deposition (Peck 1974; Coles 1988), and spore densities of 1750 spores m−3 (Lacey & McCartney 1994). These authors also showed that most spores were released between 08.00 and 11.00 hours GMT.

VIII. Reproductive characteristics

(a) reproduction of sporophyte

Once established as a young sporophyte, reproduction is primarily vegetative and establishment by spores is relatively uncommon under current conditions. In culture, sporelings may produce spores at the end of the second year (Conway 1949), but in the field it is normally at the end of the third or fourth year (Conway 1957). All fronds are potentially fertile; there is no morphological distinction between infertile and fertile fronds, although genotype influences the amount but not the rate of sporangial development (Wynn et al. 2000; Wynn 2002).

On fronds of the mature plant, the basal pinnae are most fertile and the first to produce spores, their numbers diminishing, and their appearance delayed, as the upper pinnae unroll. There are usually 19 spore mother cells per sporangium. The potential production of spores is large, estimated at 2 × 104 spores cm−1 sorus length. A frond 173 cm tall with 13 pairs of pinnae had 9000 sorus-bearing segments, a total sorus length of 15 870 cm and produced an estimated 3 × 108 spores (Conway 1957). As 1 mg contains c. 2 × 106 spores, a single frond may therefore produce 150 mg spores.

Sporulation is influenced by a complex of factors, including: age of plant, genotype, developmental stage of fronds, seasonal weather conditions and site environmental factors such as exposure, shade and wind (Conway 1957; Wynn et al. 2000). ‘Mast’ years occur (Page 1990) and preliminary explanations of the controlling variables suggest that these include climatic (sunny, warm springs), high light and the energy balance of the plant (Kendall et al. 1995; Wynn et al. 2000). However, attempts to interpret mechanisms have proved elusive (Kendall et al. 1995; Wynn 2002).

There is considerable variation in sporulation in space and time: in the west of Scotland two out of 19 sites had no sporulation, five sites under 10% and the rest varied between 10% and 96%, and through time sites varied from 4% to 49% over a 3-year period (Conway 1957). This variation in time and place may be related to the make-up of the frond population and the factors affecting it. Defoliation, artificial or natural, or even severe crippling by spring frost affects the percentage of fertile fronds (Wardlaw & Sharma 1963). Late-emerging fronds do not produce mature spores. Thus, in a series of plots across Britain where Pteridium was cut once (May), cut twice (May and June) and cut thrice (May, June and July), the mean percentage fertility in the replacement population was 37%, 5% and 0%, respectively, compared with 36% in uncut controls, although the variability between sites was high (Conway 1957).

Drought (which is common for example in Breckland) also checks sporangial development. In the west of Scotland a wet August and September allowed maturation of spores but prevented their discharge (Conway 1957). Temperature does not seem to be critical, for a dull and cold summer followed by a dry period in the late summer was succeeded by very good spore production (Conway 1957). Spore production is also reduced in shade (Schwabe 1951, 1953; Conway 1957; Dring 1965) but according to Schwabe once sporangial development begins even complete darkness will not check it, although other factors may. Wind exposure stops or delays spore production (Bright 1928). In one exposed area with fronds 18–23 cm high, there was 100% sterility compared with 96% fertility in a sheltered area 33 m away with 61–76 cm fronds (Conway 1957).

(b) discharge and dispersal of spores

Spores can be released suddenly through the drying of the annulus, and they may be dispersed by wind and water (Peck 1974). In a closed Pteridium stand 2.8 × 106 spores were estimated to be in 1 g of surface soil (0–2.5 cm depth and a surface area 1.7 cm2), with lower numbers in the underlying mineral layers (5–7.5 cm = 0.2 × 106; 10–13 cm = 0.1 × 106). Peck (1974) also provided estimates of dispersal distance; she estimated a reduction to 50% of source value at 0.15 km. Spore input into four adjacent areas varied from 4300 cm−2 year−1 in the middle of a Pteridium stand, between 240 and 460 in reservoirs to the south-west and 1000 in a Callunetum stand 600 m to the north, reflecting both the effect of the prevailing wind and rapid decrease in number with distance. The longest distance by water transport recorded by Peck (1974) was 2.5 km; the number of spores in the suspending waters varied with the size of the floods, run off and their incidence during the following months. Increasing discharge rates resulted in increased spore concentration (Peck 1973). Estimates of spore influx to reservoirs depended on the methodology used, with directly measured estimates being 2–3 times greater than estimates from sediment cores of known age. Peck (1973) put this discrepancy down to difficulties in scaling between short- and long-term measures, but it is equally likely that there is a considerable loss of spores through time to adverse environmental or biotic factors.

Spores are consumed by Collembola; prothalli are eaten, and rhizoids can be bitten off by Isotoma viridis and Lepidocyrtus cyaneus (both Collembola). Parasites and pathogens include the algae Chlamydomas sp., Chlorella sp., Protococcus sp. and Stichococcus bacillaris (Conway 1949), and the parasitic fungi Tilachlidium sp. Preuss (Ascomycota: Hypocreales) and Coniothyrium sp. Corda (Ascomycota: Leptosphaeria) (Hutchinson & Fahim 1958). Prothalli grown on mineral agar are potentially susceptible to attack from a wide variety of pathogens of higher plants. Moreover, spore germination and the young prothallus are both susceptible to competition from other plants (Schwabe 1951).

(c) germination of spores

Spores germinate without a period of dormancy, and if sown within a week of collection can show up to 95% germination within 5 days. Spores stored in a dry place may retain their viability for several years although they often take longer to germinate and prothallus development may be sluggish.

Germination takes place readily at cool temperatures between 10 °C and 16 °C, germination continues up to 25 °C, but it is prevented at 35 °C. Germination is slowed at low temperatures, but it is not completely stopped even at 1–2 °C, at which a few spores germinate after 4–5 weeks. Spores often survive the winter (Conway 1949, 1953, 1957; Melville 1965). Germination takes place in both the dark and the light, although rate is affected by duration of illumination. There is a time lag with shorter periods of illumination and shade (Conway 1957). High atmospheric humidity is no substitute for liquid water in the substratum. With liquid water, germination takes place even if the relative humidity is as low as 15%.

Germination takes place over the pH range 4–8, with the maximum in the range pH 5–7.5, and a reduction near the extremes. Germination is prevented at very low pH (e.g. 3.1). Spore germination is particularly high in ammonium-rich media (Conway & Stephens 1957).

There are few records of Pteridium spores being incorporated into diaspore banks. For example, in over 700 soil samples taken from bracken-infested areas, no Pteridium sporophyte has been produced (J. Ghorbani, pers. comm.), although spores of other fern species are common (Lindsay et al. 1995). One explanation on the low number of Pteridium spores accumulating in diaspore banks is that all of the spores germinate rapidly into prothalli on the soil surface, or if incorporated into the soil via dark germination (Lindsay et al. 1995). Pteridium is unusual amongst British ferns in that it does not require light for spore germination (Dyer 1990).

Even though regeneration from spores is apparently low under current conditions, it must have happened in the past because of the widescale occurrence of Pteridium. Even today regeneration from spores cannot be dismissed, as it must occur under some environmental conditions, although perhaps at a very low frequency in space or time (Dyer 1990). Observations also suggest that regeneration via the prothallus pathway appears to be correlated with major disturbance such as fire (Fritsch 1927; Lousley 1946; Oinonen 1967b; Dyer 1990). It is likely that with a long-lived, clonal, perennial such as Pteridium, the prothallus pathway occurs very infrequently, and thus its appearance is very difficult to predict.

(d) gametophyte morphology

In suitable conditions the germinating spore produces a rhizoid filament of 4–5 cells. The apical one divides forming a group and eventually, the cushioned, plate-like, heart-shaped prothallus c. 1 cm across (Fig. 5d). This may be formed in c. 18–19 days and the reproductive organs in 15–27 days (Fig. 5e–h); the antheridia develop first, followed later by the archegonia. The temporal separation in sexual maturation provides an opportunity for cross fertilization.

Experiments to determine the critical limits of survival have not been done, but experimental and field data show that survival and the rate of development are affected by temperature, light flux and duration, nutrients and moisture. At low temperatures (1–3 °C) the prothallus grows slowly, but on later exposure to higher temperatures (8–10 °C) growth becomes normal. The more mature prothalli appear to be able to withstand periods of cold without noticeable damage (Melville 1965). The filamentous prothallus can survive darkness for at least 10 weeks: then after 3 weeks of exposure to full daylight the culture becomes green. Duration of illumination affects the rate: after 3 days’ exposure to different daylengths (i) the prothallus was still largely filamentous with no reproductive organs with 6 h light day−1; (ii) it attained its mature shape with a few antheridia and no archegonia with 12 h day−1; and (iii) it had many antheridia and some archegonia with 18 h day−1 (Conway 1949).

The gametophyte responds positively to additional N supply (particularly as (NH4)2SO4) and to K (attaining dimensions of 1.5 × 1.75 cm compared with 0.5 × 0.7 cm for the control). It also grows well in the presence of Ca but may be more subject to pathogenic fungi. Prothallial growth was stimulated by the addition of glucose or sucrose and the presence of contaminant fungi (Hurel-Py 1950a,b). Volatile metabolites from other fungi were neutral or inhibitory (Hutchinson 1967). No mycorrhiza has been reported from the prothallus.

(e) reproduction of gametophyte

Water is essential for the release of antherozoids from the antheridium and as a medium for the antherozoids to swim to the neck of the archegonium where the ruptured neck canal cells release an attractive substance. This substance is not specific to the mature archegonium, since antherozoids are attracted also to the ruptured cells of prothalli and sporophytes and water in which a mature archegonium has been previously immersed (Wilkie 1954). Flooding the prothalli can induce the production of sporelings of a uniform age.

Experimental populations tested to date are cross compatible. Incompatibility has been demonstrated in three populations of Pteridium (Wilkie 1953, 1956). Two mating types have been detected in each of the populations, spores being taken from a single plant in each case. The three populations were cross compatible in all combinations of mating types and in a single gene; a multiple allele system of incompatibility is postulated. Incompatibility appears to operate between the time of antherozoid entry into the mucilage of mature archegonia and their arrival at the egg surface. Incompatibility acting in the zygote is considered unlikely. Incompatibility alleles appear to be rather weak in their action because selfing was 8%, 16% and 17%, respectively; each population produced haploid gametophytes (n = 52) and diploid sporophytes (n = 104), indicating a normal sexual cycle. Selfing in one of the populations gave rise to abnormal sporophytes, one diploid, the other haploid (Wilkie 1956).

Both apogamy (production of sporophytes without fertilization) and apospory (production of gametophytes from sporophytic tissue in the absence of meiosis) can be readily induced in Pteridium. Apogamy may be induced in ‘Pteridium var. latiusculum’ experimentally in sterile culture when sugar (glucose, sucrose, maltose or fructose) is added to the medium (Whittier & Steeves 1960, 1962; Whittier 1966; Sheffield 1992). Apospory is also rapidly induced in gametophytes grown under sterile conditions on agar in the presence of sucrose and mannitol (Sheffield & Bell 1981). Sheffield (1992) tentatively suggested that apospory might be an explanation for the origin of polyploids.

On compost enriched with ammonium salts, unfertilized prothalli showed much proliferation of plates of prothallial tissue. Even the archegonial cushion gave rise to proliferations and later archegonia were produced on these secondary growths (Conway & Stephens 1957).

(f) ecology of gametophyte

The most important factors likely to affect survival and development are drought (especially dry, cold winds) for the prothallus, and drought and frost for the young sporeling (Conway 1957).

(g) hybrids

None reported.

IX. Herbivory and disease

(a) animal feeders and parasites


Pteridium supports different faunal assemblages depending on geographical location. A comparison of fauna on Pteridium between Britain, Australia and Papua New Guinea (PNG) found Hemiptera, Diptera and Lepidoptera in common to all three countries, but some groups were specific to each region: Thysanoptera in Britain, Hymenoptera in Australia and Coleoptera in PNG (Shuter & Westoby 1992). In Britain, the invertebrate fauna is relatively low; 27 core species of arthropod have been recognized as feeding exclusively, or with some regularity, on fronds (Table 15). In addition a further 11 species may either feed on the rhizome or have been recorded as occasional feeders. The most surprising feature of this list is the scarcity of Coleoptera (only one species, and even this is uncommon, Table 15). Two genera appear therefore to have been associated with the plant for a long time; Dasineura galls (the species is not yet clear) and a large number of Chamobates mites (Oribatida) which appear to be the same species as on present-day Pteridium. Both have been found on preserved Pteridium (nearly 2000 years old) from the Vindolanda Roman excavations in Northumberland.

Table 15.  Arthropods known to feed, or suspected to feed, on Pteridium in the UK. This list has been provided by J.H. Lawton (pers. comm.), who has updated and corrected earlier lists (Lawton 1976, 1982, 1984, 2000). Here, monophagous refers to species which feed only on bracken; oligophagous to species which feed on bracken and other ferns; and polyphagous refers to species which feed on bracken and a wide range of other plants, including angiosperms. Comments on the status of each species refer only to their presence on bracken. The first 27 species make up the core assemblage exploiting the above-ground parts of the plant. Additional species follow at the end; these species either feed below ground on the rhizome, or there are few records of them feeding, or possibly feeding, above-ground
Taxonomic group Notes
(a) Core species
Collembola (springtails) Bourletiella viridescens Gisin. All stages feed on the surface of the pinnae, although exactly what they eat is unknown. Polyphagous. Very common and widespread.
Hemiptera (bugs) Bugs have piercing mouthparts used to feed on plant fluids (the contents of cells, phloem or xylem). All life-history stages exploit bracken. There are two main suborders.
 Heteroptera (‘True’ bugs)Monalocoris filicis (L.). Fern bug. One of only a few insects world-wide known to feed on the developing sporangia. Oligophagous. Can be common; widely but patchily distributed.
HomopteraDitropis pteridis (Spin.). Monophagous. Very common and widespread.
(Plant hoppers, leaf hoppers, aphids, etc.)Macrosiphum ptericolens Patch. Bracken aphid. Probably monophagous, but possibly oligophagous (overwintering host, if any, unknown). Erratically distributed, with wide population fluctuations.
First described for Britain by Lawton & Eastop (1975).
Philaenus spumarius (L.). Spittle bug. Polyphagous. Common and widespread.
Lepidoptera (butterflies and moths) The caterpillars of several species of moths feed on bracken. Unless otherwise stated, all live externally on the plant and exploit the pinnae.
 ‘Macro-moths’Ceramica pisi (L.). Broom moth. Polyphagous. Scarce but widespread.
Euplexia lucipara (L.). Small angle shades. Polyphagous. Scarce but widespread.
Lacanobia oleracea (L.). Bright-line brown-eye. Polyphagous. Rare.
Petrophora chlorosata (Scop.). Brown silver-line. Monophagous. Common and widespread.
Phlogophera meticulosa (L.). Angle shades. Polyphagous. Scarce but widespread.
‘Micro-moths’Olethreutes lacunana (D. & S.). Caterpillars live in small shelters formed by spinning silk to fold the tips of the pinnae. Polyphagous. Scarce but widespread.
Paltodora cytisella (Curt.). Caterpillars mine the rachis, giving rise to characteristic gall-like swellings. Monophagous. Common and widespread.
Diptera (‘True’ flies) The larvae of several species of flies feed on bracken. They are drawn from three families.
 AnthomyiidaeChirosia albifrons (Tiens). Larvae live communally, forming necrotic pits and ‘open’ mines in the tips of newly unfolding pinnae. Probably monophagous. Never common but reasonably widespread.
Chirosia albitarsis (Zett.). Larvae mine the rachis, particularly basal part. Probably monophagous. Rare.
Chirosia histricina (Rond). Larvae form characteristic blotch-mines in the tips of the pinnae. Probably monophagous, but may be oligophagous. Very common and widespread.
Chirosia parvicornis (Zett.). Larvae mine the tips of the pinnae, causing them to role up into characteristic gall-like structures. Oligophagous. Very common and widespread.
Cecidomyiidae (gall midges)Dasineura filicina (Kieff.). Larvae form small, rolled galls on the edge of the pinnae. Galls black when mature. Probably monophagous. Very common and widespread.
Dasineura pteridicola (Kieff.). Larvae form small, rolled galls on the edge of the pinnae. Galls green when mature. Probably oligophagous. Less common than filicina, but widespread.
Agromizidae (leaf-mining flies)Phytoliriomyza hilarella (Zett.) Oligophagous.
Phytoliriomyza pteridii (Spencer). Monophagous.
Both species form wiggly mines in the pinnae. Can be identified only from adults; mines cannot currently be distinguished. P. pteridii appears to be the commoner. Phytoliriomyza mines are common and widespread.
Hymenoptera, Symphyta(Sawflies)The caterpillars of several species of sawflies chew the pinnae.
Aneugmenus fürstenbergensis (Kon.). Oligophagous. Scarce.
Aneugmenus padi (L.). Oligophagous. Common and widespread. Parthenogenetic.
Aneugmenus temporalis (Thom.). Oligophagous. Scarce.
Stromboceros delicatulus (Fall.). Oligophagous. Reasonably common and widespread.
Strongylogaster lineata (Christ.). Oligophagous. Common and widespread.
Tenthredo ferruginea Schr. Polyphagous. Scarce, widespread.
Tenthredo sp. Caterpillars of several other polyphagous species of Tenthredo have been recorded from bracken in the literature, at least one of which (never associated with an identifiable adult) occurred at Lawton's long-term study site at Skipwith Common (Lawton 2000).
(b) Additional species
Lepidoptera Three rhizome-mining species in the genus Hepialus (Swift moths), namely H. fusconebulosa
(De Geer), hecta (L.) and sylvina (L.), may be regular feeders on bracken. Indeed, the literature records
H. fusconebulosa as monophagous (the other two are reported to be polyphagous).
Diptera (‘True’ flies)AnthomyiidaeA fourth species of anthomyiid fly in the genus Chirosia (C. crassiseta Stein), apparently mining the rachis, but has never been observed by Lawton.
Other species Chirosia flavipennis (Fall.), Strongylogaster macula (Klug.), S. xanthocera (Stephens), and adults of a beetle, the Garden Chafer Phyllopertha horticola (L.) are recorded as feeding, or possibly feeding, on the above-ground parts of the plant in Britain, but they have never personally been encountered by Lawton.

On Skipworth Common, Yorkshire (a classic study site), regular sampling of the fauna in a pure Pteridium stand was undertaken in 1972 and 1973 (Lawton & Eastop 1975). One of the commonest insects was a large rather strikingly vivid dark green aphid, first recorded in May 1972 and identified as Macrosiphum ptericolens, previously known in North America, where it is widespread, mostly on Pteridium. It was rare in early summer but increased in numbers from August onwards. From mid-August until the end of September all the fronds on the site had aphids on them, often in large numbers; a summer frond density of 34 fronds m−2 had a population of 4500 aphids m−2.

Spore-feeding insects have been described (Srivastava et al. 1997), including a Lepidopteran species Oruza divisa, and two Hemipterans (Bryocoris pteridis and Monalocoris filicis), as well as some species of general fern spore feeders within the Lepidoptera (e.g. in the genera Pachyrhabda, Calicotis, Thylacosceles, and possibly Psychoides). Nectar feeders are described in Section V1.

Ixodes ricinus, the sheep tick, is commonly found on the Pteridium frond or in the humid environment of the Pteridium litter, waiting for the next vertebrate host. The tick can act as a host for various organisms responsible for disease in lambs (pyaemia), in sheep and lambs (Louping ill) and in humans (Lyme disease). Lyme disease is caused by a spirochaete Borellia burgdorferi (Brown 1995).

Little is known about the change in invertebrate community after Pteridium is controlled. In a local survey, numbers of Hemiptera and Coleoptera were significantly greater in lowland heath compared to an adjacent Pteridium patch, whereas Diploda showed the opposite trend (Table 16a). Where the Pteridium had been controlled, the Araneae, Collembola, Hemiptera and Coleoptera increased, whereas the Diploda, Isopoda and Opiliones declined (Table 16b).

Table 16.  Comparison of invertebrate fauna in (a) lowland Calluna heath vs. dense Pteridium stands (n = 8 ± 1 SE), and (b) the same Pteridium stand where control treatments had been applied (n = 16). Pitfall traps were set out at Cavenham Heath, Suffolk for 14 days between 4 and 14 May (late spring) and 7–21 July (mid summer); the cutting treatments had been applied for 12 years and the asulam applied once in 1978 and repeated in 1984 (Lowday & Marrs 1992a); data from S.J. Stevens, reported in (Pakeman & Marrs 1992)
(a)Invertebrate groupLate-springMid-summer
Hemiptera 7.0 ± 3.0 2.0 ± 0.7 8.0 ± 1.40.4 ± 0.2
Coleoptera16.0 ± 4.410.0 ± 2.912.0 ± 4.34.0 ± 1.0
Diploda 1.0 ± 0.213.0 ± 2.2 0.5 ± 0.33.0 ± 1.1
UntreatedCut once per yearCut twice per yearAsulam appliedLSD (P < 0.05)
 Diploda13.1 3.5 2.9 5.9 2.6
 Isopoda31.412.8 7.111.310.9
 Opiliones13.4 3.0 1.1 5.9 4.5
 Araneae17.824.616.614.0 6.1
 Collembola35.970.865.9 4.322.5
 Hemiptera 1.9 4.6 4.6 4.1 1.2
 Orthoptera 0 0.06 0.25 0.06 0.11
 Hemiptera 0.4 5.3 8.9 5.8 2.7
 Orthoptera 0 0.13 0.6 0.25 0.31
 Coleoptera 3.6 6.9 3.111.5 3.9

Anti-herbivore defence compounds appear to have differential effects on insect populations. In early spring Pteridium appears extremely unpalatable and poisonous especially to non-specialist feeders such as locusts. However, it does not appear to be particularly poisonous or unpalatable in September, when ecdysone levels are high. Habitat heterogeneity (i.e. frond dimensions) appeared to be an important factor in structuring arthropod communities, and was more important than patch size in faunal communities (Rigby & Lawton 1981).

Insects as biological control agents

The use of insects as biological control agents has been proposed (Lawton et al. 1986; Lawton 1988, 1990). Two species of moth were selected, Conservula cinisigma and Panotima sp. near angularis from the Cape mountains of South Africa. This region was selected as having a climate as near to Britain as can be found in Africa and with the same subspecies of Pteridium. The chances of success were discussed by Lawton (1988), who considered that the introduction of moths would prove much less of a threat to native flora and fauna than either the continuing spread of Pteridium or the widescale aerial spraying of herbicide. However, there were substantive problems with their introduction. First, there would almost certainly have to be enablement legislation introduced, and thereafter if there were a successful introduction, there would be two further problems: (i) it would be irreversible, and (ii) the impact on Pteridium would be uncontrolled. Irreversibility means that once present it would be very difficult to control, and areas of Pteridium would be damaged without regard to whether it was regarded as a positive landscape feature or not. As Pteridium-infested land often needs substantial restoration aftercare after control, it is difficult to see how such restoration could be implemented given the potential patchy nature of Pteridium in Britain. It is possible, given that no bio-control agent has eliminated its target weed species, that a gradual decline in Pteridium vigour would be accompanied by colonization by other species. However, this process has yet to be demonstrated.


One species, the Azores Bullfinch (Pyrrhula murina Godman) eats Pteridium fronds when other food sources (plants, seeds and sporangia) are in short supply (Ramos 1994); on three transects (c. 500 m long) that were observed for two periods in 1991 and 1992, c. 5% of Pteridium fronds showed bite marks, compared to > 50% of Osmunda regalis fronds. It was suggested that these birds prefer O. regalis because it is acyanogenic, and that they tend to choose the lower (older) parts of Pteridium where toxic substances might be in lower concentrations.


Cattle, horses and sheep eat Pteridium but in general it is poisonous or harmful. Goats are less sensitive (Kingsbury 1964). Rabbits have been recorded as eating Pteridium rhizomes but have never been seen eating Pteridium fronds, even though the fronds can be found in the faeces, especially those of young rabbits (Bhadresa 1982). A wide range of syndromes in mammals can be induced by Pteridium (Smith et al. 2000):

  • 1Cyanogensesis, which probably merely restricts intake rather than causing serious harm (Section VI);
  • 2Induced thiamine deficiency, which causes a nervous condition especially in monogastric animals such as pigs and horses, but has been induced experimentally in sheep (Evans et al. 1975). This is caused by a Type 1 thiaminase (Evans 1976b). In cattle ingestion of an amount (green or dry) equal to the animal's weight over 3–4 months results in overt symptoms followed by death in 1–4 days;
  • 3Acute haemorrhagic disease, which occurs most frequently in weaned calves but is also known in older cattle and sheep (Moon & McKeand 1953);
  • 4Degeneration of the retina, e.g. bright blindness in sheep (Watson et al. 1965);
  • 5Enzootic haematurea, which covers a range of conditions resulting from tumours in the bladder. Common in sheep and cattle after prolonged intake of Pteridium;
  • 6Upper alimentary tract and other carcinomas (Pamukcu et al. 1980; Hirono 1986);
  • 7Induction of breast and lung neoplasia by ecdysones (El-Mofty et al. 1994);
  • 8Induction of hyperplastic nodules of the liver (Hirono et al. 1984).

The exact cause of the syndromes 3–5 remains unknown, but a range of chemical constituents of Pteridium (ptaquilosides, quercetin and shikimic acid) has been implicated (Evans 1976a; Smith et al. 2000). Chemicals in Pteridium, known to have carcinogenic, cytotoxic, mutagenic, tumorigenic and teratogenic activity, can induce apoptosis and symptoms were radiomimetic (Evans 1976a, 1986; Santos et al. 1992; Cross et al. 1996; Ngumuo & Jones 1996; Rzymowska et al. 1999; Simán et al. 2000). Ptaquiloside, for example, has been shown to have a DNA-binding capacity and cleaves DNA (Kigoshi et al. 1992). There may also be a synergism between secondary compounds and the incidence of papillomavirus in some cell transformations (Pennie & Campo 1992; Beniston et al. 2000). Pteridium spores have also been shown to cause carcinogenic effects when fed to mice, and DNA lesions in vivo and in vitro (Evans 1987; Villalobos-Salazar et al. 1995; Simán et al. 2000).

For humans the health significance of Pteridium poisoning is potentially considerable. In Japan, Brazil, and across northern America (especially the Appalachians), Pteridium is consumed as the developing frond, i.e. the crozier or ‘fiddlehead’ (Fig. 7f). At this developmental stage the quantities of toxin are at a maximum, and in Japan there is the highest incidence of gastric tumours in the world (Evans 1976a; Marlière et al. 2000). High levels of gastric cancers have been identified in other areas where there are large Pteridium infestations, e.g. Costa Rica and North Wales (Buckley 1989; Villalobos-Salazar et al. 1989), although transfer pathways and causal relationships have not been proven. Some effects can be transmitted experimentally via milk (Evans et al. 1972; Villalobos-Salazar et al. 1990; Alonso-Amelot et al. 2000). Transfer to humans through drinking water has also been investigated, and although water soluble extracts of Pteridium contain carcinogens, Galpin & Smith (1986) thought this is not a major cause for concern. However, more recently Rasmussen et al. (2003a,b) has detected ptaquiloside throughout the plant–soil system in Denmark (e.g. 108–3795 µg g−1 in fronds, 213–2145 µg g−1 in rhizomes, 200–8500 µg g−1 in soil solutions). This chemical is water soluble and easily leached from fronds and they estimated a flux of 260 mg m−2 transferred to soils. These fluxes were considered a potential risk for the contamination of groundwater supplies, especially on sandy soils.

(b) plant parasites: fungi

Fungi that live on Pteridium include, on the prothallus: Tilachlidium sp., Pythium sp. Pringsh. (Oomycota, Pythiaceae), Completoria complens Lohde (Zygomycota, Completoriaceae) and Coniothyrium sp. Corda (Ascomycota, Leptosphaeriaceae); on the adult: Corticium anceps (Bres. & Syd.) Gregor (Basidiomycota, Corticiaceae), Cryptomycina pteridis (Rebent.) Höhn (Ascomycota, inc. sed), leaf spot Sphaerella polypodii (Rabenh.) Fuckel (Ascomycota, Mycosphaerellaceae), leaf spot Ascochyta pteridis Bres. (Ascomycota, inc. sed), Melasmia imitans Peck (Ascomycota, Rhytismataceae), and Gloeosporium pteridii Karak and G. leptospermium Peck (both Ascomycota, Dermateaceae). Twenty-six species of fungi have been identified from dead, or dying, standing fronds (Table 17).

Table 17.  Fungi detected on bracken parts (Ellis & Ellis 1985); taxonomy revised from Index Fungorum ( i.c. = Incertae sedis
SpeciesTaxonomic groupPart of plant affected
Crocicreas cyathoideum (P. Crouan & H. Crouan) S.E. Carp.Ascomycota: HelotiaceaeDead fronds, especially in branch axils
Cryptomycina pteridis (Rebent.) Höhn.Ascomycota: i.c.Dead fronds
Dasyscyphus pteridialis GraddonAscomycota: HyaloscyphaceaeDead fronds
Dasyscyphus pteridis (Alb. & Schw.) MasseeAscomycota: HyaloscyphaceaeDead fronds, especially in branch axils
Hyaloscypha flaveola (Cooke) Nannf.Ascomycota: HyaloscyphaceaeLower surface, damp, dead fronds
Melittosporium pteridinum (Phill. & Bucknall) Sacc.Ascomycota: RhytismataceaeDead stems
Micropodia pteridina (Nyl.) Boud.Ascomycota: HelotiaceaeBlackened bases of old stems
Microscypha grisella (Rehm) H. Syd. & Syd.Ascomycota: HyaloscyphaceaeLower surface, damp, dead fronds
Mollisia pteridina (Nyl.) P. KarstenAscomycota: DermateaceaeDead stems
Mollisia pteridis (Alb. & Schwein.) GilletAscomycota: DermateaceaeBlackened bases of old stems
Pezizella chrysostigma (Fr.) Sacc.Ascomycota: HelotiaceaeDead petioles
Psilachnum pteridigenum GraddonAscomycota: HyaloscyphaceaeDead fronds
Diaporthopsis pantherina (Berk.) Wehm.Ascomycota: ValsaceaeDead petioles
Didymella lophospora (Sacc. & Speg.) Sacc.Ascomycota: Pleosporales i.c.Dead fronds
Didymella prominula (Speg.) Piroz. & Morgan JonesAscomycota: Pleosporales i.c.Dead stems and fronds
Leptopeltis litigiosa (Desm.) L. Holm & K. HolmAscomycota: LeptopeltidaceaeDead petioles
Leptopeltis pteridis (Mouton) Höhn.Ascomycota: LeptopeltidaceaeDead petioles and veins
Monographos fuckelii L. Holm & K.HolmAscomycota: DothideaceaeDead fronds and petioles
Mycosphaerella pteridis (Desm.) J. SchrötAscomycota: MycosphaerellaceaeDead fronds
Phomatospora endopteris (W. Phillips & Plowr.) Sacc.Ascomycota: Xylariales i.c.Dead fronds
Rhopographus filicinus (Fr.) Nitschke ex FuckelAscomycota: Pleosporales, i.c.Dead petioles
Scirrhia aspidiorum (Lib.) BubákAscomycota: DothideaceaeDead petioles
Ascochyta pteridis Bres.Ascomycota: i.c.Dead fronds
Camarographium stephensii (Berk. & Broome) BubákAscomycota: i.c.Dead petioles
Chalara pteridina Syd.Ascomycota: i.c.Dead petioles
Coniothyrium pteridis A.L. Sm.Ascomycota: LeptosphaeriaceaeDead stems and fronds

Pteridium prothalli are susceptible to many fungi but only two naturally occurring species are pathogenic to prothalli. During examination of Pteridium-infested areas in the west of Scotland these were Tylachidium sp. and Coniothyrium sp. The disease caused by Botrytis cinerea Pers. (Ascomycota, Sclerotiniaceae) is similar to the disease in higher plants (Hutchinson & Fahim 1958). Generally the behaviour under fungal attack is similar to that in higher plants.

Curl tip disease is associated with Ascochyta pteridis, Phoma aquiline Sacc. & Penz. (Ascomycota, inc. sed), Septoria sp. Sacc. (Ascomycota, Mycosphaerellaceae) and Stagonospora sp. (Sacc.) Sacc. (Ascomycota, inc. sed) (Angus 1958; Burge & Irvine 1985; Fisher 1996), and leaf roll is associated with Cryptomycina pteridis (Gabel et al. 1996). Other endophytes have been isolated Aureobasidium pullulans (de Bary) G. Arnaud (Ascomycota, Dothioraceae), Cylindrocarpon destructans (Zinssm.) Scholten (Ascomycota, Nectriaceae), Phoma sp., Ramichloridium schulzeri (Sacc.) de Hoog (Ascomycota, inc. sed), Stagonospora sp. and Sordaria fimicola (Roberge ex Desm.) Ces. & De Not. (Ascomycota, Sordariaceae; Petrini et al. 1993). Corticium anceps has also been shown to be parasitic in the field but only under the specialized conditions of high humidity.

Several attempts have been made to develop fungi as myco-herbicides; Phoma isolated from dead or dying fronds proved pathogenic under glasshouse conditions (Angus 1958), and both Phoma sp. and Ascochyta pteridis have been assessed (Burge & Irvine 1985; Burge et al. 1986; Womack et al. 1995). This assessment considered a range of different fungal isolates, and the development of emulsion-based CDA application systems. However, although reasonable infection rates were obtained in laboratory culture, these failed to be translated into effective control methods suitable for field application (M.N. Burge, pers. comm.).

In Transvaal, Pteridium is listed as a host of Pythium sp., a species that attacks tobacco, sugar cane, tomatoes and other cash crops. Corticium anceps does not cause malformation of the fronds, but since badly affected portions become very brittle and readily break off, diseased fronds commonly have an irregular and lopsided appearance.

In short, Pteridium, like most ferns, is remarkably resistant to disease.

X. History

(a) history and spread of pteridium

Pteridium is an ancient plant with records of Pteridium-like fossils in Oligocene deposits (Hungary) and Miocene deposits in England (Long & Fenton 1938) and Oregon (Graham 1963). There are single site records for Pteridium spores from the Ludhamian (Tiglian, the earliest interglacial) and Cromerian interglacials and more substantial records from all the substages of the Hoxnian (Holsteinian), accompanied at Gort in substage II by leaves with dichotomously branched veins, an incurved membrane and a lower surface covered in long multicellular hairs. Pteridium spores have been recorded in the three substages of the Ipswichian (Eemian), the last interglacial. Records in the late Weichselian (the Last Glacial) are sparse with spores recorded in scattered localities in southern and western Europe (Huntley & Birks 1983). In zone IV (Pre-Boreal, c. 9000 bp) Pteridium has been recorded at low frequency in Skye, the Caingorms, Tyrone and Hampshire. At Crane's Moor (Hampshire) spore frequencies in the following zone reached 5–10% of local tree pollen. The Flandrian (Holocene/Post glacial) expansion of site records did not occur until zone VI (Boreal, c. 8500 bp). At this point Pteridium was widespread throughout Spain, western France and the British Isles, with particularly high concentrations in the Skye region of north-west Scotland (Huntley & Birks 1983), and it almost reached its northern limit, at least in the British Isles. Spore frequency remained low throughout zones VI and VIIa (late Mesolithic) with Pteridium presumably confined mainly to the forest ground layer and natural clearances and margins, although with very high values in some western localities (Huntley & Birks 1983) suggesting open conditions, or only very thin forest cover. In zones VIIb and VIII, from the beginning of the Neolithic onwards (Godwin 1981), frequency increased greatly due to forest clearance. It is not surprising that the detailed analysis through prehistoric tree clearances regularly shows Pteridium making a quick response to forest opening (Turner 1964, 1965; Godw. Hist.; Rymer 1976). By 1000 bp the spore record is considerably reduced, especially in northern Scotland, reflecting climatic cooling and reduced spore production (Huntley & Birks 1983).

In undisturbed communities (similar to pre-Neolithic communities), Pteridium probably had a subordinate status in all communities except in open woodland, where the shade was moderate and the soil suitable. This implies that even when soil and climate are suitable for Pteridium, the natural woody vegetation excluded Pteridium by competition (McVean 1958). The basis for this deduction is partly the rarity or absence of Pteridium in vegetation cores from islands in West Highland freshwater lochs in comparison with the nearest similar vegetation on the mainland (McVean 1958). He drew the conclusion that the unmanaged vegetation (analogue for the Mesolithic) is unfavourable for Pteridium establishment from spores, rhizome spread or sporophyte survival, except in natural gaps. Where there were large open areas available on the mainland, there were opportunities for establishment, and both dispersal and vegetative spread. Thus, in the climax vegetation, Pteridium may be a subordinate component of the community or, where dominance occurs, this is in small patches. However, as yet there is no really convincing proof of this hypothesis.

During the Neolithic period and into the Bronze Age (VIIb, c. 4000 bp) there was an increase in the occurrence of Pteridium, associated with records of scrub development, vegetation clearance and charcoal use. By the Romano-British Iron age, Pteridium had expanded, and macrofossil evidence suggests that by this time it was being used for bedding (Turner 1965; Smith 1970; Page 1982; Brown 1999). This increase in Pteridium is therefore correlated with human activity, directly by felling, cutting, burning, cultivation and abandonment of land, and indirectly through grazing by domestic stock and changing the ‘balance’ between carnivores and small herbivores, including the later-introduced rabbit. Pteridium was important after clearance from 1600 bc. Later there is an indication of a reduction in Pteridium as if some measure of partial control was taking place (Walker 1966). Intense clearance began around 1300 ad (Moore & Chater 1969). This general explanation for the rise in Pteridium in the northern hemisphere has also been confirmed for P. esculentum after the arrival of Maori settlers in New Zealand at c. 800 bp (McGlone 2001; McGlone et al. 2005). More recent increases in Pteridium have been ascribed to land-use change. In County Mayo, Ireland, expansion occurred after a period of woodland clearance in the late 18th century, cropping with oats and potatoes peaking in the mid-1800s, followed by a decline and a change to marginal pasturing with subsequent Pteridium increase (Little & Collins 1995).

Current observations, archaeology and documentary records and the palynological record of the British flora and vegetation collectively show Pteridium as a ‘camp-follower of man’; it expanded as a result of the removal of factors limiting it, or the creation of opportunities for expansion. Man learned to use Pteridium for a range of purposes and its harvest must have at least restricted its spread and at most reduced its cover and intensity. However, as man has stopped using Pteridium as a resource in recent times, it has become regarded as a weed. When the pollen records of a range of unbiased samples were compared, Pakeman et al. (2000) concluded that the current abundance of Pteridium was less than, or at worst, equivalent to maximum historical records. Indeed, recent surveys suggest a slight deline in abundance (Pakeman et al. 1995, 2000).

(b) uses of pteridium

Pteridium has been used in a variety of ways by many cultures, and this has varied in time (Rymer 1976). The rhizomes have been used as either food or as a source of bread in aboriginal Australian, British, French, Japanese, Lapp, Roman and Siberian cultures (Rymer 1976; Veitch 1990). In Japan, the young croziers are still eaten after boiling (which has been shown to reduce the levels of toxins; Hirono et al. 1972, 1973) and bundles are sold in shops in spring. As a starchy substance, the rhizomes have been used for the preparation of glues and the brewing of beer. The rhizomes contain about 46% starch: they froth with water and are used as soap in country districts of France. Pteridium has also been used to produce strong yellow dye, and on the Isle of Man it was boiled with linen to bleach it. The fronds have also been used for packing fruit in baskets and for protection in gardens against the winter frost.

In Scotland it is used as thatch for houses that could last for 20–30 years, cordage for securing heather, stacks and sheds and as fuel for domestic heating. Pteridium has also been employed routinely for animal bedding and, indeed, human bedding since Roman and Viking times. Pteridium has also been used for animal feed; Aitken (1888) reports cut Pteridium heaped into stacks like beehives, and when this material was fed to Highland cattle as silage, the cattle suffered no ill effects. When Pteridium is cut early a silage of reasonable quality can be produced, but it is not generally attractive to stock and its digestibility is low (Watson & Smith 1956). The dry matter was close to hay in starch equivalent, and when cut late it was little better in nutritional value than straw. In Wales the young green fronds were used as pig food, and dried fronds were chopped up with straw and hay and given to horses.

There are many examples of the use of Pteridium as a fertilizer. At Stonefield in 1779 the rent included ‘16 cartloads of pulled fern deliverable at the tenants’ expense to the mansion house in Fernoch for the buildings of the said lands of Fernoch’ (L. Rymer, pers. comm.). It is also used as manure by cutting, wetting and trampling down and then mixing with calcium-rich materials and soil (A.S.W.). Recently Pteridium has been tested as a mulch, green manure and as a source of potash in organic agriculture and potato production (Taylor & Thomson 1998; Donnelly 2003), and for compost in modern horticulture (Pitman 1995; Davies 2004). Dead fronds of ‘Pteridium aquilinum var. esculentum’ provided a useful mulch for Sinapsis alba (white mustard), Trifolium repens (white clover), Lolium perenne (perennial ryegrass), Eucalyptus fastigiata (brown barrel) but not Leptospermum polygalifolium (yellow tea tree) (Taylor & Thomson 1998).

Pteridium has also been used industrially: it has been processed into a board, similar to beaverboard, resembling Bakelite; and as basic ash, rich in potash, that was used for soap and glass-making. Pteridium was also used as faggots for fuel in baking ovens; it has been used to fire brick and lime kilns. For this use, fronds were cut at the end of June when the levels of potassium were at a maximum; it is half-dried, then burned slowly in a pit. It has also been assessed for its suitability as a crop to produce energy through combustion or the production of biogas (Lawson et al. 1986).

Pteridium has been used as a medicine. It has been utilized as an antihelminthic, but according to Langham (1579), there are at least 21 uses including ‘Burnings, Cattle galled, Festers, Gnats, Horsesicke, Kanker, Miltpaine, Mother suffocat, Nosebleeding, Purgation, Sinewes griefes, Skinne off, Sores, Wormes, Wounds, Makes women barren’. It has also been used as a cure for rickets, and as an aphrodisiac (Cameron 1900; Rymer 1976). The fronds have a characteristic smell and are repellent to insects, serving well for wrapping up fruits and vegetables. It is also a plant of fokelore; ‘fern seed’ apparently can confer the power of invisibility, and burning bracken it is reputed ‘doth draw downe the rain’ (Rymer 1976).

In the past, Pteridium was such an important resource that its harvesting was once a right, and an appreciated privilege, protected in some places through byelaws to prevent over-exploitation. Where this occurred the Pteridium harvesting was restricted to ensure that the frond harvest was taken at the end of the growing season, at a time (Stage 5, Fig. 4) after the carbohydrates and nutrients had been re-translocated to the rhizomes, and there was minimal impact on the frond production in the following year. Thus, at Lakenheath Warren Pteridium was not to be cut until after the 29th of August each year (Crompton & Sheail 1975). In this way there would be a minimal effect on future productivity and hence a sustainable crop was ensured. Thus, in the past, Pteridium was very useful to man though now under the modern economic regime it is a pest to be destroyed or reduced.

The classification of Pteridium as a weed is a relatively recent phenomenon. Its abundance and distribution are generally regarded as being relatively stable until the last 200 years when a combination of changing factors (land management practices) reduced control and allowed it to expand. Pteridium is not recorded as a weed until complaints about its spread were first noted in the early 19th century and its control was attempted by flooding with spring water (McTurk 1837; Murray 1837). Poel (1951) more recently tested this method in two small-scale studies; the first killed most fronds within 4 months, but there was no effect on the rhizomes or the adjacent Callunetum. The second compared two 1-m2 quadrats; irrigation reduced frond numbers from 50 m−2 to zero in less than 16 months; unfortunately, effects on rhizomes were not reported.

Since then, there have been increased generalizations that Pteridium was spreading, caused by, inter alia: a change from cattle to sheep grazing in Wales and the highlands, a change from heavy wether sheep to lighter ewes grazing in North Wales, muirburn, severe winters, acid rain, elevated nitrogen loads, and even myxomatosis which reduced the rabbit population. The change in grazing systems from heavier to lighter animals with a lower trampling effect and hence reduced damage to developing fronds reduced the limitation on Pteridium performance (Pakeman & Marrs 1992). There is also a possibility that Pteridium has been severely affected by occasional very hard winters (e.g. those in the 10th century and 1940, Watt 1940), and a reduced frequency of these has allowed Pteridium to expand.

In general terms, Pteridium has ceased to be a useful plant, and has become in Britain a weed to be checked or destroyed.

XI. Conservation and management

(a) conservation

Pteridium is generally regarded as a weed and as having a low conservation status, or at least lower than the communities that it replaces. However, there are some Pteridium stands where there is a high conservation value (reviewed in Pakeman & Marrs 1992). It also provides an attractive and valued vista in autumn, providing a golden brown hue to the landscape.

Plant species

In some instances, Pteridium communities can harbour local rarities; these are usually woodland species, probably relicts from previous woodland communities (Section III).


In some areas, Pteridium stands have an understorey that provides a refuge for food plants (Viola sp. and Melampyrum pratense) of butterfly species of high conservation interest such as the high brown fritillary (Argynnis adippe Denis & Schiffermüller), heath fritillary (Mellicta athalia Rothemburg), pearl-bordered fritillary (Bolorio euphrosyne L.) and small pearl-bordered fritillary (Bolorio selene Denis & Schiffermüller) (Warren & Oates 1995). The Pteridium canopy appears to form a pseudo-woodland canopy that in conjunction with the food plants contribute to the survival of the butterflies. However, where the Pteridium canopy becomes too dense, the litter appears to have adverse effects on butterflies. Conservation management of the Pteridium at such sites must tread a delicate balance. Management that contributes to a heterogeneous canopy structure appears best and grazing, especially when large herbivores are present (deer, cattle and horses), appears ideal.


Pteridium stands in some areas hold populations of adders (Vipera berus L.) and common lizards (Lacerta vivipara Jacquin), and in some areas the very localized sand lizard (Lacerta agilis L.) and smooth snake (Coronella austriaca Laurenti) (STOG 1988). Reptiles often use bracken litter for hibernation but also require bare areas or areas of short vegetation for egg laying and/or basking. Thus a mosaic of habitats is needed to maintain these species, and where Pteridium invades and becomes dominant, this can be problematic for these populations.


Pteridium stands generally have little ornithological importance. Ratcliffe (1977) listed 15 species breeding in Pteridium compared with 25 and 33 on submontane acidic grassland and Calluna-dominated moorland, respectively; some species are, or can be, positively associated with Pteridium stands. These include: nightjar (Caprimulgus europaeus L.), ring ouzel (Turdus torquatus L.), tree pipit (Carduelis flavirostis L.), twite (Carduelis flavirostis L.), whinchat (Saxicola rubetra L.), warblers (Phylloscopus sp.) especially willow warblers (P. trochilus L.), and some raptors. Others appear negatively associated: curlew (Numenius arquata L.), golden plover (Pluvialis apricaria L.), merlin (Falco columbarius L.), red grouse (Lagopus lagopus L.), greenshank (Tringa nebularia Gunnerus), short-eared owl (Asio flammeus Pontoppidan) and hen harrier (Circus cyaneus L.). Data are conflicting as twite and merlin showed strong affinities for nesting in the southern Pennines (Haworth & Thompson 1990) and a negative relationship elsewhere (Ratcliffe 1977). Although Pteridium stands were listed as the third most important habitat for nightjar (Gribble 1983), Pteridium control was included as part of the management plan to increase nightjar populations at Minsmere (Burgess et al. 1990).


Pteridium stands can provide cover for fox (Vulpes vulpes L.) earths, rabbit burrows and badger (Meles meles L.) sets, and frond material may be used for bedding by badgers and fallow (Dama dama L.), roe (Capreolus capreolus L.) and red (Cervus elaphus L.) deer. Pteridium provides food and cover for the Skomer vole (Clethrionomys glareolus skomerensis Barrett-Hamilton) (Pakeman & Marrs 1992).

Little is known about small mammal activity, field (Microtus agrestis L.) and bank (Clethrionomys glareolus Schreber) voles, woodmice (Apodemus sylvaticus L.) and the insectivorous common (Sorex araneus L.) and pygmy (S. minutus L.) shrew and hedgehog (Erinaceus europaeus L.) have been found in Pteridium stands.

Conservation management

As Pteridium communities tend to be mid- to late-successional (Marrs et al. 2000), it is likely that they will provide different conservation opportunities, and cause different problems in different situations and at different times. This will inevitably require different conservation management options to be implemented when appropriate.

(b) pteridium control

Pteridium is very difficult to eradicate and usually at least a two-stage control process is required. Where Pteridium is dense, there needs to be an initial control stage, and thereafter there will almost certainly need to be a second phase of follow-up control, possibly integrated with a restoration phase to re-establish semi-natural vegetation. Once a suitable vegetation type has been established, a maintenance phase is needed to ensure that the required vegetation is maintained and the Pteridium is kept at a low level.

Where Pteridium is present at low levels at the outset, control can be less intensive but management of the surrounding vegetation is needed to keep it in good condition and prevent expansion. Where a Pteridium front is invading other communities, expansion can be kept in check by cutting, herbicide use (Pakeman et al. 2002) or by the development of competitive vegetation (Watt 1955; Ninnes 1995).

Essentially, there are three main strategies, which are used to control bracken. These are described below.

(1) Mechanical control

Mechanical control can range from severe disturbance (e.g. ploughing and cultivation) through cutting and bruising/rolling to less severe treatments such as stock treading.

Where land is not too steep or rocky a good reduction in bracken cover can be achieved by ploughing; the aim is to cut the bracken rhizomes and expose them to frost damage (Snow & Marrs 1997). Some regeneration will occur, so a follow-up program using another method must be employed. Clearly, this method is not recommended for sites with a valuable ground flora or with archaeological remains.

Cutting is a common form of mechanical control. Here, fronds are cut during early summer, before and up to the point of maximum frond expansion. The aim is to ensure a maximum withdrawal of carbohydrates and nutrients from the rhizome reserves (Hunter 1953; Williams & Foley 1976). When this strategy is used it is advisable to cut the fronds before the new assimilates start being translocated from the fronds to the rhizomes in large amounts (late July/early August in Britain) (Williams & Foley 1976). Cutting can be carried out one, two or three times annually and needs to be repeated for at least three years (Braid 1959; Williams 1980). The positive advantage of cutting is that it helps break up deep Pteridium litter and helps natural regeneration (Marrs & Lowday 1992; Marrs et al. 1998c).

Crushing/bruising is a variant of cutting; it may be less effective than cutting as it does less damage to the litter layer, but is especially suitable for difficult terrain, which might damage a cutter. Like cutting, this method is unsuitable for eradicating bracken and follow-up by other means is necessary.

Damage can also be inflicted by stock treading of the fronds. It can be manipulated either by fencing or by siting winter-feed appropriately. The livestock (i) damage the rhizome buds and developing fronds which are close to the surface or just emerging and (ii) disturb and break-up the litter, encouraging frost penetration to the rhizomes and the regeneration of vegetation.

Burning of bracken litter can also be used to ease cultivation and seeding success. Burning of dead litter without follow-up is of no benefit, as it constitutes an unnecessary fire risk, may increase frond production and has adverse, although temporary, effects on landscape value.

(2)Herbicidal control

Herbicide action is unlikely to have a significant effect on the amount of rhizome carbohydrate reserves, so herbicides, which attack frond buds on the rhizome, are most successful. Only asulam (Asulox) and glyphosate (many formulations) have label recommendations in the UK.

Asulam is a selective herbicide and can be sprayed in all densities of bracken because it has little permanent effect on underlying vegetation, except other ferns and some bryophytes (Rowntree 2004; Rowntree & Sheffield 2005) and it can be applied by air (> 5000 ha year−1 between 1990 and 1998; Pakeman et al. 2000). Glyphosate is a non-selective herbicide and should be sprayed only in areas of deep litter bracken with little underlying vegetation: it will kill any grass, Calluna or other plants present.

Both chemicals have to be applied after full frond expansion has occurred but before any dieback of the tips to ensure maximum absorption and translocation into the below-ground rhizome system. This period occurs between mid-July and late September depending on altitude, locality and season. Both chemicals can give over a 95% reduction in fronds in the year following treatment. An advantage of glyphosate is that it produces visible symptoms soon after application, allowing treatment of missed strips in the same season. In all cases, the spot treatment of missed areas in the same season or in the following year, and the treatment of regenerating areas, will considerably increase the period of effective control.

Asulam [methyl (4-aminobenzenesulphonyl) carbamate] is the most widely used herbicide for bracken control; it is translocated into the rhizome and accumulates in both active and dormant buds, where it effects a lethal action (Veerasekaran et al. 1976, 1977a,b, 1978). Asulam frequently produces a very good reduction in fronds in the year after spraying, but there is often rapid frond recovery unless other treatments are applied in subsequent years (Robinson 1986, 2000; Lowday & Marrs 1992a). It must be stressed that bracken will regenerate quickly after herbicide application (Marrs et al. 1998a), so further treatments will be necessary to maintain control or progress to eradication. Repeat spraying on a 5- to 7-year cycle (i.e. when recovery may have produced a complete canopy) is not effective at controlling bracken on a long-term basis. For eradication of bracken, an annual follow-up spot treatment of asulam has been recommended, and this should be applied to all regrowing fronds without respite until no further fronds are produced (Robinson 2000).

(3) Inhibition by other vegetation

Usually where dense bracken is to be controlled, managers want to remove the bracken and replace it with some other vegetation. There is some evidence that this approach can work in some situations (Watt 1955; Lowday & Marrs 1992a; Ninnes 1995). Tree planting within Pteridium stands is an obvious way of manipulating the succession which will inevitably reduce Pteridium cover (Marrs et al. 2000). However, Pteridium can maintain itself as a component of the field layer and it will expand after felling or coppice (Harmer et al. 2005).


This account has had a long gestation. The original draft was compiled by the late Dr A.S. (Sandy) Watt FRS, who found he had too much to say for such an important species within the original limit of 12 pages. He was encouraged by Professor C.D. Pigott to expand this to a ‘sort of mini monograph’. With failing health this became an impossible task and, in 1981, A.S.W. passed the original hand-written manuscript to R.H.M. in the hope of a major and rapid revision. R.H.M. has pruned these handwritten manuscripts (which are archived at the University of Liverpool) to about 10% of their original length, and has considerably updated the final version. R.H.M. has attempted to maintain the flavour of the original; however, all errors, omissions and changes in emphasis are his. Professor Sir John Lawton, Dr S. Isaacs and Dr D.J. Thompson read and corrected parts of an early draft, and the final version owes a great deal to assistance provided by successive editors, the late Professor A.J. Willis and Dr A.J. Davy, and referees including Dr M.C.F. Proctor, Professor R.J. Pakeman, Dr S.E. Sheffield and D.T. Streeter. Finally, R.H.M. owes an enormous gratitude to Professor J.A. Thomson from the National Herbarium of New South Wales, Sydney, who has provided wise counsel, early access to his taxonomic revisions of Pteridium, and correction of mistakes.

The editing phase has been supported financially by both the University of Liverpool, and the Royal Botanical and Horticultural Society of Manchester and the Northern Counties. R.H.M. thanks Dr Catherine Collins and Ms Emma Cox for substantive editorial assistance, Ms K. Lancaster and Ms S. Yee for graphics production, his three colleagues Dr Hugh McAllister, Dr Mike Le Duc and Professor Robin Pakeman for long-term support on Pteridium work, and A.S.W.'s family for encouragement.

The literature on Pteridium is voluminous and the final selection is to some extent a personal one; hence an extended interactive literature data base, encompassing A.S.W.'s and R.H.M.'s information, is available at Most references are available in pdf format.