Woodlands

In the 1980s and 1990s the Forestry Commission and Greenpeace recorded symptoms on trees, principally conifers but including beech, all over Great Britain.⁷ In some years at least two-fifths of the youngish beeches in this country fell short of the Central European standards for a healthy beech-tree. The shortfall varied from year to year in no very obvious pattern, nor was it related to the geography of any form of pollution. No evidence emerged that this state of affairs was really abnormal – that British beeches had measured up to those standards in the past.

Tansley’s book on British vegetation gives the impression that woodland soils more acid than pH 4.0 were rare in the 1930s. They are not rare now: in the 1970s I measured pH of between 3.0 and 4.0 in about one-fifth of the woodland soil samples that I examined from eastern England. This is a surprisingly large change to have occurred in only 40 years, given that at least twice as much acid deposition must have occurred in the century before Tansley. Has acid rain acidified woodland soils? Or were methods of measuring pH in Tansley’s time less sensitive to low pH?

Since 1971, woodland soils in the rest of Britain have been getting less acid. The Bunce–Kirby comparison reports a mean pH of 4.98 from 1,648 plots in 1971, and 5.31 from the same plots in 2001. The methods on the two occasions were comparable and the results appear to be real. The more acidic soils were more affected: some 12 per cent of soils had pH below 4.0 in 1971, but only 2 per cent in 2001.⁸ Is this an effect of contamination by lime as an agricultural fertiliser (cf.)? The data did not include any from East Anglia and Essex, where my 1970s measurements need repeating.

From the seventeenth century onwards coal was the main urban and industrial fuel, and when railways came was the main transport fuel too. In the 1950s and 1960s, encouraged by the Clean Air Act, coal was partly replaced for heating by heavy oil, which was less smoky, but worse for sulphur dioxide. (Electricity, too, was dirty because of the coal and oil wastefully burnt to generate it.) From the 1970s onwards, oil was replaced by natural gas, which contains almost no sulphur.

After 1975, lichenologists noticed a dramatic change in acid-sensitive lichens in London. Pollution suddenly declined with the ending of heavy oil for heating, of coal-burning railways, and of Bankside and Battersea power stations. Species not seen for 200 years returned even to central London.⁹ From 1974 to 1999 The Lichenologist published annual bibliographies of articles on lichens and air pollution, documenting a similar recovery in cities all over the world.

When I first went to Australia my hosts took me into the bush and explained that Australia was a different planet, in which most trees were eucalypts, and fire and termites did the recycling job done by fungi and worms in Europe. Another unfamiliar feature, they said, was that Australia had droughts lasting years at a time: the leafage of eucalypts died back, economising moisture, and grew again when the rains returned.

But was this so unfamiliar? In the 1940s many hedgerow oaks in Norfolk had towering dead branches, which fascinated me. In the 1960s these ‘dying’ oaks were held up as an example of the ill effects of ploughing around them, toxic sprays, roadworks, ‘old age’ (but few were more than 200 years old) etc. In the 1990s they were still ‘dying’: similar trees were held up as an example of the polluting effects of Gatwick Airport. Those in Norfolk are mostly still there: over the years some of the dead boughs have rotted away, but others remain, and the trees have grown new leafage at a lower level.

Dead boughs on oaks last for about a century (Fig. 131). The bark rots and disappears, then the sapwood, leaving a sharp-edged core of heartwood. On oaks in Madingley Wood, photographed by D.E. Coombe in 1951, some boughs then already reduced to heartwood are still there now.¹⁰ I suspect that the ‘dying’ oaks were survivors of the great droughts of 1911 and 1921, aggravated by a plague of defoliating caterpillars, by newly arrived oak mildew, and perhaps by frost and honey-fungus. Silviculturalists all over Europe debated whether oak had any future;¹¹ their French colleagues waxed eloquent on the grande misère du chêne. Oaks in Sussex were stag-headed long before Gatwick Airport was thought of.

Portraits of historic oaks, such as Queen Elizabeth’s Oak at Huntingfield (Suffolk), show that they have passed through phases of dieback in the past: a ‘high top bald with drie Antiquitie’, as Shakespeare put it. In Sherwood Forest, successive waves of dieback, probably on the same trees, have been remarked on since the seventeenth century.

The 1921 drought was combined with other factors that have not recurred. Subsequent droughts have not produced oak dieback on the same scale. However, the ‘oak decline’, which attracted attention and concern in the 1990s, is likely to be related to a cluster of hot dry summers.

The drought of 1975–6, though it failed to reactivate oak dieback in Sherwood, was followed by a dramatic dieback of non-woodland ashes, especially in the east Midlands; by the 1980s there were whole landscapes of half-dead ash-trees. But ash does not remain stag-headed for long: by now the dead boughs have disappeared, and the trees have grown new crowns.

In 1989 the Conservators of Epping Forest asked me to investigate why the great pollard beeches of the Forest were dying. The matter attracted much attention: ‘Epping Forest in Danger’; ‘up to half’ the trees were said to be affected; I was told that a hapless councillor had been voted out of office for failing to keep them alive.

In Epping Forest there were several hundred manifestly unhealthy beeches and some dead ones. It was argued that the beeches of the Forest displayed an advanced state of Waldsterben symptoms; that this had the same significance and cause as in Central Europe; and that this was a recent and abnormal state of affairs.

Had the Continental parallel been correctly applied? Are the symptoms pathological at all, or are they part of the normal behaviour of beeches, which are not immortal? If pathological, are they specific to the Waldsterben syndrome, or can other adverse influences bring them about? Are they more severe in parts of the Forest particularly exposed to pollution? Have they increased recently?¹²

Epping Forest, anciently Waltham Forest, is a huge, uncompartmented wood-pasture common (Fig. 183). It sits on a flat-topped ridge of Claygate and Bagshot Beds overlying London Clay. The trees are mainly hornbeam on lower ground, oak in between, and beech on the plateau. Beech is dominant over an area of about 3 miles by 1¼ (5 × 2 kilometres). Soils under beech are silty and seasonally wet, very unlike the well-drained sites on which beech typically occurs. They are acid: most pH measurements range from 3.7 to 4.2, typical for woodland soils in south Essex.

In prehistory Epping Forest was strongly dominated by lime, as shown by a pollen record from a bog in the present beech zone. The beech zone contains two Iron Age hillforts, connected with major settlements probably outside the present Forest. In Anglo-Saxon times the wood-pasture probably reached its medieval and early-modern form. Lime declined and then disappeared, replaced by oak and then beech.¹⁴ Domesday Book shows the future Forest as the biggest concentration of woodland in Essex. By the fourteenth century it had reached almost exactly its present size.

Waltham was declared a Forest in the early twelfth century. The physical Forest lay between the king’s palace at Havering and his royal abbey of Waltham. The land was owned by a score or more of private or institutional lords of manors, and used by several hundred commoners who had rights of pasturage and woodcutting. To these other uses the king added perhaps 200 fallow deer and a few red deer. In the mid-thirteenth century he consumed, on average, about 20 fallow and four red deer annually from Waltham Forest.

Epping Forest produced wood, but not much timber. Since it was uncompartmental and the commoners’ animals could not be excluded, the trees were pollarded. Chapman & André’s map of 1772–4 shows most of the Forest as wood-pasture, carefully differentiated from woodland (Fig. 183). There were a number of plains, areas of heather, grassland and bog, mainly on low or very ill-drained sites, grazed by deer and the commoners’ cattle, sheep, pigs and illicit goats. It was crossed by one main road, the present A121, bordered by trenches (narrow clearings, p.171f) to give travellers a sense of security from highwaymen.

The Forest and its institutions changed little for seven centuries. In 1543 Henry VIII made a short-lived park on land whose ownership he had stolen from Waltham Abbey. He left behind a standing, an observation tower for ceremonial hunts, now miscalled Queen Elizabeth’s Hunting Lodge. (The great oaks from which it is built probably came from elsewhere.)

Decline came in the nineteenth century from external causes. In 1830–4 the local turnpike trust split the Forest along its length by a new road, the present a104 (a11), which even then showed scant respect for the Forest’s amenities. In 1851 the sister Forest of Hainault, where the Crown owned most of the land, was privatised and most of it grubbed out and made into poor-quality farmland. Epping Forest was saved by its fragmented ownership, but individual landowners began to encroach upon it. Its growing significance as a place of public recreation led to the formation of the Commons, Open Spaces and Footpaths Preservation Society. The City of London Corporation, whose purchase of a property carrying grazing rights gave them a legal standing, were persuaded to intervene. The result was the Epping Forest Act of 1878, which abolished the rights of the Crown and the landowners and transferred the freehold to the City of London.

This was the first big victory in Europe of the modern conservation movement. However, it resulted in a transformation of the Forest and a series of problems that still continue. The early Conservators had little sense of history or ecology: the science of ecology itself had hardly begun. They were strongly prejudiced against pollards, and stopped the commoners from cutting them. The beeches grew up crowded and dense, suppressing other trees and ground vegetation – except holly, which has infilled wherever there is a space. The plains were not grazed enough, and the heather was overgrown with oak and birch. Although since 1980 the Conservators have recognised the losses resulting from these changes, they had gone too far to be easily reversed.

Beech is the only tree in the Forest commonly showing signs of damage. The original Waldsterben of Central Europe principally affects spruce and fir (Abies); damage to beech there is less conspicuous. There have never been any significant conifers in Epping Forest; those planted around its edges have not particularly suffered.

Dead or manifestly unhealthy beeches occur in groups (Fig. 132). A typical group centres on a cluster of dead trees or their remains; these grade through a zone of trees with more or less severe symptoms, to the ‘normal’ beeches of the surroundings. All the groups together amount to at most 5 per cent of the beech in the Forest. Hornbeams and oaks within a beech group are much less affected than beech; holly and birch flourish through lessening of shade. Group damage is not closely related to the age of the trees. It may be commoner in maiden beeches, dating from around the decline of pollarding, but affects pollards also.

Group damage is mainly in the northern, beech-dominated, half of the Forest. The sites tend to be just off the central plateau. They are usually on or near spring-lines, marked by patches of rushes or the wetland grass Molinia cærulea. They are associated not with vehicle roads, but with constructed horse-riding tracks.

Group damage goes back at least to the 1960s. On my first visit, in 1974, I noticed the Broadstrood group, which was evidently not recent. This group is no longer active, and the gap has been invaded by birch. Near High Beach there was a boggy area with ‘dying’ hornbeams and oaks as well as beeches; this too has not markedly got worse. The Hill Wood group is said to have begun in the late 1960s; some then attributed it to the effects of a World War II explosion. This group was aggravated by the 1975–6 drought, but later showed recovery. In the hot summer of 1989 some beeches on the edges of groups withered, but there was no general extension of the groups; indeed, there was a good deal of recovery, as marginal trees produced new shoots from lower in the crown.

Epping Forest was not much affected by the storms of 1987 and 1990. Trees bordering groups have sometimes been uprooted or broken by windblow. This is to be expected – the death of one tree exposes its neighbours – but is not an important factor in extending the groups. Beeches have often died after uprooting, which is unusual and perhaps indicates pre-existing root trouble.

Of the five groups that began before 1970, three had stopped extending by 1989. At least eight others date from before 1980. Some of the big groups extended further by 1992. There was a general slow extension during the 1990s.

Symptoms of reduced ‘vitality’ on beech were published by D. Lonsdale,¹⁵ following the practice of A. Roloff in Germany. They are:

crown thinness – the amount of sky that can be seen through the canopy;
branching pattern – in vigorous trees long-shoots predominate, whereas in less vigorous trees short-shoots take over;
cluster-shoot formation – chains of short-shoots arising from other short-shoots;
whether the leafage becomes thinner at the top of the tree;
whether there are dead twigs or branches at the outside of the canopy;
leaf rolling;
fall of leaves while still green; and
premature autumn colours.

Most of these ‘symptoms’ were present, but not always in the same tree. They were more marked in the damaged groups than in the general Forest. But their application is not straightforward. The list was intended mainly for use with younger trees; in trees of the ages usual in Epping some reduction of vitality may be normal. Epping Forest trees are wild, not mass-produced in a nursery, and some of the characteristics in which they normally vary overlap with those regarded as measures of vitality. Crown density varies from tree to tree because some trees have a strongly directional distribution of the foliage.

If vehicle fumes damage trees directly, damage should be more intense in places more exposed to fumes. In practice, none of the damaged groups directly borders a main road. The busy Epping New Road is bordered for miles by beeches, hornbeams and oaks no different in condition from those in the rest of the Forest.

Two places are specially exposed to fumes. Woodredon Hill is a busy, narrow main road climbing a steep hill. Trees come right up to the road. Beech is super healthy, except for slight advancing of autumn colours. Careful search detected slight abnormalities on oak and hornbeam. In the south of the Forest the North Circular Road crosses a narrow part of the Forest in a cutting, and intersects the Epping New Road. There are slight but clear symptoms on beech, oak and hornbeam – which could as easily be due to the digging of the cutting as to traffic fumes.

Group killing can hardly be attributed to the direct effect of traffic pollution. Lack of association with roads is the more significant in that groups near roads are more likely to be discovered than those away from them. Several groups, however, adjoin horsepaths.

Outside areas of group damage and of special exposure to pollution, most of the beech symptoms occur, but are never severe, and rarely more than two symptoms per tree. Chlorosis is probably the most frequent. In oaks, a common symptom in 1989 was patches of premature leaf-yellowing.

Although older, the beeches of the Forest at large in 1989 were no worse in terms of vitality, and probably better, than the average of those in Forestry Commission and other surveys of Britain in earlier years. There was no sign that the symptoms were getting worse or that the trees are suffering actual harm.

If general air pollution were the cause, trees on the western side of the Forest ridge should be more exposed and more affected than those on the east side. There was no sign that this was so.

For most of its history the Forest was very different from today. Beech was less strongly dominant. The trees were much smaller, pollarded on a cycle of about 13 years. Although in places they were crowded, the Forest as a whole was less wooded and more savanna-like. The present Sunshine Plain – a narrow, boggy, open area – is almost the only survivor of the plains (mapped by the Ordnance Survey in 1881) that wandered irregularly across the beech zone. These were permanent: one is mentioned in a charter boundary of 1062.

After 1878 the Conservators suppressed pollarding and allowed grazing to decline. Trees, especially beech, were allowed to grow up very much bigger than before. The trees that overgrew the plains are, for the most part, oak and birch; beech is uncommon on former plains.

Pollarding often ended well before 1878. Views of the Forest, shortly before then, depict pollards of many more than the normal years’ growth.¹⁶ The pollard beeches of the Forest today have grown up for between 130 and 190 years; their bollings (permanent bases) vary from over 400 years old down to about 130. The oldest maiden (unpollarded) beeches date from about 1878.

The plains are not there by chance. Claygate and Bagshot Beds vary from clay to gravel, producing spring-lines and waterlogged patches. Here the growth of most trees is difficult and was easily prevented by the animals which grazed the Forest. Most damage groups are close to plains and on similar soils.

Beech is not abundant in the Forest because the environment favours it. The Forest is acidic and wet, and at best marginally suitable. Beech was favoured by the rise and fall of the wood-pasture ecosystem, which allowed it to predominate over competing species. It is passing through a phase of abundance that may not last for ever: in Writtle Forest it died out as a native.

Epping Forest is the classic locality for the decline of lichens in relation to pollution. J.M. Crombie, in a pioneering article, noticed that many of the lichens collected by Edward Forster in c.1790, which included even Lobaria pulmonaria, had disappeared when he first knew the Forest in the 1860s; one-quarter of them had disappeared by 1885. Further drastic losses were found by lichen surveys in the 1910s and c.1970.¹⁷

Sulphur dioxide output in London was increasing and shifting eastward to locations nearer the Forest, especially Liverpool Street Station, which opened in the 1840s. Houses – all burning coal – began to invest the south of the Forest in the 1870s, and by the 1930s had spread northwards. All but the more resistant lichen species were progressively lost.

Table 19 illustrates this progression. The lichens present in the Forest at each date have been classified on a 10-point scale of sensitivity to sulphur dioxide; from this it is possible to estimate the mean acidity of the atmosphere in winter, in terms of the sulphur dioxide concentration.

Although by 1970 acid rain in Epping Forest was severe, there are two worse degrees on the Hawksworth–Rose scale. As trees in Sheffield, the Welsh coalfield valleys and London show, beech can survive and even flourish in a greater intensity of sulphur pollution than Epping Forest had at its worst.

In the absence of fuller data, Paul Moxey kindly gave me some records by Professor Hawksworth in 1989. These, though incomplete, show a change for the better (as in many other places): the Forest had regained some lichens not seen for more than 70 years.

Paul Moxey measured the acidity of rain at High Beach, near the middle of Epping Forest, day by day from 1983 to 1989 (Table 20, by his kind permission).

Rain in Epping Forest is more often acidified (pH < 5.6) than not, but the pH varies according to which dust or gas the rain has picked up on a particular day. Acidity was lessening; very acid rain (pH < 3.6) happened only sporadically. De-acidified rain (pH > 5.6) was getting more frequent in the late 1980s. The change was too recent to have been responsible for bringing back the lichens: rain in Epping Forest would have been more acid still in the middle of the century.

Hyde Park – very polluted: If pollution is bad for beeches in Epping Forest, it should have been impossible in the middle of London. Hyde Park and Kensington Gardens have a history of very severe acid rain, as shown by the erosion of Victorian statuary; they are surrounded and intersected by main roads. In 1989 there were almost no obvious lichens, even Lecanora conizæoides being confined to the relatively alkaline bark of ash trees.

In 1986 Dr Roloff declared: ‘of the few beeches in central London (e.g. in Hyde Park), most are dying. Here a vitality loss because of ozone seems to be possible.’ By 1989 there was a noticeable recovery: the nine beeches that I found, about 60–100 years old, displayed thinning of the crown, fastigiation, dieback and leaf-curling, but with one exception these symptoms were slight to moderate. The one tree faring worse was a copper beech next to a huge road junction. Hyde Park, though worse than Epping Forest, was thus not an impossible environment for beech.

Hyde Park has benefited from the decline of sulphur dioxide. In 1989, Professor Hawksworth found 12 species of lichens, though they had not yet grown big enough to attract attention.

Pindus Mountains – very unpolluted: In early September 1989 my duties on the Grevená Archaeological Survey took me to the Pindus Mountains around Perivólia and Vovoússa, northwest Greece: a remote and rustic place, not far from Albania; it is 110 miles (180 kilometres) southwest of Thessalonica, 180 miles (290 kilometres) northwest of Athens, and 220 miles (350 kilometres) east of Táranto. Here, if anywhere in Europe, should be a place to investigate the ‘normal’ condition of trees, unaffected by pollution.¹⁸

Beech is widely the dominant tree at altitudes of 4,330–6,000 feet (1,300–1,800 metres). The beeches are huge and magnificent. Many are well over 300 years old. Luxuriant lichens, including abundant Lobaria pulmonaria, confirm the lack of atmospheric pollution.

‘Symptoms’ were very frequent: chaining of short-shoots, dieback at tops, chlorosis, leaf-curling, early leaf fall. These are not confined to old trees. They vary with slope and aspect. In wet places, as at the edge of a small bog, the symptoms were very severe: some trees were withered and a few dead. Reduced ‘vitality’ is evidently chronic, not acute, and rarely kills the tree.

The Pindus beeches are in a ‘worse’ state than those of Epping Forest, but there is no suggestion that this is abnormal. They have been in a state of reduced vitality for centuries, as shown by narrow annual rings. Reduced vitality is one of the reasons for their long lives.

Epping Forest reached its present state by a unique sequence of historical events. As the lichen history shows, acid rain is an integral part of the history of the Forest, like pollarding and the presence of beech itself. The trees were already exposed to acid rain when they began to grow up out of the pollarded state.

Beeches can withstand the recent degree of acid rain without overt damage to their health or even reduction of growth. They have survived greater acid rain in the past. They withstand – with survivable signs of distress – more severe pollution in Hyde Park.

Vehicle fumes, even where locally concentrated, have no visible effect on beech, and very little on hornbeam or oak. If the Forest were suffering a general decline, caused by a widely distributed form of atmospheric pollution such as ozone, symptoms of reduced vitality should be either randomly distributed, with local variation according to individual trees being more or less resistant, or else related to the intensity of pollution. In practice the random symptoms are very mild, and there is no reason to regard them as abnormal. The Pindus observations show that they can exist in the absence of pollution.

Group damage is better related to soils than to evident sources of pollution. It is related to springs and to horsepaths, both of which may interfere with the roots, and to the distribution of plains. I infer that group damage is a manifestation of an unfavourable set of soil factors, which when there was grazing prevented tree growth altogether, creating and maintaining the plains.

The symptoms, though severe, are not progressive. The tree may die, or live with the symptoms indefinitely, or recover. If pollution were the cause, affected trees would only get worse unless the cause diminished.

Lime, the aboriginally dominant tree of Epping Forest, disappeared a thousand years ago. Wood-pasture favoured beech, which became established even in moderately wet places. It could live there indefinitely provided that it was pollarded. Pollard trees, not allowed to form big tops, could withstand having their roots restricted by waterlogging.

After pollarding ceased, the trees grew up for a while, and maiden trees infilled between them. They had not room to form enough root to keep in balance with their enlarged tops. This would not matter in a normal year, but root-restricted trees would be at the mercy of the weather; they could be damaged in unusually wet and unusually dry years, of which there have been many since 1968.

Mounted visitors used to ride all over the Forest; horse trampling destroyed vegetation and soil, especially along wet tracks. By the 1960s this was causing trouble, which was met by constructing gravel paths. On wet ground, churning, excavation and compaction of the soil could all have damaged the shallow delicate roots of the beeches. Some, at least, of the group damage began at the time when this disturbance was at its worst.

It may be argued that pollution in some way interacts with soils – that beech could withstand waterlogging if it did not have to contend with pollution as well. This is a mere hypothesis; the soil theory covers the facts, and there is no reason to suspect any influence of pollution. Group damage occurs in similar wet places in the unpolluted Pindus. The puzzle is that beech should grow so widely on the unsuitable soils of Epping Forest, not that it should sometimes fail to prosper.

Time has passed and knowledge has increased, and there could be complicating factors. Beech is strongly ectotrophic-mycorrhizal: waterlogging, soil disturbance or compaction could damage the fungi rather than the tree directly. In the 2000s one suspects that a Phytophthora may be an intermediary – but in the absence of overt symptoms this would be difficult to prove.

Is group killing abnormal? Dead beeches are usually replaced by birch thickets, among which young beeches sometimes come up. To an American ecologist this would be a good example, the best in Britain, of gap-phase regeneration, one of the textbook methods by which forests replace themselves in the absence of human intervention. A big sugar-maple, for example, rots at the base and crashes down, forming a gap in which short-lived trees, especially other American maples, establish themselves and live for a few decades before being replaced by a new generation of sugar-maple. (American beech short-circuits the process by being clonal (Fig. 11): when a big beech crashes down the gap is immediately filled by its own suckers.) In the mountain forests of spruce and fir (Picea rubens and Abies balsamea) of New England, gaps are started by the assaults of spruce beetle, mistletoe and root diseases; they get bigger as honey-fungus, wind-uprooting and base-breakage take the trees on the edge of a gap. Here too, the abundance of gaps can depend on the long-term history of the stand as a whole.¹⁹

Group damage is part of the process whereby beech retreats from unsuitable soils. Had gaps had not been started by horse and horsepath damage, something else would have initiated them. In the long term beech will probably retreat on to well-drained sites, renewing itself there by gap-phase regeneration. The Forest will not revert to its original natural state dominated by lime.

In May 1981 I was in a wood of many kinds of tree in Connecticut that a month before had been coming into leaf. Now it was almost leafless, and alive with the sound of myriads of tiny jaws. My host showed me the gypsy moth caterpillars (Lymantria dispar), then about 1½ inches (4 centimetres) long, and told me they would grow to 3 inches (8 centimetres) before pupating. I did not see how they could: they had begun on the black cherries and had already eaten almost everything except tulip-tree, which they would not touch.

Gypsy moth was in the last year of an 11-year population cycle that had repeated itself several times. In 1868–9 Monsieur E.L. Trouvelot, astronomer, had brought some caterpillars from Europe to teach them to be silkworms. He got no silk out of them, but lost a few of his livestock; he told the authorities, who were unconcerned. Twenty years later the caterpillars began marching across the United States at a rate of 13 miles (20 kilometres) a year. Attempts to control them by drenching the woods with poison, or by introducing more than 20 other insects, bacteria, or viruses to eat or parasitise them, have made little difference.²⁰

Gypsy moth mainly troubles gardeners who squelch on caterpillars and whose prized and weakly trees die, and foresters whose tree growth is delayed. Its ecological effects are not dramatic, and are difficult to follow in a region where most woods are recent and changing through historical development à la Epping Forest. Most trees can withstand several defoliations: they grow new leaves and carry on as before, but with a shorter growing season and narrower annual ring for that year. A tree that is growing poorly, for instance because of competition, may decline further if defoliated and be gobbled up by honey-fungus.

In Britain, where gypsy moth was last seen in 1907, other caterpillars defoliate oaks. In woods that are monocultures of oak, June can turn into December. At least four species of tortrix-moth caterpillars are involved; defoliation results from unusual abundance of common caterpillars, rather than the appearance of rare caterpillars.

Staverton Park, which is effectively an oakwood, is perhaps the only place in eastern England to be frequently attacked. I recorded attacks in the following years:

This pattern of attacks would probably be usual in north and west Britain, where oakwoods are common; there are reports of such woods being defoliated one year in three. Individual trees are affected haphazardly from year to year; they tend to be clustered rather than scattered.

Freestanding oaks, and oaks in woodland of other trees, are much less prone to attack. However, 1980 was a vintage year. In Hayley Wood many oaks were leafless in June, and defoliation spread to ash, maple, hazel, elm, sallow, aspen and sometimes birch. In Buff Wood nearby, these trees were affected, but not the oaks. In northeast Norfolk attacks spread to hedgerow oaks.

What are the consequences? For ecology probably little, except that caterpillar-eating birds have a feast. Oaks and elms are built to withstand defoliation. In most years they expand their leaves rapidly in May (or latterly April) and then stop growing for several weeks, producing lammas shoots in August (latterly July) with a second flush of leaves. Oaks respond to caterpillar attack by advancing the lammas growth, which for some reason is never defoliated. The oaks of Staverton remain conspicuously healthy despite repeated defoliation, which must be reckoned among the adverse factors that promote long life.

However, a tree that loses its leaves loses the substance that went into the leaves and would have been withdrawn had the leaves fallen naturally – plus the substance that the lost leaves would have made. Since excess substance goes into the year’s annual ring, this should result in a narrow ring. Martin Bridge and I cored a number of oaks in Hayley Wood in 1981, and the results are shown in Table 21.

The annual ring for 1980 was reduced by about one-third (36 per cent) compared to that for the unremarkable years 1978 and 1979. The ring for 1976 (the drought year of the century) was reduced by 53 per cent. That is to say, an extreme caterpillar year had about two-thirds the effect of an extreme drought year. However, the caterpillars, as ever, attacked some oaks more than others, and the sample of 26 trees included two that declined by less than 10 per cent from 1979 and one that increased.

Caterpillar attacks annoy the forest economist, who hates trees to grow slowly. They annoy the dendrochronologist, who likes growth rates to depend on weather (that affects all the trees in a region alike) rather than insects that attack a tree here and a tree there.

Everyone is familiar with the whitish bloom, like a thin coat of whitewash, on oak leaves in summer, especially on coppice and vigorously growing shoots (Fig. 137). This is the visible part of the fungus Microsphæra alphitoides, a mildew that spreads by wind-borne spores. Nineteenth-century mildew specialists knew it only as an American fungus. It suddenly overran Europe in 1908, much like two American mildews that had devastated European vines some decades earlier. It is now commoner on deciduous oaks in Crete and Britain than in America.

Oaks are not vines, and mildew at first sight seems to be trivial, at worst reducing the growth rate. But it may explain the Oak Change that happened at about the same time. In the Middle Ages and long after, oak had been a tree of established woodland, replacing itself in existing woods; millions of small trees were harvested to make timber-framed buildings (Chapter 11). In the twentieth century oak was a pioneer tree: it established itself easily on abandoned fields or along railways, but a young oak in an existing wood was a rarity. It is difficult to create a replica of a medieval building because of the lack of small oaks.

Maybe oak, especially Quercus robur, was always a light-demanding tree, encouraged in prehistory (as Vera would have it,) by particular levels of browsing animals, and in historic times by coppicing. Where there are still abundant oaklings, either the wood is still coppiced (Bradfield Woods, Fig. 76) or the oaklings are concentrated along wide rides (woods on the Blean, Kent). But resuming coppicing has not brought back young oaks to Hayley, Chalkney and many other woods; nor has resumption of browsing by deer.

More plausibly, mildew, in effect, makes oak more sensitive to shade: an oakling may succumb to shade and competition if it has to contend with mildew as well. Mildew appeared when there had been abnormally little felling for half a century, and oaks were not regenerating well. If this interpretation is right, a seemingly trivial introduced disease has had a profound and irrevocable effect on the ecology of oak.

Dutch Elm Disease: Elms used to be the third commonest, and usually the largest, non-woodland tree in England. An epidemic of Dutch Elm Disease flared up in the 1960s and within 20 years had killed 90 per cent of the great elms in England.²¹ A microscopic fungus, Ceratocystis (Ophiostoma) novi-ulmi, poisons the tree and blocks its water-conducting system (Fig. 135). It is carried from tree to tree by elm bark beetles, which breed under the bark of newly dead trees. It is controllable by immediately burning infected trees and to some extent by injecting a fungicide, but only East Sussex County Council and a few private landowners had the energy to do this.

Forty years on, the disease continues to smoulder. The most susceptible elm is English elm, the familiar elm of middle and south England, which no longer exists as a big tree except in East Sussex and among the urban elms of Scarborough and Edinburgh. Surviving big elms are of other species, especially some of the elms of East Anglia, the east Midlands and east Kent; they tend to be in woods rather than freestanding. However, elms have not been exterminated or even much diminished in their distribution. Most are clonal and continue to sprout from the roots. Wych-elm, the exception, grows from seed and within 15 years produces more seed; in regions like the Cotswolds it still flourishes as a small tree.

Ecological effects are less than might be expected. Gaps left by dead elms have been filled by elm suckers (deer permitting) or by other trees such as ash and maple. A few creatures that live on big elms have diminished, such as the white-letter hairstreak butterfly, the agaric bracket-fungus Rhodotus palmatus, and certain lichens. It is no longer so easy to demonstrate the wetwood condition produced by the bacterium Erwinia nimipressuralis, previously the commonest disease of elms, which produces the dark colour and sour smell of elm heartwood.

There was a previous Ceratocystis epidemic in the 1920s and 1930s, especially investigated in Holland (whence the name Dutch Elm Disease), and one in the 1830s and 1840s. Each epidemic attracted similar attention, with calls for infected trees to be felled and burnt, articles on the connection with bark beetles, an editorial in The Times, and attempts to make the government responsible. From time to time Elm Disease has flared up elsewhere in Europe, and lately has been gaining on growth of the elms: a big elm is now rather rare except in outlying corners such as Norway or Crete. The disease was unknown in America until it was introduced on imported logs in the 1920s, and has produced a widespread but patchy epidemic there. It seems to be unknown in Japan.

There can now be little doubt that Elm Disease is responsible for the well-known Elm Decline, which marks the early Neolithic period all over north and middle Europe. What else can have reduced elm pollen production – and only elm – by at least half, without affecting other trees, in at most four years at a site?

Successive waves of Elm Disease have each killed some of the elms (a larger proportion each time round) and then mysteriously declined to an inconspicuous level. At least four organisms are involved: elms, the fungus, two or three species of bark-beetles, and viruses that parasitise the fungus. These are variable and some of them mutate, and from time to time a combination of circumstances generates a virulent form of the fungus that escapes from its limiting factors. A new factor is the tendency to ship logs and bark at random round the world in coals-to-Newcastle fashion. Thus was the disease brought to America, and it is widely held that the virulent strain of the 1960s arose in America and was brought back to Europe.

HENRY WADSWORTH LONGFELLOW (1839), REFERRING TO CASTANEA DENTATA, THE SWEET-CHESTNUT OF EASTERN NORTH AMERICA

As a student I was taught about Endothia parasitica, chestnut-blight fungus, which produces cankers on sweet-chestnut that spread and girdle and kill the branch or trunk. Its spores ride the wind, an insect vector not being essential. It is said to have come to America from Japan, where it occurs on Castanea crenata but is not a problem. (American horticulturalists, such as the great Luther Burbank, were not content with American chestnut, but must needs import quantities of Japanese material.) It spread unopposed through C. dentata, converting it from a great forest tree to a small understorey species. Chestnut does not easily rot, so the woods are still full of the mighty remains of multi-stemmed self-coppicing stools. It reached southern Europe to ravage C. sativa and the commerce in edible chestnuts, which in the Apennines had been a staple foodstuff.

Many years later, I was astonished when an Italian colleague told me that chestnut-blight was no longer troublesome. Chestnuts have been rescued by a virus that attacks and cripples the fungus. The disease still exists, but the cankers are limited and no longer kill the tree (Fig. 136). Chestnut is tenacious of life, and even a three-quarters-killed tree may fully recover. Once infected by a virus-infected fungus, a tree is permanently armoured against further Endothia attack.²²

The northern Apennines are covered with chestnut-woods, stretching over hill and valley, except for beech on the higher mountains. They are often outgrown from chestnut orchards; there are remains of the alberghi that dried the nuts and the water mills that ground them into flour. A few groves are still productive; most are converted to coppices or neglected. Many of the great trees display high dead tops, relics of an Endothia attack defeated by the unseen ally. Woodcutters are familiar with cancro del castagno, which has left iron-hard lumps of canker, embedded deep in the trunk to take the teeth off an unwary chainsaw.

This is the story of chestnut-blight in the European heartland of chestnut, in Italy and southern France. The disease never reached outlying populations, such as the introduced chestnuts of England and Crete. On Athos, in north Greece, it arrived in the 1990s; the monks, who earn their living from chestnut coppice, have introduced the virus. With American chestnut the story has no happy ending. Although the virus got there, attempts to disperse it have been less successful, and the spreading chestnut-tree is now rarer than the village smithy.

Oak wilt: Forty years ago John Rishbeth lectured on a fungus that killed oaks in North America in the manner of Elm Disease. Ceratocystis fagacearum is mainly wind borne, and like Elm Disease persists for less than a year in logs. Since the 1920s oak wilt has been spreading out from a focus in Minnesota. It is assumed to be an introduction from an unknown country, but has still not been found outside America. It attacks two of the three groups of American oak species, red oaks and live-oaks, but white oaks were less affected. By 1983 it was killing the willow-oaks (Quercus phellos, a red oak) on the streets of Washington.

Oak wilt smoulders on. There is a second focus in middle Texas, where I have known it since 1983 (Fig. 138); it attacks the motts (clonal patches) of live-oak (Q. fusiformis) that dot the savanna. It usually kills an entire mott to the ground, but there may be regrowth if cattle allow it. (Elm motts are of Ulmus crassifolia, which seems to be unaffected by Elm Disease.) In 25 years it has killed about 15 per cent of the live-oaks, and seems to be slowing. It sometimes kills ‘Spanish’ oak (Q. texana), a red oak, but I have not seen it attack a white oak. It is carried partly by insects that are attracted to wounds on trees. It invades long-standing plant communities in the manner expected of an introduction.

For 40 years there have been rumours of oak wilt reaching Europe, and speculation on what would happen. It would probably not have the devastating effect of Elm Disease. There are no native red oaks in Europe; European deciduous oaks are white oaks and likely to be resistant. Whether the evergreen oaks of southern Europe, of the live-oak group, are susceptible is not known.

The dying pines of Japan: The mountains around Hiroshima look as if they have been hit by an atomic bomb: mile upon mile, millions of dead pines cover the slopes (Fig. 139). This is the latest stage in the spread of a microscopic, deadly nematode worm, said to have been introduced from America in the 1930s, which has gradually devoured the red pines of Japan.

Aka-matsu, red pine, Pinus densiflora, which resembles Scots pine, is a familiar and much-loved tree; its curved timbers are a feature of old-fashioned Japanese houses. It (and its edible matsutake fungus, was encouraged by a peculiar compost-gathering land-management practice that has no parallel in Europe. One of the profoundest changes between sixteenth-century pictures and the present landscape is the huge loss of red pine. Declining land management left the pines mainly on ridgetops that would have been their natural habitat, but the micro-worm has sought them out even there.

Pines have been replaced by evergreen oaks, hollies and others among the multitudinous evergreen trees of Japan. The parasite apparently does not matter in America, nor does it much affect other Japanese pines. What would happen if it reached Europe?

Bloody fluxes: Phytophthora is a large genus of microscopic fungi of the phycomycete kind (if indeed they are fungi), which thrive in temporarily wet conditions. Some ‘damp-off’ seedlings, or blight potato leaves and tubers. Those that attack full-grown trees are traditionally called Ph. cinnamomi or cambivora, terms that cover a group of species varying in host and in virulence. They spread by water-borne spores; they attack the cambium and may kill the tree. A diagnostic symptom is rusty-red, sticky exudations on the bark, looking like dried blood (Fig. 140).

The best-known European tree Phytophthora is ink disease of sweet-chestnut, which kills roots and sometimes the tree; it was a well-known problem before chestnut blight came. It favours wet seasons and waterlogged soils. It may be a native fungus, attacking what is usually an introduced tree.

In the 1960s John Rishbeth and I took an interest in an avenue of horsechestnuts belonging to our Cambridge college. Individual trees developed yellow, stunted leaves and died in two or three years. At first the cause seemed to be bacterial wetwood in the trunk spreading outwards and interfering with sap conduction. Later research found a Phytophthora damaging the cambium at the base of the trunk, which allowed wetwood to gain the upper hand and kill the tree. Phytophthora, in the manner of its kind, soon disappears, leaving other fungi to invade the dead tissue. The trees were planted c.1881, began to die at about 70 years of age, and by 2004, 26 out of 53 trees had died. They are not scattered at random: groups of dead trees alternate with stretches of healthy trees.

This set me looking for dried blood on wild trees. In the 1970s Phytophthora-like exudations were common on birch and sallow, and also found on beech, alder, willow, holly, cherry, aspen and oak. Some host trees were faring badly or dying from some other cause, but most recovered. The fungus probably entered through a small wound or twig scar, killed a patch of cambium, and then often died out, leaving a wound which was covered over by further growth of the tree. These limited attacks seem to have become less common since 1980.

As a student I learnt of the devastating effects of Ph. cinnamomi on the native vegetation of southwest Australia, especially on jarrah, one of the great timber eucalypts. Many years later I met Dr Frank Podger, whose articles I had read, and who explained how it had progressed by 1996. The scene is the mighty jarrah forests east of Perth, plant communities with hundreds of species, virtually every one peculiar to that small region, many of them lignotuberous (p.16n) and all highly fire-dependent. The climate is mediterraneoid, with hot, dry summers and warm, wet winters. Logging began in the nineteenth century: its effect might not have been too disastrous had it not happened that into this new planet, as it were, a common tropical root parasite descended in the 1920s.²³

Phytophthora first takes out the banksias, small trees along whose roots it spreads to other species. Then it rots the blackboys (Xanthorrhœa) – an extraordinary plant, which flowers in profusion after each fire – and zaps the cycads. Dryandra, a common genus of small trees, is not much affected, and only two of the six common eucalypts are susceptible. Ultimately it gets all the jarrah (Eucalyptus marginata), without hurting the other great eucalypt, marri (E. calophylla). The result is an impoverished native ecosystem, dominated by marri, from which jarrah and a random set of other species have been subtracted.

It was discovered, too late, that Phytophthora is spread by making gravel roads using gravel from infected pits. From these artificial sources it spreads at about 3 feet (1 metre) a year, faster in wet seasons. Its spread can be slowed, especially by keeping dirty vehicles out of uninfected areas. The cause is not lost: in an area about the size of Yorkshire nearly half is still unaffected.

In the 1990s another Phytophthora came to public attention in England through reports of sudden death of alders and oaks. Alder disease is a classic Phytophthora, attacking the cambium of the root and lower trunk, exuding dried blood, and more virulent near water. If the alders are felled the regrowth may escape. It was first noticed in the 1970s; it has killed probably about 10 per cent of the alders in some areas, such as Herefordshire and southeast England, and is beginning to attack East Anglia. It is widespread in Europe, and has attracted attention in Hungary where there are alder plantations. The causal species is reported to be a hybrid between two introduced Phytophthoras.²⁴

At the same time dieback of oak was attracting renewed attention. On the Continent, there was some evidence for attack on oak roots by what was named Ph. quercina, which may be involved here too. Phytophthora-type exudations are scattered here and there in both dying and normal oaks.

Horsechestnut Phytophthora is now widespread.²⁵ A spectacular example is in Hatfield Forest, whose owners, the Houblons, whimsically introduced this tree in the 1850s. For a Balkan cliff endemic it fared surprisingly well on the Essex boulder-clay, and became the tallest tree in the Forest; but Nemesis has caught up with it. Year by year the horsechestnuts succumb, leaving mighty snags that are an excellent deadwood habitat (Fig. 143).

The scare of the year 2004 was the so-called Phytophthora ramorum, supposedly from Central Asia. On invading California and Oregon, it affected many trees, especially the local species of oak. At the time of writing it was reported in western England from rhododendrons and several other species of garden trees, but there was no sign that it would attack native oaks.

The first problem with environmental mishap and tree disease is defining normality. The public has been trained by 250 years of Enlightenment to expect trees to be upright and single-stemmed, not to have dead branches, not to be rotten, to come into leaf and to lose their leaves at the right time of year, and to die of old age. This may be what foresters and gardeners would like, but does not always agree with the agenda of trees as wildlife, nor with the concordat worked out between people and trees over the centuries.

High expectations can distort the views of scientists too. Reports of the declining health of beeches and other trees amounted to little more than that trees were failing to conform to standards of perfection set in Germany. One hardly expects the normal state of a human being to be perfect health: why should this be true of other creatures?

The Pindus beeches explode the implication that any tree that does not measure up to Central European standards of vitality must be a damaged tree. To achieve those standards, a beech needs to avoid being damaged and also to have the right climate and soil. Nature did not intend the Pindus, and maybe not England, to be the perfect environment for beech. In Central Europe, the tree is in the middle of its range, and perfect vitality may be its normal state when young. In England, beech is at the edge of its range, and perfect vitality may be unusual. In the Pindus, beech is at another edge of its range, and demonstrates that it can survive – indeed can be the dominant tree – despite poor vitality.

Much beech has been planted and suffers from the disadvantages of planted trees, especially trees introduced beyond their range or from foreign sources. People who plant beech often have no clear idea of where it will prosper. Exiled into a climate to which they are not adapted, dug up from a nursery bed, stuck into a hole in the wrong soil, exposed to hostile fungi and nematodes, separated from their mycorrhizal partners – it is surprising how often planted trees succeed at all, let alone conform to external standards of good condition.

Great storms are rare but normal events, to which trees are adapted. So are defoliations by native caterpillars. Stag-headed oaks are part of normal behaviour, a response to a rare but normal combination of events, except that a minor contributory factor may have been oak mildew.

Dutch Elm Disease is a borderline case. Epidemics may be part of the normal interaction between the fungus and its host: some might argue that the clonal habit of elms is an evolutionary response to the disease. Abnormality is shown by epidemics getting more frequent and more severe, and spreading round the globe. As with other conditions involving tree sprouting and reproduction, its effect is made worse by the increase of deer.

Plants and animals often come to terms with parasites through evolution. Over successive generations, an annual plant species develops a reaction to infection, for example by secreting chemicals that attack the invading fungus. The fungus then develops a means of avoiding damage by the chemicals. This interaction works up to a point – if the life cycles of the host and parasite are not too dissimilar. In my plant physiology days I experimented with a variety of barley, bred for resistance to mildew, called Maris Concord, in the year when, alas, a Maris-Concord-resistant strain of mildew appeared. This problem has dogged plant breeders over the years.

The interaction works less well with long-lived or clonal hosts. In evolutionary terms, plant pathogens can run rings round trees. To develop genetic resistance to poplar-mercury rust, aspen has to start again from seed, which it does perhaps once in a hundred years, in which time the rust will have gone through 200 generations. A big elm clone may have had to cope with at least three epidemics of Elm Disease.

Disease resistance in trees has to be of a more structural and unspecialised kind. Sometimes it depends on symbiotic mycorrhizal or decay fungi. Dr Rishbeth invented a means of preventing the root fungus Fomes annosus from killing pine plantations. Usually it starts in a stump and spreads through the soil to attack living trees. New stumps can be deliberately infected with the harmless fungus Peniophora gigantea, which occupies the stump and denies it to Fomes.

Normal dynamics are circumvented by modern humanity’s flair for mixing up the world’s pathogens – human, animal and plant. Oak mildew shows how a seemingly insignificant introduction can profoundly upset the normal course of nature. The other oak conditions, dieback and wilt, seem not to be a problem for Britain, destructive though they are in other continents. Pine nematode, though it spreads relatively slowly, has changed the face of Japan. Phytophthora attack on the native vegetation of Western Australia has been called ‘a biological disaster of global significance’: it is comparable to the havoc wrought on small marsupials by introduced cats and foxes.

1980	none (despite unusually severe attacks elsewhere)
1981	slight to moderate
1982	slight
1983	severe: 70% of oaks attacked to some degree, c.10% wholly leafless
1984	slight
1985	only one tree attacked
1986	severe on about six trees
1987	severe
1988–9	slight or none
1990	very slight
1991	slight
1992	severe but local
1993	very severe, except in The Thicks where oaks are mixed with hollies
1994	moderate but not in The Thicks