Herbivores and the Holocene did the lowland European forest have a closed canopy

The steppe-tundra present at the end of the last ice age was open and dominated by grasses and herbs. When the temperature rose, animals adapted to very cold conditions, like the mammoth Mammuthus primigenius, moved away. As they did so a more familiar fauna of large ungulates, including the auroch Bos primigenius and the tarpan Equus przewalski gmelini which were, respectively, the wild ancestors of cattle and the domestic horse, took their place. Fossil remains show that these two species remained prominent well into the Holocene, roaming and grazing primeval vegetation along with other wild indigenous ungulates, including the European bison Bison bonasus, red deer, elk, roe deer and wild boar (see Fig. 5.12). These facts have long been known, it is the size of their populations and the nature of the forests in which they lived that are the critical issues raised by Frans Vera (2000, 2002). Until very recently it was widely assumed that until large-scale human intervention started, the lowlands of central and western Europe, which have a temperate climate, would have been covered with closed canopy forest wherever trees could grow. This view is largely based on studies of modern abandoned fields and pastures, which in the absence of heavy grazing pressure give rise to just such a wood or forest. Once established, such closed canopy forests are maintained by a variety of mechanisms, including patch dynamics, small regeneration gaps and regeneration waves, discussed in Section 9.2 and 9.3.

Vera, however, has posed an interesting re-interpretation of the available evidence, suggesting that much of Europe was a more open and park-like wood pasture, not unlike that now found in much of the New Forest, southern England and, to a certain extent, in the Oostvaardersplassen Reserve in The Netherlands (Vines, 2002 provides a good review). Vera suggests that these have developed in response to the grazing of a considerable population of large herbivores, particularly aurochs, horses and bison, keeping the woodland more open. Had populations of these animals become even greater, extensive areas of grassland or heathland would have formed as a result of the retrogressive succession of woodland.

Vera's view is that many of the lowlands of central and western Europe and eastern North America were covered by a grazed park-like landscape containing groves of trees c. 7000 years ago, before the arrival of humans (Hodder et al., 2005). Individual groves enlarged their boundaries at the speed with which shrubs such as blackthorn could spread by means of their underground rootstocks (around 0.1-0.5 ha in 10 years). Isolated individual shrubs such as hawthorn would also appear as a result of animal dispersal of seed, helped by reduced grazing pressure in the summer (rapidly growing grass means less extensive grazing). These prickly shrubs gave shelter which allowed the seeds of animal- or wind-dispersed trees to develop. Jays, for example, transport acorns considerable distances, bury them and often fail to return to their food stores. On poorer soils trees could grow up protected by shrubs such as juniper or bramble. Thus, open park areas would be gradually invaded by forest. The groves would be dense, producing tall unbranched trees in the centre (as found in preserved bog oaks) with more branched trees at the edge. Within the groves, mature oaks were not replaced by young shade-tolerant trees because these were destroyed by large grazing herbivores wherever a gap began to form and grasses moved in. Eventually gaps became larger as older trees were windblown or died as a result of fungal infection, and individual groves degraded into grassland. Thus the presence of a considerable population of r

-1. PARK PHASE

Open landscape with thin scatter of trees

4. BREAK-UP

Spread of thorny scrub excludes herbivores; young trees grow up overtopping scrub

2. SCRUB PHASE

Canopy opens up as trees die; main vegetation shifts from woodland to grassland

3. GROVE PHASE

Tree-dominated; closed canopy shades out the shrubs; herbivores return preventing tree regeneration

Figure 5.17 Essentials of Vera's model of primeval forest succession, together with the scrub phase added by Kirby (2003).

large grazing herbivores (probably helped by fires in dying areas) encouraged a shifting sequence: grassland - thorny shrubs - grove - grassland - thorny shrub - grove... (Fig. 5.17). This process, which Vera terms the cyclical turnover of vegetation, may well have been important in the long history of many continental European forests. The case, however, must not be pressed further than the evidence warrants. Closed canopy forests do exist and are discussed later (Section 9.4).

Kirby (2003) suggests that most grove trees would be expected to live for 300-400 years but this does not preclude older veteran trees (> 450 years old) so typical of the British landscape. Kirby also suggests that some grove trees would survive as solitary trees until the young trees of the next grove phase over-topped the veterans and shaded them out.

Evidence bearing upon this problem is drawn from the pollen records of undisturbed prehistoric forests, population trends in forest reserves studied over a long period, and documentary evidence which in the case of forest charters of the Merovingian and Frankish kings goes back to the seventh century. Documentary evidence, particularly that provided by Forest laws and regulations in the various European countries, has to be treated with considerable caution as the meaning of various familiar words has changed with time, and the reasons for some prohibitions demand considerable understanding. Pasture, for example, is now thought of as consisting of grassland affording food for cattle and sheep, whereas in the Middle Ages the concept applied to anywhere animals could obtain food, including such trees as the oak, wild apple, wild pear and wild cherry.

Vera's ideas require a re-thinking of the evidence which has been previously interpreted as showing a dense forest. His view is that the open parkland explains why hazel, pedunculate oak and sessile oak (and other light-demanding species) have been well represented in pollen records for thousands of years, along with that of shade-tolerant species such as limes, elms, ash, common beech and hornbeam. In closed-canopy forests and forest reserves where large gaps are not present, oaks tend gradually to diminish because their seedlings, unlike those of the shade-tolerant trees, cannot grow at the low light levels present in the limited gaps which do form. He also contends that a partial explanation for the very high proportion of tree pollen dating from this period is that grazing may have been so efficient that production of grass pollen per unit area was greatly reduced. Svenning (2002) counters this by pointing out in a review of north-west Europe that in many studies non-tree pollen correlates well with other measures of openness such as beetle, snail and plant macro-fossils and concludes that forested conditions were the norm with open vegetation being restricted to floodplains or poor soils (sandy or calcareous) and in the continental interior of north-west Europe. It is also possible that the frequent pollen from light-demanding species may have come from relatively few trees growing in a limited number of openings in an otherwise dense forest. Some of the debate is hampered by just what a landscape driven by large herbivores would have looked like. Was it open savanna with less than 30% cover of trees or mainly a wooded landscape (> 70% tree cover) but with larger, shifting glades?

An insight into a developing situation of this type can be gained by measuring the trunk diameters of the main tree species present. When results of this kind were plotted in 10-cm diameter classes for trees at Dalby Soderskog in Sweden, a very clear picture emerged (Fig. 5.18). Here the number of small trees of pedunculate oak was very low; together with the high proportion of large trees, this indicated an increasing recruitment failure. In closed-canopy sites these light-demanding species are being replaced by shade-tolerant trees including ash, beech, elm, lime and sycamore. In such situations these species have a high proportion of smaller (and younger) individuals, so the curve of their diameter classes forms a reversed J, indicating successful regeneration.

Pendunculate oak (n = 1020)

20 40 60 80 100

Wych elm (n=7500)

Pendunculate oak (n = 1020)

Wych elm (n=7500)

20 40 60 80 100

20 40 60 80 100

20 40 60 80 100

Diameter at chest height (cm)

20 40 60 80 100

Figure 5.18 Percentage distribution into diameter categories for the main tree species growing at Dalby Soderskog, Sweden. Only trees with a trunk diameter of > 10 cm at chest height are included. (Redrawn from Malmer et al., 1978. From Vera, 2002. Arboricultural Journal 26.)

20 40 60 80 100

20 40 60 80 100

Diameter at chest height (cm)

20 40 60 80 100

Figure 5.18 Percentage distribution into diameter categories for the main tree species growing at Dalby Soderskog, Sweden. Only trees with a trunk diameter of > 10 cm at chest height are included. (Redrawn from Malmer et al., 1978. From Vera, 2002. Arboricultural Journal 26.)

Vera's ideas raise many interesting questions and may help explain a number of features of European woodlands. For example, does the existence of large open spaces explain why green woodpeckers Picus viridis feed on the ground, favouring ants? Or why there are so many common bird groups that feed on open-ground insects, such as starlings Sturnus vulgaris, wagtails Montacilla spp., pipits Anthus spp., and many corvids like crows and magpies (P. R. Hobson pers. comm.).

How Vera's ideas apply in the British Isles has been discussed in detail by Rackham (2003), Kirby (2003) and Hodder et al. (2005), who welcome many of Vera's insights though questioning a number of important points. In addition to the openness of the forest, a second area of debate is whether the openness here could have been driven solely by large herbivores. Conditions in the British Isles were markedly different from the continent, in that the number of large animals surviving the last glaciation was very restricted. Native horses and bison have not been present in this interglacial. Elk and bear probably died out in the Mesolithic (Middle Stone Age) and the auroch, which lived more in fens than forests, in the Bronze Age. Even the remaining red and roe deer, wild boar and beaver were uncommon by historic times. England was also more densely populated than continental Europe and more than half the original wildwood had gone by the Iron Age or even earlier. Oak may not always behave as Vera supposes, indeed in some English woods from which deer were largely excluded, occasional thickets of sessile oak resulted from the great mast year of 1976 and similar events. It is suggested that in the UK abiotic disturbance such as fire and windthrow may have had important roles in maintaining open spaces (Hodder et al., 2005). Certainly the open landscape in the New Forest is currently maintained by such means as heather burning, and the numbers of grazing animals are managed. Hodder and Bullock sum up the debate as follows:

The difference between the closed forest hypothesis and Vera's alternative of cyclical dynamics may be a matter of degree. While there is general agreement that the original-natural forest [in the UK] may have been more open than was previously thought, this is not equivalent to saying that a wood-pasture landscape would necessarily dominate the landscape. The balance of opinion is towards predominance of closed forest with localised, longer lasting openings.

On the whole it would seem that the British Isles landscape probably consisted of large tracts of wooded pasture with groves covering expansive areas. Rivers and floodplains would have acted as corridors for the large herbivores which migrated from one feeding ground to the other. This would certainly help explain the persistence of plants restricted to old woodlands (the British Ancient Woodland indicators) that are estimated to require many hundreds of years to invade isolated woodlands. These species and their dynamics do not fit a landscape made up of shifting, patchy groves in a sea of grass and scrub. To envisage genetically viable metapopulations of woodland plants in discontinuous, patchy landscapes would be almost impossible, particularly for species such as toothwort Lathraea squamaria. It is possible, however, that these plants have previously persisted in small woods on steep ground or islands inaccessible to large grazing animals.

Mitchell (2005) discusses the openness of European primeval forests using palaeoecological evidence from Ireland where only two native large herbivores, the wild boar and the red deer, were present in the early Holocene. This was in marked contrast with England and Scotland, which at that time also possessed roe deer, elk, reindeer, horses, aurochs, and beaver, and continental north-west Europe where bison and fallow deer were also present. This gives a quite exceptional opportunity of testing Vera's view that many large herbivores were required to create forest gaps and an open park-like landscape in which the high abundance of oak Quercus and hazel Corylus pollen found in European pollen diagrams could develop. The comparison is conclusive: the similarity between data sets from Ireland and continental Europe is so close as to make it clear that large herbivores were not required to maintain large oak and hazel populations in the primeval landscape. Mitchell goes on to demonstrate that percentage tree pollen is a reliable indicator of canopy openness in both Europe and eastern USA, where the palaeoecological data demonstrate that open canopy forest has only ever been maintained by human exploitation.

Rackham inclines to the view that in cases where beech is not involved, the landscape was less dynamic than Vera supposes. Beech was in any case mainly found in south-east England and concentrated in wood pastures in historic times. Moreover, topography and substrate conditions play a major role in the long-continued distribution of trees and grassland the world over. In surviving wood pastures the woods tend to be on hilltops and the pastures in the valleys, while individual trees favour rocky places, the shelter of thorny shrubs and grassland soil.

Vera's ideas do not answer all our questions about how woodlands used to look, and indeed have resulted in even more that still need to be answered. Whatever the eventual outcome, it is this type of innovative thinking that helps ecological science to develop.

Was this article helpful?

0 0
Worm Farming

Worm Farming

Do You Want To Learn More About Green Living That Can Save You Money? Discover How To Create A Worm Farm From Scratch! Recycling has caught on with a more people as the years go by. Well, now theres another way to recycle that may seem unconventional at first, but it can save you money down the road.

Get My Free Ebook


Post a comment