I’m currently engrossed in Vera’s book Grazing Ecology and Forest History, which I cannot rate high enough for its readability, lucidity, and coherency. Whilst on my holiday last week, I read through the third chapter on the study of palynology and how this relates to interpreting how treed landscapes may once have looked, and I have to be honest when I say that the entire field was somewhat (though not wholly) new to me, and what was suggested within the book certainly made me think. For this reason, I’m going to write a little bit about pollen studies and how it can effectively be used, or even ineffectively used, to determine what our landscapes once looked like, with regards to what trees existed, and in what abundance / distribution, according to Vera.
Vera begins by ‘setting the scene’, by describing how the Swedish geologist, Von Post, in 1916, produced what is considered the first pollen diagram. Prior to the utilisation of pollen, typically accumulated in peat bogs and lakes (which are regional pollen sinks), larger parts of plants were used in an attempt to understand what the landscape once looked like (up until the last Ice Age some 12,000 years ago) – leaves, fruits, tree stumps, and possibly even larger seeds were three means of how the landscape’s vegetation history was being deciphered, and again these were usually found within peat bogs. In this sense, prior to 1916, the vegetation composition of a landscape was being understood through assessing how plant macro-fossils were distributed (vertically) in peat bogs. For example, if a stump of a pine tree was found below the leaves of a willow tree, one could suggest that pine trees existed prior to willow trees in the geographical area. After 1916, pollen, which readily remains desposited in such aforementioned naturally-occurring sinks, could instead be used. Granted, pollen generally only persists for wind-pollinated species, with the exception of poplars, so one cannot, in theory, decipher the exact presence of tree (and plant – grasses, etc) species – one can instead only interpret, based on the facts gathered.
By-and-large, following Von Post’s landmark pollen diagram in 1916, studies into pollen presence had suggested that the landscape was once almost wholly covered in trees (where tree cover was possible, due to biotic and abiotic factors). This is because, when pollen studies have been undertaken, the large majority of pollen found has been from tree species (usually, non-arboreal pollen amounts for no greater than 5-20% of total pollen in the sinks). Historically, and prior to 1934, when Firbas published a paper on how one can also identify and use the pollen of grasses and shrubs to determine landscape composition, there was also a choice to ‘ignore’ the pollen of non-tree species. Because of these factors, scientific opinion was generally that grasslands and wood pastures are an advent of agriculture and man’s influence upon the landscape, in place of wild ungulates (auroch, bison, boar, deer, Przewalski’s horse, and so on) influencing upon the vegetation composition. Therefore, prior to modern man, the European and American landscape was largely void of expansive steppes and pastures, where the land was potentially habitable by trees. The wild ungulates were thus not seen as responsble for carving the landscape, and thus only existed in low numbers within treed landscapes. Instead, these wild herbivores followed the regression and regeneration of trees in the landscape.
However, where it gets interesting is when one looks at what tree species were present in the pollen records. Before we look further at this however, we must recognise that the dense forest will generally be host only to shade tolerant tree species (beech, lime), assuming it has reached its ‘climax’ (prior to this climax, more light-demanding species will be present, initially with birch, hazel, pine, willow, and so on, and then with species such as oak). For this reason, if we assume that the historic landscapes were covered with high forest, we can assume that much of this high forest will be of climax species, as man was not historically around to carve apart such landscapes with cattle and for arable activities. Despite this, this is not what the pollen records show. In fact, hazel (Corylus avellana) and oak (Quercus robur) contribute quite significantly to pollen records, and as neither species will regenerate in high forest (because they are not shade tolerant), how is it possible that large tracts of the landscape were high forest? Unless the species were able to regenerate significantly enough in high forests to feature so readily in pollen records, which goes against the species’ understood biology and ecology, there must have existed landscapes where significant light was able to reach the floor. This is where Vera suggests that the landscape could very well have been shaped by wild grazing animals, who kept large areas adjacent to groups of trees or forests open (where there was the ‘mantle and fringe’ vegetation), and the thorny scrub that grew within such a grazed landscape enabled for hazel to grow in thickets and oak to succeed within such thickets (of hazel, and particular thorny scrub, on which ungulates would not generally graze). Oak, in particular, can in modern day be observed not to regenerate in high forest, but in grazed areas amongst thorny scrub (I myself saw this the other day at Dunwich Forest and nearby heathlands, where oak was regenerating not amongst high forest, but within the gorse and bramble scrub).
Vera also raised concerns over interpreting the high amounts of tree (arboreal) pollen in pollen records as meaning the landscape was largely comprised of trees. This is because the pollen sinks, as already stated, are generally regional (peat bogs and lakes, of which large lakes are more often used). Because tree pollen is released early in the season, and is usually released in high abundance at an elevated level in the canopy, there is a much greater chance of tree pollen travelling greater distances, where it will reach these regional pollen sinks. Conversely, grasses release pollen during the summer, and at levels just above the ground, where winds are less strong and there is a greater chance of the pollen not travelling too far (because of the lower wind speeds, and the trees and shrubs in leaf ‘trap’ the pollen in situ). As a result, a regional sink, such as a peat bog, even if large areas of land, even almost adjacent to the bog, were grassland or pasture that were bordered by trees, there is still a very high probability of non-arboreal pollen not accounting for more than 10-25% of the total pollen distribution in a sample. Not only this, but even if we assume that the landscape was wood pasture where animals grazed, the suppression of the grass by grazing herbivores and the fact that open-grown trees have much larger, fuller crowns, means that pollen ratios between arboreal and non-arboreal sources will likely register as if the area was instead a forest (for example, the total crown area of an area of wood pasture and of high forest may not be all that different). Trees in wood pasture will also have more clearance for pollen to travel great distances, and thus end up in these regional sinks at high levels. Even modern-day records suggest exactly this, and in this sense a wood pasture can be interpreted as, if assessed on pollen records alone, high, dense forest. Of course, this suggests that pollen records only tell part of the story, and it is easy to mis-interpret findings based on pollen studies.
If you have found this post interesting, then please do consider buying the book. There is no way that I can give the whole picture here, and instead I have only given a fragment. Hopefully, it makes sense, and hopefully it gives an indication of why suggesting that the landscape was once comprised of massive expanses of high forest is perhaps not entirely accurate. In the modern day, there is no doubt that grazing by cattle has suppressed the regeneration of forest, and man’s conservation efforts with heathlands and grasslands has also stopped forest regeneration; as has man’s carving-up of the landscape for building and development. However, historically, when wild herbivores were still actually in existence, as man hadn’t pit-falled the last auroch to its death, the landscape may have not been covered exclusively by high forest where conditions allowed. Considering that fire is not seen as a massive driver behind the regression of forests to grassland and then back to a form of woodland at a later date, and the beaver is not considered to have been the only mammalian influence behind the loss of forest patches (again, according to Vera and the sources he immersed himself in), perhaps wild ungulates had more of a role in shaping the landscape than is generally considered. Food for thought, no doubt. Graze on that literary resource, and head out for pannage in your local library.
Source: Vera, F. (2000) Grazing Ecology and Forest History. UK: CABI Publishing.
If you have found this post interesting, then please consider sharing it, leaving a comment below, or alternatively posting over on Arbtalk.