Land cover and soil properties also influence the ease of progress. However, land cover may have changed considerably over time. Some examples are given by Bell et al. (2002) – swamps can be drained, forests can be felled, and scrubland cleared. A comprehensive discussion of soil forming and transformation processes such as gleying or podsolisation can be found in Waugh (2000, 271–3). As a result of landscape modifications, reconstructing the vegetation for a given period of the past is difficult. Howey (2007) describes a rare case for which this was possible. In Michigan, biologists translated the notes of official surveys conducted between 1816 and 1856 into a digital land cover map. This map was used as an approximation of the landscape for a study dealing with the years 1200–1600 AD. Still, any reconstruction of past vegetation cover is not as detailed as modern geographic data.
Alternatively, soil or geological data may be used to reconstruct the vegetation cover and to identify swamps, the crossing of which is assumed to be very costly. Since soil or geological data are closely connected with the land cover, these two issues are considered together.
Table 3 presents some of the multipliers found in physiological studies. Soule and Goldman (1972) recorded walking data in a landscape with light brush, heavy brush, swampy bog and on a dirt road. These data are compared with the energy expenditure required for walking on a blacktop road. Givoni and Goldman (1971) tentatively propose some multipliers for ploughed fields and hard snow. Walking on sand is analysed in several studies including Pandolf et al. (1977); multipliers in the range of 1.6 and 2.7 were suggested, so a multiplier of 2.0 seems an appropriate compromise.
The multipliers proposed in Table 3 are a lot lower than those published by the European Commission (1995–2010). However, the physiological studies measure energy expenditure, whereas the European Commission publication focuses on travel time. The study of the European Commission assumes a travel time of 60 minutes per km in an evergreen landscape with broadleaved tree cover and for regularly flooded areas with trees or shrubs. In their accessibility model (of the modern world), this is the worst case; for cultivated and managed areas they calculate 36 minutes per km, whereas in their model, the fastest walking speed on a landscape without paths can be achieved on sparse herbaceous or sparse shrub cover or bare ground (24 minutes per km).
In general, vegetation cover and soil properties vary widely and Langmuir (2004, 150–3) specifies many landscape features that impede progress: Steep vegetated slopes can cause problems particularly when they are wet, and care should be taken to avoid areas of turf without firm attachment covering slippery rock. Boulder fields can provide loose insurmountable obstacles and descent across such a field requires extra concentration and agility. Similarly, handling loose rock requires a combination of care, balance and alertness. Wet rock can pose a problem, if it is made greasy by lichen or other vegetation. Furthermore, peat bogs and tussock grass may impede progress. It seems beyond the scope of any archaeological LCP study to take all these features into account.
Most multipliers given above assume that the path is not yet there. However, often paths are formed by repeated trampling or by construction work, and in these two cases, some of the multipliers are no longer appropriate. Creating a path by repeatedly and regularly walking on a given route is hardly possible in some types of vegetation, such as dense forest, and fairly easy on others such as grassland. For this reason, vegetation cover should be mainly considered in the context of path formation. Yu et al. (2003) present some tables for calculating the cost of building modern roads. These costs depend on the land cover, and the authors experiment with two different configurations: In the first configuration the costs are 1 for farmland or grass, 2 for forest, and 10 for urban areas, whereas the second configuration sets the costs for building roads on farmland or grass to 2, those within forest to 3 and in urban areas to 6.
However, even if a path exists, wetness and snow or ice play an important role, and this can be avoided only by extensive construction work such as tunnels. Pandolf et al. (1977) found in their experiments that the energy expenditure for walking in 35cm of snow is four times as high as on a tarred road.