20. Steppe-tundra

20a Steppe-like steppe tundra

Storage (tC/ha)	Ecosystem component
0 tC/ha	Dead standing trees, coarse woody litter, leaf litter and other debris
5 tC/ha	Above and below-ground vegetation
55 tC/ha	Soil organic carbon
60 tC/ha	Total carbon storage

.

20b Tundra-like steppe tundra

Storage (tC/ha)	Ecosystem component
0 tC/ha	Dead standing trees, coarse woody litter, leaf litter and other debris
5 tC/ha	Above and below-ground vegetation
55 tC/ha	Soil organic carbon
60 tC/h	Total carbon storage

.

This sparse dry-climate vegetation type, or rather this set of vegetation types, was widespread during the Last Glacial in the mid latitudes of North America and Eurasia (see the QEN pages and maps). As this vegetation does not apparently exist in the present-day world (it is widely described as a 'no present analogue' biome). The methods used to infer the likely character of the 'steppe-tundra' biome are described in more detail in the QEN pages, but basically its character must be inferred indirectly from knowledge of the habitat preferences of the individual plant species that were present in this vegetation, and from related zoological and sedimentological evidence.

The terms 'steppe' and 'tundra' tend to imply a dense sward with organic-rich soils, and so in this sense the term is misleading. For convenience, the steppe-tundra can be divided into two types, a more 'steppe-like' variant and a more 'tundra-like' variant.

For the tundra-like vegetation, analogies have been drawn with a treeless vegetation that presently occurs in scattered patches on well drained south-facing hillslopes in north-eastern Siberia, although the modern-day equivalent is thought to have too dense a ground cover of vegetation (Khotinsky 1984). In terms of biomass, the nearest modern equivalent may be the rather dry types of tundra that occur in the high Arctic (E. Zelikson, pers. comm.). According to E. Zelikson (pers. comm.) who has worked on this palaeovegetation type, ground cover would have amounted to no more than about 50%, with mainly herbaceous plants but a few scattered low shrubs and occasional stunted trees in sheltered spots. Peat accumulation would have been negligible, and the soil would have had a much lower organic content than most present-day tundra. Certainly, very few peat lenses have been discovered of around last glacial maximum age, in contrast to the widespread peat and organic-rich soil deposition occurring in tundra zones today. Where organic material of last glacial age does occur, it tends to be in thin localised wedges of organic debris (leaves, sticks etc.) sorted and concentrated by stream action (Zelikson pers. comm.).

The more steppe-like variant, containing a higher proportion of steppic species, would seem if anything to have had an even sparser vegetation cover. E. Zelikson (pers. comm.) suggests that the best analogy would be with semi-desert transitional steppes that occur today at the northern fringes of the central Asian desert. Present-day values of soil carbon from this vegetation zone give values of around 10-25 tC/ha (e.g. Zinke et al. 1985). Where glacial-age pollen-bearing profiles are preserved in loess, this vegetation seems to have been growing only on partially-developed soils, poor in organic matter and certainly nothing like the rich chernozem soils that characterise the present-day moist steppes (Zelikson pers. comm.). In general, the organic matter content of these steppe-tundra palaeosols (e.g. in mid-latitude eastern Europe, which probably had a relatively high precipitation compared to regions further east at the same latitudes) less than 5 or 10% of that preserved in the interstadial-age chernozem soil horizons that occur in the same sequences (E. Zelikson pers. comm.). This may have 'thinned-out' the soil organics present over the whole 2-3m thickness of the loess accumulating during the glacial interval, in contrast to the 1m thickness of the palaeosols.

Vegetation

Storage (tC/ha) Location/type Author(s) 6 tC/ha (1.) 'tundra-like' steppe tundra 5 tC/ha (1.) 'steppe-like' steppe tundra

(1.) Estimate based on consideration of palaeoevidence and nearest apparent present analogues. See section 1 of this thesis for a more detailed discussion of the evidence and the nature of steppe-tundra.

Soils

Storage (tC/ha) Location/type Author(s) 45 tC/ha (1.) 'tundra-like' steppe tundra Zelikson et al. (pers. comm.) 55 tC/ha (2.) 'steppe-like' steppe tundra Zelikson et al. (pers. comm)

(1.) Derived from the value for dry tundra (31 tC/ha) given by Zinke et al. 1984, but giving a fairly arbitrary addition for the possibility of slightly moister conditions and higher carbon storage. Estimate based partly on consultation with palaeoecologists and palaeogeographers of the Russian Academy of Sciences at Moscow.

(2.) Taken from consideration of the low organic content and poor soil development of preserved sequences, relative to the normal interglacial and interstadial state. Also based on general advice from Russian workers (e.g. Zelikson) on the character of closest modern analogue soils of Eurasian temperate semi-desertic regions. The Russian workers suggest that in fact this value may in fact be somewhat too high. G. Jalut (University of Toulouse, pers. comm. October 1994) reports that indeed, from a thorough study (so far unpublished; submitted manuscript) of steppe soils in southern Spain it appears that indeed, carbon storage of a steppe-tundra could well have been lower than this.

Conclusions; There is a high degree of uncertainty surrounding this soil reservoir, which is unfortunate given that steppe-tundra vegetation types covered so much of the Earth's surface during glacial maxima. Some workers (e.g. Van Campo et al. 1993) have suggested that its average carbon storage was similar to present-day average for steppe or tundra, but this conflicts with the many sources of circumstantial evidence for a much sparser vegetation (see QEN pages and also sources cited above). The view which prevails here is also based on opinions expressed by G. Jalut and E. Zelikson (pers. comm.). What can be said with confidence is that its carbon storage was well below what would be the average for modern-day steppes and tundras, and that it must have approached values typical of semi-arid zones in the world today. For the 'steppe-like' steppe tundra, the values of present-day temperate desert from North America are thought to represent vegetation types too shrubby and densely vegetated to resemble the more grassy glacial vegetation. The central Asian desert fringes seem likely to offer the closest analogues (G. Jalut & E. Zelikson pers. comm.).