The world’s largest alpine ecosystem, golf course-like Cyperaceae-pastures of southeastern Tibet cover 400 000 km² and are considered to be natural. Global Change Scenarios classify subtropical high mountains as vulnerable due to self-reinforcing feedback effects of human impact. It is therefore astonishing that human impact has yet to be recognized as an integral part of palaeoecological research in Tibet. Thus palaeoecology is challenged to disentangle superimposing effects of climatic and cultural impacts.

Hypothesis: Kobresia pygmaea grazing-lawns are a synanthropic replacement of grassland and forests.

To test this hypothesis we will contrast proxies that potentially carry a human signal with proxies that are independent of human impact.

Taking the monsoon-driven preponderance of environmental changes in Tibet for granted, we subject our hypothesis to divergent results depending on whether they are based on proxies excluding the bias of superimposing climatic and cultural signals, or whether they imply the bias. In order to test our hypothesis in the light of these two positions we derived the following a priori criteria:
i) If the Kobresia ecosystem is natural, exclusion of grazing should not lead to a vegetation change. Vice versa, we would have to expect a vegetation change under grazing exclusion if this biome was grazing-driven.
ii) Under the climate scenario, natural forest in normal sites in this “alpine” biome should not reproduce successfully. The same is true for reforestation trials on normal sites. By contrast, the presence of vigorous forests and successful reforestation trials would indicate the climatic potential for forests.
iii) In a climate-driven scenario we should expect a forest pollen decline after independent proxies show a transition from the mid-Holocene optimum to the Late Holocene climatic decline. In contrast, forest pollen decline during the mid-Holocene optimum should be an indication of human impact.
iv) In an exclusively climate-driven scenario, an increase in human indicator pollen is not synchronous with decreasing forest pollen. On the other hand, a synchronous increase would be expected in a human-driven scenario.
Methods: Testing our hypothesis we performed (1) vegetation monitoring in grazing exclosures; (2) vegetation records as a baseline for the “Indicator-Species Approach”, compared with climatic data; (3) forest surveys and experimental reforestation trials; (4) and integrated pollen analyses, palaeopedological and anthracology results.

Results/Conclusions

The Kobresia pygmaea ecosystem is a grazing-driven plagioclimax, developing since 8000 years after the onset of pastoralism.