Articolo in Italiano

Let me provide a different perspective with a quote from our paper “Key ecological parameters of immotile versus locomotive life”:

“According to the paleontological evidence, an inherent feature of the ecological community of the mammoth steppe was its spatial and temporal instability [42]. The megafauna of the mammoth steppe underwent repeated regional extinctions after
which it recovered at the expense of migrations from distant refugia. … Such spatial and temporal dynamics is consistent with
the proposition that due to its powerful destabilizing impact on the regional environment and climate the megafauna disrupted conditions favorable for its own existence. Then the megafauna got extinct and could re-colonize the same territory only after the ecosystem has recovered in the course of succession in the megafauna’s absence.

Despite the paucity and incompleteness of the data about environmental characteristics of the mammoth steppe, possible mechanisms of how the megafauna could destabilize the environment can be easily outlined. For the first, the grass layer sustained by the megafauna is not able to efficiently store moisture in soil as compared to the forest (or tundra in higher latitudes); thus the grass ecosystem cannot function as the biotic pump of atmospheric moisture [43]. Besides soil moisture
control, big herbivores can change the vegetation species composition such that grasses using C4 photosynthesis might begin to dominate (for example, some Artemisia species [44, 45]), which have a lower transpiration rate per unit carbon mass fixed than the С3 plants. Low evaporation reduces the intensity of the biotic pump. In the absence of the biotic pump precipitation on land is scarce and irregular; droughts and floods are frequent that may cause a decline in primary productivity and disruption of the food resources.

For the second, according to the available data from long-term experiments, an elevated fertilization-caused productivity of high latitude ecosystems (in particular, tundra) leads to a rapid depletion of soil organic carbon [46]. This implies that the high productivity of grasslands that is necessary to feed dense populations of big animals could be of transient nature and accompanied by soil degradation, depletion of food resources and extinction of large animals after which the successional
recovery of the ecosystem began. Notably, neither the biotic pump nor soil degradation effects have been considered in mammoth steppe studies [13].”

Megafauna on land is, generally, a big nuisance for ecosystem stability. Humans are just another (big) point in the row.

Mark, you are right in many ways, but climate change played a decisive role in the Mammoth tundra-steppe degradation, bat not human population growth. The tundra-steppe existed until trees could grow massively in conditions of a cold climate and lack of precipitation. The growth of trees in a continental climate was hindered by a soil permafrost and a lack of precipitation. A small snow layer did not interfere with the mega phytophages nutrition in winter.
As the climate warmed, precipitation increased, the snow layer increased in winter, and it became more difficult for phytophagous species adapted to grass and shrub vegetation to feed. There is little grass under the forest canopy, and many species were not adapted to eating coarse twig food in winter.
The habitats of some species adapted to the grass vegetation conditions (for example, in Eurasia, these are horses, saigas) have been reduced to the territory of the modern steppe and forest-steppe, where less precipitation falls in a continental climate. During the formation of the broad–leaved forests zone, horses, bison, deer and moose were preserved in it, in the taiga zone – only deer and moose, whose digestive systems are adapted to feeding during the winter only with coarse twig food.
In the taiga zone, moose and deer play a completely different role than ruminants in grasslands ecosystems. For moose and deer, grasses and early successional deciduous trees and shrubs species are more preferable as food than coniferous species, and moose and deer accelerate succession by eating early successional species, accelerate the change of these species to conifers, which are not only themselves less edible for animals, but also conifers litter decomposes worse than the early successional species litter.
In the massively humans populated broad-leaved forests zone, horses and bison were completely exterminated by humans in the XIX-XX centuries, horses and saigas were almost completely exterminated in the steppes (efforts are currently being made to restore populations of these species), deer and moose were preserved in the less humans populated the taiga zone.
Some human-domesticated phytophagous species are now distributed much further north than their wild ancestors. For example, in the taiga zone in the Arkhangelsk region there are several small populations of feral horses. There are no the existence signs of wild horses in those places during the Holocene period.
Thus, grasslands symbiotic systems where the animals that live on them are vital for their health, exist only under a certain combination of climatic conditions, and other ecosystems are formed under a different climate. Taiga ecosystems, indeed, are less balanced than grasslands ecosystems, but with the taiga temperature and precipitation ratio characteristic, grasslands ecosystems are impossible.

Mihail