The impact of feedbacks between vegetation and other climate systems components is and emerging area of climate science indeed recognised by the IPCC as an important consideration in future climate modelling. Claussen (2009) reviews the effect of vegetation-climate feedbacks on the Earth System and reviews the evidence for vegetation being a secondary amplifying factor in orbital forcing.
Summary of paper
The review sets the context on vegetation dynamics by briefly reviewing the vegetation-albedo negative feedback of Daisyworld (Lovelock and Watson, 1983). On Daisyworld, vegetation mediates global temperatures for the range of insolation when the planet is conducive to organic life, as we know it. Daisyworld is a climate model originating from a thought experiment about the emerging properties of the Earth from its constituent climate system components. The author focuses on the late Quaternery, in particular the geologically recent glacial-interglacial cycles of the Holocene in search of validation of the Daisyworld model.
Biogeophysical and biogeochemical feedbacks
Ignoring the anthropogenic contribution to climate change over the period in question, the author uses biogeophysical and biogeochemical feedbacks as a central them. The interdisciplinary names of the feedbacks themselves imply a coupling between different spheres. Biogeophysical feedbacks are energy, radiation and water and directly affect near-surface energy, moisture, and momentum fluxes via changes in surface structure and plant physiology. Biogeophysical first came to the attention of climate scientists in the 1970s in work led by Charney et al (1975) in their work on the understanding the extent of the Sahara desert over time when studying the general circulation model. Biogeochemical feedbacks affect the chemical composition of the atmosphere and so could be characterised as the atmosphere-biosphere coupling, although at its heart is the carbon cycle (#ref).
Evidence from ecosystems that can survive in subarctic regions
Given increasing insolation and the orbital forcing of the Milankovitch cycles, the northern limit of the tree-line would be expected to vary but to move northwards during the Quaternary. In other words, palaeobiological evidence for the extent of the northern tree-line and pollen counts should support cooler mid-Holocene winters. This is not supported by climate models that ignore vegetation dynamics, although by tweaking the sea-ice climate feedback (itself consisting of two positive feedback mechanisms; sea-ice albedo and sea-ice heat flux), some parallel between models and palaeobiological data can be discerned. Berger (2001) hypothesises vegetation-snow albedo feedback as the cause. With such conflicting results, Claussen suggests that it is the cycle of biogeophysical feedback in summers and winters that may explain the discrepancies. He goes back in geological time to the Eemian to test the effect of biogeophysical feedback but finds no conclusive proof.
Evidence from the subtropical deserts
Palaeobotanic reconstructions suggest that the Sahara region was warmer and greener region in the mid-Holocene than it is today. A desert-albedo feedback mechanism has been used to explain a dampening of the decrease in precipitation and the opposite, an amplification. Subtropical heat has more of an impact on atmospherics than boreal biomes, and links are made with monsoons and global equatorial atmospheric circulations. Claussen makes the point that like the daisies on Daisyworld, vegetation dynamics tend to maintain the stable equilibrium for longer than would otherwise be, but that once a tipping point was reached, abrupt climate change follows. Like pollen counts for boreal biomes, the global distribution of desert dust by Aeolian processes informs the debate. All the evidence does indicate a desertification of the Sahara in the mid-Holocence about 5000bp, but the rate of desertification varies greatly between models. Looking back in geological time, Claussen finds a correlation between climate change in North Africa and Dansgaard-Oeschger events and concludes that deserts expand and retreat following precessional forcing trends, with abrupt changes in vegetation, again supporting the smoothing effects of vegetation dynamics on Daisyworld.
The role of biogeophysical and biogeochemical feedbacks in climate models remains unsolved at the time of publication of this paper. Bothe feedbacks can be positive or negative depending upon couplings, boundary conditions and the starting position (the inputs). However, there is discernible evidence for the contribution vegetation dynamics makes in amplifying orbital forcing. Modelling vegetation-snow albedo feedback in synergy with sea-ice-climate feedback yields results similar to the palaeobotancial evidence. There is more agreement in models on subtropical deserts - biogeochemical feedbacks are assumed to be negative over long time periods, but the modelling of intermediate states, especially encompassing peat bogs and wetlands.
Claussen concludes that vegetation dynamics are probably important but the science is yet young.