The Vegetation: Pattern and Succession 211 



Controlled Perturbations 



Experimental alteration of selected ecosystem variables reveals their 

 relative importance in the resilience or fragility of tundra and provides 

 clues to the recovery time and thresholds beyond which the system does 

 not quickly recover. Heavy fertilization increases primary production 

 and plant nutrient concentrations initially, but subsequent increases in 

 standing dead and litter tie up nutrients and reduce light penetration and 

 photosynthesis, so that within 3 to 8 years little treatment effect upon 

 primary production can be observed (Schultz 1964, 1969). Clipping and 

 removal of all aboveground vegetation or addition of excess litter (to 

 stimulate accumulation of standing dead) alter primary production and 

 plant nutrient concentrations only slightly (Chapin 1978). Even multiple 

 defoliations of single tillers have relatively small effect upon regrowth 

 (Mattheis et al. 1976). All the above manipulations cause changes that 

 are within the normal range of conditions in lemming cycles, and the tun- 

 dra is highly resilient to these perturbations. 



The anticipated effect on the tundra of the Trans-Alaska Pipeline 

 System, which carries hot oil, led to an experiment in which a wet mea- 

 dow substrate was heated in situ. Alteration of soil temperature mark- 

 edly affects ecosystem function, but the nature of the response depends 

 upon the similarity of the perturbation to those that occur naturally. A 

 10 °C soil temperature rise for one summer month at Barrow increased 

 thaw depth, decomposition rate, nitrogen availability, plant nutrient ab- 

 sorption rates, and primary production (Chapin and Bloom 1976). Ten 

 years later little treatment effect could be observed. However, when soils 

 were heated for one full year, the increased thaw depth caused melting of 

 ice in permafrost, subsidence of the ground surface and ponding of 

 water. Rapid decomposition depleted soil oxygen and soils became an- 

 aerobic, killing all vegetation within one year. The site did not recolonize 

 during the 10 years following the experiment. Thus, although temporary 

 summer changes in soil temperature stimulate nutrient cycling and 

 primary production, a chronic year-round soil temperature change in ice- 

 rich soils leads to ponding of water, death of the vegetation and a long- 

 term change in ecosystem function. Experimental heating of a relatively 

 ice-free soil in interior Alaska throughout the year caused no subsidence 

 and increased primary production 3- to 5-fold (McCown 1973), an effect 

 comparable to that caused by temporary heating of the ice-rich soil at 

 Barrow. Thus the detrimental effects of soil heating appear to be caused 

 primarily by a perturbation in excess of any natural fluctuation which 

 triggers a chain of circumstances associated with melting of ice, soil sub- 

 sidence, and change in soil chemical and physical environment. 



