Carbon and Nutrient Budgets 467 



plant roots and rhizomes has been variously estimated as 6897o (Billings et 

 al. 1977 using gas exchange), 30 to 70*^0 (Bunnell and Scoullar 1975 using 

 computer simulations), and 31% (Chapter 9 from litter bag weight loss). 

 Belowground plant respiration may constitute a proportionately larger 

 CO2 source in tundra than in ecosystems of warmer climates (Coyne and 

 Kelley 1975, 1978, Billings et al. 1977). For example, roots contribute 17 

 to 30*^0 of soil respiration in a tallgrass prairie (Herman 1977, Redmann 

 and Abouguendia 1978) and 35% in a temperate forest (Edwards and 

 Sollins 1973). Root respiration is estimated to be 10% of total soil respi- 

 ration in a tundra-like heath ecosystem (Chapman 1979). In tundra the 

 period of active decomposition is only slightly longer than the period of 

 primary production, whereas in many temperate ecosystems decomposi- 

 tion continues throughout the year, and primary production exhibits a 

 more restricted season. Thus, the short duration of the arctic summer 

 may affect decomposition even more than primary production when 

 compared with temperate ecosystems. 



The saprovore-based food web is clearly the predominant pathway 

 of energy flow in the coastal tundra ecosystem at Barrow (Figure 12-1; 

 Chapter 11), as it is at Devon Island (Whitfield 1977) and in grassland 

 systems (Woodmansee et al. 1978). Even during a lemming high at Bar- 

 row, herbivores consume only 20% of the net primary production. 



In the Barrow region, 60 to 80% of the annual ecosystem respiration 

 occurs when plants are photosynthetically active (Coyne and Kelley 

 1975). Because photosynthesis occurs 24 hours a day, much of the respir- 

 atory CO2 is fixed immediately by photosynthesis so that the net CO2 

 flux between atmosphere and canopy is small. The CO2 concentration 

 within the canopy is seldom reduced more than 5% below free atmos- 

 pheric values (Coyne and Kelley 1975). In mid-latitude forest and grass- 

 land ecosystems a larger proportion of the respiratory CO2 is released at 

 night and during the non-photosynthetic season, so that the bulk atmos- 

 phere plays a greater role as net CO2 source and sink than in arctic tundra 

 (Coyne and Kelley 1975). During the winter, particularly in April and 

 May, there is a net CO2 flux from the ground to the atmosphere (Kelley 

 et al. 1968, Coyne and Kelley 1974). 



Plant forms differ considerably in the role they play in energy flow 

 in the coastal tundra at Barrow. Although mosses constitute 40 to 70% 

 of the maximum standing crop of aboveground plant biomass, they are 

 responsible for only 6% of the carbon fixed by wet meadow vegetation. 

 Clearly mosses do not play a major role in carbon flux in the Barrow sys- 

 tem in the short term. 



Estimates of the total carbon flux for the Barrow intensive site sug- 

 gest that during the period of study this ecosystem was not in steady state 

 but fixed three times as much carbon annually as it lost in respiration of 

 nonphotosynthetic organs and organisms (Figure 12-1): 



