142 F. S. Chapin III et al. 



sequent development of "primes" may be slower, and they may take one 

 year longer to complete their life cycles. Production of prime tillers is the 

 means by which a tiller system branches, although prime tillers often do 

 not produce daughter tillers and become "dead ends" or side branches 

 of the main rhizome axis (Figure 5-1). 



After production of approximately 18 nodes, the apical meristem 

 typically differentiates into an inflorescence. The last four nodes and the 

 inflorescence normally develop in the third (V3) or fourth (V4) season of 

 aboveground growth, although induction and initiation occur at the end 

 of the preceding season (Mattheis et al. 1976). The last growing unit de- 

 velops into the flag leaf and the elongated lower internode of the culm. 

 After flowering, the shoot dies. The tiller rhizome and roots continue to 

 live for several years and may occasionally produce VO tillers (AUessio 

 and Tieszen 1975a, Shaver and Billings 1975). 



The tiller system comprises all the vegetative offshoots of the orig- 

 inal seedling or tiller, and in the tundra at Barrow consists of more than 

 20 to 30 tillers. Living tillers are often interconnected by dead rhizomes. 



Seasonal Growth Patterns 



Shoot Growth 



The general pattern of leaf turnover in Dupontia is similar to that of 

 temperate grasses (Evans et al. 1964, Langer 1966, Milthorpe and 

 Moorby 1974) and does not represent any unique adaptation to arctic 

 conditions. Dupontia produces leaves continuously through the growing 

 season, although more rapidly early in the season (Figure 5-2) (Mattheis 

 et al. 1976). Leaves that are not fully exserted by season's end lie quies- 

 cent until spring and then resume growth. Laboratory studies suggest 

 that leaf growth is largely supported by reallocation of carbohydrate and 

 nutrients from simultaneously senescing old leaves (McCown 1978), so 

 that leaf production may represent a large sink for carbohydrates and 

 nutrients only in the spring of the first year of aboveground growth. This 

 hypothesis is supported by '^C labeling studies (Allessio and Tieszen 

 1975a, 1978) and computer simulations (Miller et al. 1978c) but lacks 

 documentation of nutrient reallocation patterns. 



Flowering tillers differ from vegetative tillers in their pattern of leaf 

 turnover in that all leaves of flowering tillers senesce relatively early in 

 the season. The accelerated senescence presumably represents a develop- 

 mentally programmed redistribution of materials to reproductive struc- 

 tures, as in temperate graminoids (Williams 1955). Thus the pattern of 

 shoot growth of Dupontia may maximize reutilization of nutrients from 

 senescing leaves. 



