The most important species of arthropods on the study area are the two species of Tipulidae. 

 All of the other values given in Table XVII represent taxonomic groupings of several to many species. 

 On the intensive study area the smaller, carnivorous crane fly larva, Pedicia hannai, was more 

 abundant than the larger, saprovorus species, Tipula carinifrons. 



All of these arthropods showed similar seasonal changes in abundance, with greater abundance 

 recorded in the early and late periods of the season than in mid-season. In the Diptera, pupation, 

 emergence of adults, and breeding all occur in July; recruitment of new larvae into the population 

 occurs soon thereafter. Larvae lost due to predation or other mortality, or due to emergence, are 

 replaced at that time. The annual cycle of abundance results from highly seasonal recruitment 

 combined with mortality distributed over the entire season. If recruitment is insufficient to replace 

 animals lost to mortality, the population shows a net decline over the season. This was the case 

 for Nematocera larvae and for Pedicia hannai in 1970. 



Even with a highly synchronous period of adult emergence the number of dipteran larvae in the 

 soil never approaches zero. In fact, the lowest numbers were approximately 50% of the greatest. 

 These data indicate life cycles of more than one season, with larvae of several generations co- 

 existing in the soil at any time. This is confirmed by data on size distribution of larvae; for ex- 

 ample, the early season size distribution of Pedicia hannai showed a clear bimodality with the 

 peaks representing larvae of different ages: 



The numbers of mites and CoUembola showed similar seasonal changes. Although the life 

 cycles of these groups have not been studied in arctic regions, it is unlikely that they have cycles 

 of longer than one year. The most common pattern for north temperate species of CoUembola is two 

 generations per year. Perhaps the simplest interpretation of these data would be that these are annual 

 species, with some reproduction occurring throughout the season but the peak of reproduction occur- 

 ring at mid-season, and peak recruitment following thereafter. The abrupt decline in abundance of 

 mites and CoUembola between mid- and late August corresponds to the onset of nightly freezing of 

 the ground surface. 



The control plots at site 1 differed from site 2 in a number of ways (Table XVIII). Pedicia 

 hannai was far less abundant, and those present were small so the biomass of this species was 

 negligible. Larvae of other nematoceran species, on the other hand, were more abundant than they 

 were on the intensive sites. Early and late in the season mites and CoUembola were present at 

 densities similar to those at the intensive site; however, the magnitude of variation seen during the 

 season was very large. Early and mid-August values were two to three times above those of early 

 June, and the late August decline was very steep. 



The Diptera complete larval development and pupation in the organic-rich, near-surface layers 

 of the soil. The adults emerge from the soil to complete the life cycle above ground. The timing 

 and extent of emergence of adult Tipulidae was measured using emergence traps and sticky-boards. 

 Two emergence traps, each covering 0.76 m^ were placed on each of the control and manipulated 

 plots. These gave a quantitative assessment of emergence, but sampled a limited area. The 

 sticky-boards (one board, 1 m x 10 cm, on each control plot) sampled a larger but undetermined 

 area, and gave an index of emergence. Thus, the two sampling methods yielded complementary re- 

 sults. 



At site 2, the first Pedicia hannai was detected on a sticky-board on 3 July, and in an 

 emergence trap on 6 July. The median dates of those captured in emergence traps were 9 July (males) 

 and 10 July (females). This sexual difference in timing of emergence has been observed in prior 



46 



