380 G. O. Batzli et al. 



soil horizon, regardless of thaw depth (Chapter 5). Furthermore, when 

 soil temperatures and thaw depth increase as a resuh of human disturb- 

 ance, nutrient availability and plant production increase (Bliss and Wein 

 1972, Challinor and Gersper 1975, Chapin and Van Cleve 1978), rather 

 than decrease as the nutrient-recovery hypothesis predicts. For these rea- 

 sons, the links between depth of thaw, nutrient availability in soil and 

 nutrient concentrations in plants that are proposed by the hypothesis 

 (steps 7-9, Figure 10-13) do not seem tenable. A more likely explanation 

 for the decline in plant phosphorus concentration observed by Schultz 

 (1964) in the years following peaks in the lemming population is that in- 

 tensive grazing and grubbing for rhizomes sharply decrease plant phos- 

 phorus reserves. Simulations suggest that plant nutrient reserves may be 

 severely depleted by grazing (Chapin 1978). The involvement of other 

 ecosystem components need not be invoked. 



Nutrient levels in forage may influence both litter size and the timing 

 of reproduction of lemmings (steps 9-14, Figure 10-13). Phosphorus, 

 calcium and nitrogen all have been implicated by a model of nutritional 

 physiology of lemmings (see Nutrition and Energetics). Apparently lem- 

 mings have adapted to low nutrient availability through high forage in- 

 take rates, low digestive efficiency of energy and selection of mosses as a 

 calcium supplement. Even if low nutrient quality of forage sometimes 

 prevents lemming population growth, it probably is only one of several 

 factors which can do so. Poor snow conditions and high weasel densities 

 may also prevent population growth during winter. Hence, forage qual- 

 ity may influence the rate of lemming population growth, but other fac- 

 tors unrelated to nutrition may be equally important. 



In 1973 the depth of thaw averaged about lO^o less than in 1972. 

 Low air temperatures caused peak standing crop of aboveground Du- 

 pontia to be 25*yo lower, but concentration of phosphorus was 200*7o 

 greater (Chapter 5, Table 5-4). These changes are similar to those ex- 

 pected during the course of a lemming cycle according to the nutrient- 

 recovery hypothesis. Yet in 1973 the lemming density was only 5 to 10 

 ha"', about half that of 1972, and the population declined during the 

 winter of 1973-74. The nutrient-recovery hypothesis predicts that it 

 should have increased. 



Our general conclusion is that the nutrient-recovery hypothesis, as 

 developed by Schultz, should be modified. Lemming activity does not 

 appear to control the nutrient concentration of forage by changing depth 

 of thaw, nor do trends in lemming populations necessarily follow trends 

 in forage quality, at least as indicated by phosphorus concentration. 

 Nevertheless, the hypothesis has been valuable because it pointed out the 

 importance of considering vegetational quality as well as quantity for 

 herbivore populations. 



The quality of available forage is difficult to evaluate. About 40 spe- 



