The Herbivore-Based Trophic System 375 



four out of five winters populations did not increase even though preda- 

 tion was low. We conclude that high predation rates are not a necessary 

 condition for the maintenance of low populations. 



In summary, predation contributes to population decHnes and may 

 be sufficient to prevent increases at low densities, but it is not sufficient 

 to account for summer declines following a peak. Furthermore, relaxa- 

 tion of winter predation will not necessarily lead to population increases. 



Nutrition 



Weber (1950a) observed many dead lemmings and devastated vege- 

 tation following the spring 1949 population peak; he therefore proposed 

 that exhaustion of food supplies and subsequent starvation caused the 

 decline. However, Thompson (1955a) noted that after the peak in 1953 

 vegetation grew rapidly, even though total primary production was only 

 half that expected with no grazing. Most of the dead lemmings that 

 Thompson found appeared to be victims of predation. He suggested that 

 lemming populations declined because of high predation rates and low 

 reproductive rates, which resulted from low vegetative cover and poor 

 forage availability. Pitelka (1957a, b) supported and expanded Thomp- 

 son's views to include the possibility of changes in forage quality as a fac- 

 tor influencing lemming reproduction. Finally, Pitelka (1964) and 

 Schultz (1964, 1969) proposed the nutrient-recovery hypothesis to ac- 

 count for the cyclic nature of lemming population dynamics. According 

 to the hypothesis the nutrient concentration in vegetation declines fol- 

 lowing a lemming high and does not increase sufficiently to support good 

 reproduction by lemmings for two or three years. The hypothesis is com- 

 plex, involving several components of the ecosystem, and it will be con- 

 sidered in detail. Schultz's descriptions contain a few gaps concerning the 

 mechanisms which drive the nutrient-recovery hypothesis, so we have 

 embellished it slightly in the following treatment and tried to make expli- 

 cit all the major causal links. 



The fundamental interactions and mechanisms of the nutrient- 

 recovery hypothesis are summarized in Figure 10-13. Intensive grazing 

 takes place during the winter buildup of the lemming population and 

 continues until the population crashes during the summer, apparently as 

 a result of the combined effects of habitat destruction and predation. At 

 snowmelt soluble nutrients released into the meltwater from urine, feces 

 and clipped vegetation are rapidly taken up by growing plants. Thus, nu- 

 trient concentrations are high in the early summer forage. Later in the 

 summer this growth becomes standing dead material, and locks up some 

 nutrients that would otherwise be available the following summer. 



In addition to releasing nutrients the lemmings' intensive grazing re- 



