(Left) Ecological pyramids 



FIG. 13-8 



classes (soil animals of tropical broe 



1941); (b) dry biomass, Silver Sprint 



(c) per cent total metabolism by different tax 



(a) numbers by size 

 .aved forest, Williams 

 fream (Odum 1957a); 

 --.. _, of a hypotheti- 

 cal soil population in a meadow (Macfayden 1957); (d) gross 

 productivity, Lake Mendota (Lindeman 1942). 



2702 



5867/m2 



Bacteria 



P 480 



Sun 118,872 g-cal/cm2/yr 



a 



PYRAMID OF NUMBERS AND 

 BIOMASS 



When the total animals in a community are 

 grouped according to an arbitrary series of size 

 ranges (Elton 1927, Allee et al. 1949), there are 

 always a larger number of small individuals present 

 than large ones. Plotting these data gives a pyramid 

 of numbers. 



Pyramids of numbers arranged by trophic levels, 

 rather than size, have special interest in respect to 

 food coactions. In a bluegrass field, the number of 

 green plants at the producer level (P) vjz.s over 

 8 times the number of herbivorous invertebrates in 

 the level of primary consumers (Ci) : the number 

 of primary consumers was 2 times the number of 

 spiders, ants, and predatory beetles at the secondary 

 consumer level (C2) ; and the number of secondary 

 consumers was over 100,000 times the number of 

 birds and moles among the tertiary consumers (C3), 

 because of the great difference in size of individual 

 animals at these two levels (Odum 1959). In Michi- 

 gan, the ratio between numbers of rodents and their 

 hawk and owl predators was found to be approxi- 

 mately 1000:1 (Craighead and Craighead 1956). 



In evaluating the importance of trophic levels in 

 the community, biomass is a more important index 

 than number of individuals. Of 13 ratios of fresh 

 weights between different trophic levels (Lindeman 

 1941, Juday 1942, Turcek 1952, Birch and Clark 

 1953), the majority fall between 2:1 and 12 :1. These 

 early studies furnish no certain evidence that the ratio 

 varies between different trophic levels : this needs 

 further investigation. rMthough these are not given 

 as established limits, as far as we now know, it is ex- 

 pected that biomass ratios falling far outside these 

 will represent inadequate or incomplete sampling of 

 populations, populations out of balance, or quite spe- 

 cial situations. Thus the 23:1 ratio between P and 

 Ci trophic levels in Cedar Lake Bog in Minnesota 

 (Lindeman 1941) is correlated with large annual ac- 

 cumulations of vegetation in a very old pond where 

 winter stagnation under the ice kills off much of the 

 animal life that might otherwise consume the plant 

 material. The ratio of 22:1 in dry weight between 

 P and C levels in Fig. 13-8b is due, in part, to a 

 significant fraction of the produce level being ex- 



98 Ecological processes and dynamics 



