CARBON IN THE BIOTA 293 



the net primary production that is consumed by herbivorous animals, based on 

 estimates of percentage consumption in different ecosystem types (column 3 of 

 Table 3) applied to total net primary production. Some consumption percent- 

 ages on land have been compiled by Golley; 25 we have used values ranging from 

 1% for cultivated land and 2 to 3% for desert and tundra to 4 to 7% for forests 

 and 10 to 15% for grasslands. Higher consumption percentages are reported for 

 grasslands and savannas; we judge these values (25 to 60%) to be too high for a 

 world average of aboveground and belowground consumption in grasslands that 

 mostly lack efficient harvest by African ungulates. For aquatic communities we 

 have used values ranging downward from 40% in the open oceans 55 to 35% in 

 upwelling zones, 30% in continental-shelf waters, and 15% in algal beds and 

 estuaries (with a higher harvest of plankton and lower harvest of vascular plants). 

 Mean estimated consumption percentages are 7% in continental, and 37% in 

 marine communities. 



Data on efficiencies of food use by invertebrate animals are limited, and 

 many data on vertebrates are of doubtful applicability to populations under field 

 / conditions. Slobodkin gives a range of gross growth efficiency (new 

 protoplasm/food ingested) from 4 to 13% in Daphnia-, and a net growth 

 efficiency (new protoplasm/food assimilated) of 55 to 59%. Assimilation 

 efficiencies 1 ' (food assimilated/food ingested) for marine zooplankton are 

 widely variable with many of the values between 40 and 80%; gross growth 

 efficiencies must be lower, and for these Riley 55 suggests 16% for marine 

 zooplankton and an uncertain higher value (24%) for benthic animals. For a 

 salt-marsh grasshopper and snail, 19 ' 68 gross efficiencies were 13 and 6%, net 

 efficiencies 3 7 and 14%. Engelmann 19 obtained a gross efficiency (mortality/ 

 ingestion) of 4.2 for oribatid mites. Column 5 (Table 3), animal production, 

 is based on multiplying herbivore consumption (column 4, Table 3) by gross 

 efficiencies of 15% for marine communities, grasslands with their grazing mam- 

 mals, and desert scrub with its seed eaters, and 10% for all other continental 

 communities. We suspect the result is a high estimate of "net" secondary 

 productivity — increase in carbon-containing mass by growth and reproduction 

 of primary consumers. Productivity of animals on higher trophic levels is based 

 on harvest of part of this secondary productivity and should be of the order 

 of 10% of its amount. Our estimates give the somewhat surprising result that 

 animal secondary production is less than 1% of net primary production on 

 land and 5 to 6% in the sea. The ratios of assimilation or "gross" animal sec- 

 ondary production (growth + reproduction + respiration + organic excretion) to 

 gross primary production should not be widely different from these. 



Animal-biomass values for various communities have been compiled as a 

 basis of column 6 (Table 3). Data are available for temperate forests, 1 8 ' 52 ' 73 a 

 small tropical rain forest 46 and a mangrove swamp, 26 tundra vertebrates, 50,83 

 and marine communities; 55 and data for the grassland biome were made 

 available to us by J. K. Marshall (personal communication); we have used 

 interpretations and interpolations from these to fill out the column. The biomass 



