Secondary Production and Consumer Energetics - Fundamentals of Ecosystem Science - page 60

Models to predict production of consumer species without using mean annual biomass
as an independent variable have not been developed, and probably would be difficult to
build and be subject to huge errors. To see why, note that Eq. (3.8) can be expanded by
replacing B with NM . Thus,

P 5 ð P = B Þ NM ð 3 : 9 Þ

where M usually is easily measured, and existing models ( Banse and Mosher 1980;
Benke 1993 ) can be used to predict P/B (with moderate to large error) from temperature
and body mass. Consumer density ( N ), however, varies over many orders of magnitude
and is at this point unpredictable. As a glance at any general ecology text will show, N
(the density or size of a population) is controlled by a wide range of ecological factors.
The only general model to predict animal density (from body size; Peters 1983 ) is far too
imprecise (the range in densities at a single body mass covers up to eight orders of magni-
tude!) to be used to predict density. As it seems unlikely that ecology will soon come up
with a satisfactory general model to predict abundance of individual consumer species,
I doubt that ecologists will be able to develop a useful model to predict production of
consumer species that does not include mean biomass as a predictor variable.

PRODUCTION OF A GUILD OF CONSUMERS

Studies of both animals and bacteria ( Kajak et al. 1980; Cole et al. 1988; McNaughton
et al. 1989, 1991; Wallace et al. 1999 ; cf. also Cyr and Pace 1993; Cebrian 1999 ) have shown
that the production of a functional group of consumers is correlated with the supply rate
of its food ( Figure 3.5 ). This conclusion holds whether food levels are experimentally
manipulated ( Figure 3.6 ) or vary naturally across a series of ecosystems. These results sug-
gest that production rates of guilds of consumers are often constrained by food supply.
Nevertheless, the considerable scatter around the published regressions suggests that
factors other than food supply also exert important control over production at the level
of guilds of consumers. For example, there is increasing evidence (e.g., Sterner and
Elser 2002 ) that bacterial and animal production may often be limited by nutrients
(i.e., phosphorus, in lakes) rather than energy (Box 3.1), so that the quality as well as
amount of primary production may set secondary production of consumer guilds
( Figure 3.7 ). In any case, the scatter around regressions like those in Figure 3.5 is too great
for food supply to be a practical predictor of production of consumer guilds.
In addition to the rate of food supply, whether the consumers are homeotherms or poi-
kilotherms affects the production of the guild, with homeotherms having markedly lower
rates of production ( Figure 3.5 ; McNaughton et al. 1989, 1991 ), as expected.
Because of the scatter in the relationships between food supply and consumer guild
production, it is difficult to estimate precisely the slopes of these relationships.
Nevertheless, reduced major axis regression slopes of most of the published log-log regres-
sions are between 1 and 1.5, suggesting that the production of a guild of consumers rises
approximately linearly (or perhaps a little more steeply) with increasing food supply
( Table 3.3 ). Thus, consumer production may be a constant or rising fraction of organic
inputs as organic inputs to an ecosystem increase.