However, as different patterns of growth and mortality select different patterns
of adult size and mass-specific metabolic rate, a variety of interspecific allomet-
ric slopes with varying amounts of scatter can result, depending on the species
included in the comparison (Daan & Tinbergen, 1997 ). Further implications of
this kind of reasoning for interspecific scaling of metabolism are discussed by
Kozłowski, Konarzewski and Gawelczyk ( 2003 ).
However, it may be inappropriate for ecologists to use interspecific allo-
metries based just on adults to predict rates such as growth or mortality within
ecological functional groups, as they ignore the majority of organisms –
juveniles. Instead, allometric relationships should incorporate all relevant size
classes of the different species in the community.Ecological applications
Estimating global patterns of mortality
Here, scaling relationships and life-history analysis are applied to deducing the
value of a difficult-to-measure life-history and demographic trait (mortality)
from data on other life-history and demographic variables (fecundity, develop-
ment period, reproductive rate). A specific example is the estimation of global
rates and patterns of mortality in marine epipelagic copepods, where scaling
with body size and temperature can be examined (Hirst & Kiørboe, 2002 ).
Mortality is especially difficult to measure in the marine environment, which
is physically dynamic and where planktonic populations cannot be easily fol-
lowed through time. Consequently, mortality measurements in this group are
relatively scarce, and confined to a narrow range of taxa, body sizes or environ-
ment types, which limit the ability to estimate the dependence of mortality
on temperature and body mass. Therefore, Hirst and Kiørboe ( 2002 ) predicted
mortality rates from data on fecundity rates, egg-hatch times, egg-to-adult
times and sex ratios, all of which were available from more diverse sets of
species, environments, body sizes and temperatures. Their initial assumptions
(justified below) were that, on average, the populations were in steady-state, the
mortality rate (d^1 ) was age independent, and egg-production rate was con-
stant for adult females irrespective of their age. With these assumptions the net
reproductive rate, which is the number of offspring per female that survive until
the next generation (R 0 ), is derived from the rate of mortality, the development
time (D, days¼time from being laid as an egg to moulting into adulthood), and
the egg-production rate (m, eggs female^1 d^1 ) (Kiørboe & Sabatini, 1994 ):R 0 ¼ðm= Þe^ D ( 3 : 2 )
Thus, for the population to be maintained,R 0 has to equalSþ1, whereSis the
ratio of adult males to females, as at steady-state each female must replace
herself and produce the appropriate number of males given the sex ratio. The
assumption of steady-state, averaged globally, must be approximately true46 D. ATKINSON AND A. G. HIRST