9780521861724htl 1..2

r^2 values ranging from 0.75 to 0.86. The close agreement shown for these scaling

exponents with that predicted by metabolic theory seems extraordinary when

considering the large ecological differences between streams, e.g. forest vs.

grassland; with fish vs. without fish; strong top-down control vs. weak top-

down control (cf. Huryn & Wallace,1987b; Huryn, 1998 ), and the fact that

consumers from two to three trophic levels were included in the analysis.

In contrast with the temperate stream communities used in our analysis,

however, the scaling exponents for P/B against M, and P against M for the snag

community of the subtropical Ogeechee River showed relatively large and sig-

nificant departures from predicted values. This lack of concordance between

observed and predicted scaling exponents indicates fundamental differences

between the production dynamics of the different stream communities. Similar

departures from predictions of the metabolic theory have been observed for

aquatic ecosystems with strongly size-structured trophic chains (Jennings &

Mackinson, 2003 ; Brown & Gillooly, 2003 ). For these systems, and presumably

others with a strong correlation between body size and trophic level, it has been

suggested that inefficient transfers of biomass between trophic levels may

result in a step-wise reduction in the slope of the biomass–body size relationship

(Jennings & Mackinson, 2003 ; Brown & Gillooly, 2003 ). Depending upon the

log 10 M (∝g)

log

N (individuals/m 10

2 )

Ball Creek

log 10 N = 3.01 – 0.66 log 10 M

(r^2 = 0.45, p < 0.0001)

Ogeechee River

log 10 N = 4.06 – 0.79 log 10 M

(r^2 = 0.40, p < 0.0001)

Sutton Stream

log 10 N = 3.09 – 0.53 log 10 M

(r^2 = 0.29, p < 0.001)

Stony Creek

log 10 N = 2.92 – 0.52 log 10 M

(r^2 = 0.22, p < 0.003)

–1

0

1

2

3

4

5

–1

0

1

2

3

4

5

–1

0

1

2

3

4

5

–1

0

1

2

3

4

5

012 3 4 5 6 7 012 3 4 5 6 7

012 3 4 5 6 7 012 3 4 5 6 7

Figure 4.7Log-log plots of

population density

(N¼individuals/m^2 ) against M

(individual taxa,mg/individual)

for four stream communities

(see Fig.4.2for details). The grey

line indicates the predicted

slope of the relationship

between log 10 N and log 10 M

(N/M0.75). The black line

indicates the slope derived from

least-squares regression of the

data.

BIOMASS TURNOVER AND BODY SIZE 65