9780521861724htl 1..2

and those for Stony Creek and Sutton Stream fell within two SE, of the predicted

slope of0.75. The scaling exponents for P/B versus M (individual taxa) ranged

from0.41 to 0.18 (Fig.4.8). The exponents for Stony Creek and Sutton

Stream fell within one SE, and the exponent for Ball Creek fell within two SE,

of the predicted exponent of0.25. Only the exponent of the Ogeechee River

(0.41) departed significantly from the predicted value. Although greater vari-

ability is apparent in both the departure of the observed scaling exponents from

those expected, and the relatively low r^2 of the regression models, the pattern of

concordance of observed-to-predicted scaling exponents among communities,

using individual taxa (>37 taxa per stream), was essentially the same as that

observed for the analysis using taxa summed within size classes (seven to eight

size classes per stream).

Discussion

Annual biomass turnover rates for benthic freshwater macroinvertebrates

range over four orders of magnitude, from<1to>200 (Butler, 1982 ; Benke,

1998 ; Huryn & Wallace, 2000 ). Although annual P/Bs for taxa with relatively

0

1

2

3

4

0

1

2

3

4

0

1

2

3

4

0

1

2

3

log 4

B (mg/m 10

2 )

Ogeechee River

log 10 B = 2.38 + 0.13 log 10 M

(r^2 = 0.12, p = 0.402)

Sutton Stream

log 10 B = 1.29 + 0.29 log 10 M

(r^2 = 0.27, p = 0.230)

log 10 M (∝g)

Stony Creek

log 10 B = 1.77 + 0.13 log 10 M

(r^2 = 0.06, p = 0.590)

Ball Creek

log 10 B = 1.72 + 0.21 log 10 M

(r^2 = 0.54, p < 0.04)

01234567 01234567

01234567 01234567

Figure 4.4Log-log plots of

population biomass (B¼mg/m^2 )

against M (mg/individual) for four

stream communities (see Fig.4.2

for details). The grey line

indicates the predicted slope of

the relationship between log 10 B

and log 10 M(B/M0.25). The black

line indicates the slope derived

from least-squares regression of

the data.

62 A. D. HURYN AND A. C. BENKE