Lake Michigan bacterial production, based on [3H-methyl]thymidine (TdR) incorporation and empirically determined conversion factors (5-25 x lo9 cells nmol-I), decreased with distance from shore (-2 x over 30 km), was higher at night (1.4 x -2.2 x), and decreased with depth (w 10 x over 100 m). TdR-based growth rates were consistent with independent antibiotic-and dilution-based estimates. Population size varied little and appeared controlled by balanced growth (0.02-0.33 h-l) and grazing (0.039-o. 12 h-l). Growth correlated with temperature only below 10°C. Cell size ranged from 0.015 to 0.072 pm3. Carbon content averaged 0.154 + 0.047 pg C pm-3. Net annual carbon production was 142 g C m-2 yr-l. Summer averages were 28.9 (epilimnion), 10.4 (lo-35 m), 1.6 (hypolimnion) pg C liter' d-*, and 652 mg C m-2 d-' for the water column. Flux to microconsumers averaged 8.4 pg C liter-I d-r in the summer epilimnion.Annual areal bacterial carbon demand is met by autotrophic production only if little of the latter is lost by other means. This suggests that external loads are needed, our conversion factors are high, or autotrophic production is underestimated. Although only small adjustments of those factors will satisfy the annual balance, the summer imbalance is still too large. We suggest that temporal and spatial disequilibrium of labile organic carbon supply and bacterial use is responsible for the apparent discrepancy during summer.

Bacterial production rates were estimated for the surface waters of a station 100‐m deep in southeastern Lake Michigan during 1984. Production was calculated from incorporation of [3H‐methyl]thymidine and from empirical conversion factors determined from dilution experiments performed throughout the study. The conversion factors (with typical C.V. <40%) varied between 4.7 and 18.3 × l09 cells produced per nanomole of thymidine incorporated into ice‐cold trichloroacetic acid extracts. Our estimates yielded bacteria exponential growth rates between 0.05 and 0.24 h‐1 (C.V. typically <50%) based on the empirical conversion factors. The growth estimates are much lower (0.004–0.020 h‐1) when based on measured 47% thymidine incorporation into DNA and a theoretical conversion factor. The higher growth estimates appear more consistentwith estimated grazing losses. Carbon flux estimates are less certain, due to the possible range of bacterial carbon content and growth efficiencies, but most of the higher growth estimates imply a bacterial carbon demand higher than concurrent 14C‐based primary production measurements. This may mean that a source other than recent primary production is needed to meet this demand.

Labile organic nitrogen mineralization and the apparent degree of bacterial substrate limitation were examined to consider seasonal relationships between substrate availability and bacterial activity in Lake Michigan. Accumulation rates of ammonium nitrogen in amino acid fortified and unfortified samples of epilimnetic Lake Michigan water, incubated in the dark, provided reasonable estimates of potential and actual rates of organic nitrogen mineralization. The labile organic nitrogen demand (LOND), defined as the difference between these respective rates, provided an index of heterotrophic potential. LOND ranged from ∼1–3 ng‐atoms N liter h−1 (during May–June and November) to 3–9 ng‐atoms N liter h−1(during July–October) as compared to actual organic nitrogen mineralization rates of < 1 ng‐atom N liter h−1 in some unfortified samples. The high LOND, relative to actual turnover, observed in late summer is consistent with the hypothesis that growth rates of epilimnetic Lake Michigan bacteria are strongly limited by organic substrate during late stratification.