Abstract
Bacteria are the principal consumers of dissolved organic carbon (DOC) in the ocean and predation of bacteria makes organic carbon available to higher trophic levels. The efficiency with which bacteria convert the consumed carbon (C) into biomass (i.e., carbon growth efficiency, Y) determines their ecological as well as biogeochemical role in marine ecosystems. Yet, it is still unclear how changes in temperature will affect Y and, hence, the transfer of consumed C to higher trophic levels. Here, we experimentally investigated the effect of temperature on metabolic functions of coastal microbial communities inoculated in both nutrient-limited chemostats and nutrient-unlimited turbidostats. We inoculated chemostats and turbidostats with coastal microbial communities into seawater culture medium augmented with 20 and 100 μmol L(-1) of glucose respectively and measured CO(2) production, carbon biomass and cell abundance. Chemostats were cultured between 14 and 26°C and specific growth rates (μ) between 0.05 and 6.0 day(-1), turbidostats were cultured between 10 and 26°C with specific growth rates ranging from 28 to 62 day(-1). In chemostats under substrate limitation, which is common in the ocean, the specific respiration rate (r, day(-1)) showed no trend with temperature and was roughly proportional to μ, implying that carbon growth efficiency (Y) displayed no tendency with temperature. The response was very different in turbidostats under temperature-limited, nutrient-repleted growth, here μ increased with temperature but r decreased resulting in an increase of Y with temperature (Q(10): 2.6). Comparison of our results with data from the literature on the respiration rate and cell weight of monospecific bacteria indicates that in general the literature data behaved similar to chemostat data, showing no trend in specific respiration with temperature. We conclude that respiration rates of nutrient-limited bacteria measured at a certain temperature cannot be adjusted to different temperatures with a temperature response function similar to Q(10) or Arrhenius. However, the cellular respiration rate and carbon demand rate (both: mol C cell(-1) day(-1)) show statistically significant relations with cellular carbon content (mol C cell(-1)) in chemostats, turbidostats, and the literature data.