The Xiamen Symposium on Marine Environmental Sciences (XMAS 2027)

How temperature differentially regulates bacterial production (BP) and respiration (BR)—and the resulting effects on bacterial growth efficiency (BGE)—remains poorly resolved in subtropical coastal systems, where BGE–temperature relationships diverge markedly from temperate and high-latitude patterns. Here, we combined 14 months of in situ observations with short-term temperature-controlled incubation experiments across a resource gradient in Dongshan Bay (southeastern China) to isolate direct physiological effects of temperature on bacterial carbon allocation. Both BP and BR increased with temperature, yet exhibited a consistent asymmetry: BP showed a near-linear response, whereas BR followed a unimodal thermal performance curve saturating at higher temperatures. This divergence produced a systematic increase in BGE with warming—opposing the classical prediction—underpinned by activation energies showing E_a (BP) consistently exceeded E_a (BR) across stations (bay: 0.67 vs. 0.37 eV; offshore: 0.81 vs. 0.31 eV). BGE temperature sensitivity (Q₁₀) was significantly higher at the offshore station (3.34) than the bay station (2.14), consistent with stronger E_a divergence under lower resource availability. Despite substantial differences in DOC, nutrients, and DOM composition between stations, resource availability modulated the magnitude of metabolic rates without altering the fundamental E_a (BP) > E_a (BR) ordering—confirming that resource conditions restructure metabolic intensity rather than response rules. Synthesis of a global BGE dataset (n = 394, four thermal zones) further revealed that BGE–temperature relationships are directionally opposed across climatic zones: subtropical systems show a significant positive relationship (R² = 0.59), while temperate systems show a negative one (R² = 0.17). These findings demonstrate that warming in subtropical coastal systems preferentially enhances biomass production over respiratory carbon loss, and that applying globally averaged BGE–temperature parameterizations to low-latitude systems will systematically overestimate microbial CO₂ release—with direct implications for Earth system model predictions of subtropical marine carbon cycling.