Chlorophyll-normalized isoprene production in laboratory cultures of marine microalgae and implications for global models
Chlorophyll-normalized isoprene production in laboratory cultures of marine microalgae and implications for global models
复制标题
DOI:
10.4319/lo.2013.58.4.1301
复制
发表时间:
2013-07-01
影响因子:
4.5
通讯作者:
Steinke, M.
中科院分区:
文献类型:
--
作者:
Exton, D. A.;Suggett, D. J.;Steinke, M.
We used laboratory cultures of marine microalgae to investigate the effects of growth conditions and their taxonomic position on the production of isoprene, a gas that has major effects on atmospheric chemistry and provides stress tolerance to many primary producers. Isoprene was quantified from 21 microalgal strains sampled during exponential growth, using purge-and-trap pre-concentration and gas chromatography with flame-ionization detection. Isoprene production rates varied by two orders of magnitude between strains (0.03-1.34 mu mol [g chlorophyll a](-1) h(-1)), and were positively correlated with temperature (r(2) = 0.52, p < 0.001, n = 59). Three distinct sea surface temperature (SST)-dependent relationships were found between isoprene and chlorophyll a (mu mol [g chlorophyll a](-1) h(-1)), an improvement in resolution over the single relationship used in previous models: for three polar strains grown at -1 degrees C (slope = 0.03, R-2 = 0.76, p < 0.05, n = 9), nine strains grown at 16 degrees C (slope = 0.24, R-2 = 0.43, p < 0.05, n = 27 with Dunaliella tertiolecta excluded), and eight strains grown at 26 degrees C (slope = 0.39, R-2 = 0.15, p < 0.05, n = 24). We then used a simple model that applied the SST-dependent nature of isoprene production to three representative bioregions for the growth temperatures used in this study. This approach yielded an estimate of global marine isoprene production that was 51% higher than previous attempts using an SST-independent single relationship. Taking into account the effect of temperature therefore potentially allows more precise modeling of marine isoprene production, and suggests that increasing the SST-based resolution of data beyond the three groups used here could further improve future modeling simulations.