Di-nitrogen (N2) fixed by a group of prokaryotes (diazotrophs) is the dominant process adding bioavailable nitrogen into the ocean. Although it has been intensively studied how N2 fixation is controlled by resources (bottom-up factors), it is unclear whether the grazing (top-down control) effectively impact growth and distribution of different diazotroph groups. In this study, we evaluate this question by conducting log-log regression of diazotroph biomass onto corresponding N2 fixation rates in the global ocean. The slope of regression for Trichodesmium is ~0.8, indicating that a small portion of the increase of N2 fixation does not accumulate as its biomass. That leads to a conclusion that Trichodesmium is under a substantial top-down control, although bottom-up control still dominates. We also analyze the residuals of the regression in the North Atlantic, concluding that free trichomes of Trichodesmium is subject to stronger top-down control than its colonies. The weak correlation between the biomass and N2 fixation of unicellular cyanobacterial diazotrophs indicates that the degree of top-down control on this type of diazotrophs varies greatly. The analyses obtain unrealistic results for diatom-diazotroph assemblages due to complicate nitrogen sources of these symbioses. Our study reveals the variability of top-down control among different diazotroph groups across time and space, suggesting its importance in improving our understandings of ecology of diazotrophs and predictions of N2 fixation in biogeochemical models. Measurements of size-specific N2 fixation rates and growth rates of different diazotroph groups can be useful to more reliably analyze the top-down control on these key organisms in the global ocean.
Figure 1. Schematic diagram of determining whether organisms are under top-down or bottom-up control through regression analysis of logarithmic biomass and productivity.
Figure 2. Spatial distributions of collected diazotrophic data. Volumetric (a) and depth-integrated (c) Trichodesmium biomass using directly counted abundance data. Diazotroph biomass derived from nifH-based abundance, with the color represents the dominant group (note that some data points are overlapping spatially) (e). b, d and f N2 fixation rates paired to diazotrophic biomass in a, c and e, respectively.
Figure 3. Log-log regression of Trichodesmium biomass on its N2 fixation rates. The Trichodesmium biomass data are based on directly counted abundance. a. volumetric; b. depth-integrated data. The grey area near the regression line represents 68% (±1σ) confidence interval of the regression.
Figure 4. The residuals of the log-log regression of Fig. 3 in North Atlantic. Positive residuals indicate weaker top-down control than average and negative residuals indicate stronger top-down control. Both the analyses using volumetric (a) and depth-integrated (b) data are shown.
Figure 5. Log-log regression of nifH-based diazotroph biomass to N2 fixation rate. The data are separated into three categories according to dominant diazotroph group including Trichodesmium (red) (n=54), unicellular N2-fixing cyanobacteria (UCYN, blue) (n=30) and diatom-diazotroph assemblages (DDA, green) (n=32).