Phosphate Enrichment Regulates the Interplay Between Deterministic and Stochastic Processes of Bacterioplankton Community Assembly in a Subtropical Bay Impacted by Thermal Discharge


 Increasing anthropogenic activities has caused intensive environmental issues and undesirable ecological impacts on coastal bay ecosystems. Bacterioplankton play critical roles in ecological functioning of the bay ecosystems, but much remains to be learned regarding the response of bacterioplankton community assembly to heterogeneous environmental issues in the bay ecosystems and its underlying mechanisms. In this study, we used the phylogenetic framework to analyze the bacterioplankton community assembly in the subtropical Daya Bay, where the habitats are connected by water flows and tides and their trophic status are subject to intensive environmental stress induced by human activities (e.g., thermal discharge and nutrient release from aquaculture). We found bacterioplankton community compositions (BCCs) among the sampling habitats in the Daya Bay showed obvious spatial heterogeneity. The spatial distributions of BCCs were significantly shaped by phosphate concentration among the examined environmental factors. We observed that between pairwise sampling habitats, the BCC dissimilarity significantly increased with the differences in seawater phosphate concentration. Compared with stochastic processes, phosphate enrichment imposed stronger effects of environmental filtering in determining bacterioplankton community assembly in the subtropical bay, and bacterioplankton communities tended to be higher phylogenetically clustered in more phosphate-enriched habitats. In summary, we propose that phosphate is a major environmental determinant in the subtropical Daya Bay impacted by thermal discharge and regulates the interplay between deterministic and stochastic processes underlying bacterioplankton community assembly.


Introduction
In marine ecosystems, coastal bays are essentially important to human society for 47 various services they provided, such as supporting, regulating, and provisioning  In this study, we used a phylogenetic framework to study bacterioplankton 115 community assembly and the underlying determining mechanisms based on 16S 116 rRNA high-throughput sequencing collected from a set of habitats located in the Daya   We set 12 sampling sites in the Daya Bay (S1-S12, Fig. 1), and sampled the 142 seawater in August 2017, including the marine aquaculture area seriously affected by 143 human activities (Fanhe harbour aquaculture area S1; Xiaoguiwan aquaculture area S5; Dapengao aquaculture area S10), nuclear power station thermal pollution area 145 (S9), petrochemical pollution area (S6), the estuary area (S11 and S12), river inflow 146 area (S2), and site S7 that was less affected by human activities.  present in most sampled habitats but with low relative abundance (Fig. S1). We 240 observed that bacterioplankton (sub)phylum compositions at S1 and S3 stations was 241 similar, and the BCCs was mainly composed of Alphaproteobacteria (Fig. S1, Fig.   242 2e). The relative abundance of Bacteroides was much higher in S1, S3, S11, S12 than 243 the other sites, suggesting that Bacteroides was more suitable to the habitats of the top 244 and the estuary of the Daya Bay ecosystem (Fig. S1, Fig. 2b). However,

245
Cyanobacteria had the highest relative abundance at S7, S8, S9, and S10, where was 246 strongly affected by the thermal pollution from the nuclear power plants (Fig. S1, Fig.   247 2c). In contrast with other phyla, Actinobacteria were mainly found in estuary of the 248 bay (Fig. S1, Fig. 2a).

249
The average water temperature in the subtropical Daya Bay was 29.78 ± 1.14°C

250
with the highest at site S9 (32.03°C), which was strongly affected by thermal 251 pollution of the cooling system of the nuclear power stations (Fig. S2). The average 252 salinity of the Daya Bay was 28.84 ± 3.03‰, with the highest at site S11 (33.33‰) 253 and the lowest at S1 (22.78‰) (Fig. S2). The concentration of PO 4 3--P, NO 3 --N, 254 NH 4 + -N and SiO 3 2showed a decreased tendency from the top (site S1 and S3) to the 255 estuary (site S11 and S12) of the Daya Bay (Fig. S2). Among the measured 256 environmental factors, we found that seawater phosphate was the best predictor of 257 BCCs in the Daya Bay (Fig. S3). The mantel test also showed that phosphate was 258 significantly correlated with differences in BCCs across different bacterial phyla and 259 subphyla in the water bodies (Fig. 3a). Therefore, we further explored the relationship 260 between the concentration PO 4 3--P and the relative abundance of the dominant phyla.

297
In this study, we examined bacterioplankton community assembly and the 298 underlying mechanisms in the typical subtropical semi-enclosed Daya Bay ecosystem. 299 We found that the spatial pattern of BCCs was dominantly shaped by deterministic

334
Phosphorus limitation exists not only for phytoplankton, but also for bacterioplankton.

335
Compared to phytoplankton, bacterioplankton can assimilate phosphorus more 336 effectively because of their larger specific surface area, and they have a lower 337 half-saturation constant for phosphate than phytoplankton (Currie and Kalff 1984). In

359
Heterogeneity of BCCs across phosphate gradients 360 We found that the sites with higher phosphate concentration were dominated by trait, or suite of traits, that allow them to persist in a given habitat (Webb et al. 2002).

408
Our results suggest that environmental filtering may be more important than  in bacterial communities. Ecology 87: S100-S108 doi: and Gammaproteobacteria (f).