This study relies on the shared voyage of the WTP commissioned by the National Natural Science Foundation of China. Physical, biological, chemical, and geological surveys were carried out on the R/V “Kexue” from October to December 2019. The survey included 45 stations, divided into four sections (Figure 1). Section A comprised stations A1-A9 (8°-14°N, 129°E), section B included stations B1-B18 (1°-20°N, 141°E), section C consisted of stations C1-C10 (0°, 142°-160°E), and section D comprised stations D1-D8 (2°S-4°N, 163°E). Samples were collected at depths of 5, 25, 50, 75, 100, 150, and 200m, with a total of 315 bottles collected.

CTD collected samples of the phytoplankton community structure. Samples from different water layers were placed in 1 L PE bottles, fixed with formaldehyde solution (3%), and stored in a cool place. The phytoplankton samples were shaken gently and settled in a 100 ml sedimentation column for 48 hours in the laboratory. The structure of the phytoplankton community was identified qualitatively and quantitatively under an inverted microscope (Motic AE 2000) based on Utermöhl method (Sun et al., 2002). The phytoplankton species were identified according to Jin, Isamu Y (Isamu, 1991), and Sun (Sun et al., 2002).

Dissolved inorganic nitrogen (DIN), dissolved inorganic phosphorus (DIP) and dissolved silicon (DSi) were measured by colorimetric method using Technicon AA3 Auto-Analyzer (Bran Luebbe, Germany). The DIN measured included Nitrate (NO₃-N), Nitrite (NO₂-N), and Ammonium (NH₄-N). Using the cadmium-copper column reduction method to determine NO₃-N, the limit of detection (LOD) was 0.01 μmol L^-1 (Wood et al., 1967).The naphthalene ethylenediamine method was used to determine NO₂-N with a LOD of 0.01 μmol L^-1 (Wang et al., 2022). Using the sodium salicylate method to determine NH₄-N with a LOD of 0.03 µmol L^-1 (Verdouw et al., 1978). DIP was determined as PO₄-P with a LOD of 0.02 µmol L^-1 using the phosphomolybdenum blue method (Taguchi et al., 1985). DSi was determined as SiO₃-Si. The LOD was 0.02 µmol L^-1 using the silicon-molybdenum blue method (Isshiki et al., 1991).

The carbon biomass of the cells was estimated based on the cell volume with conversion factors. A fluorescence microscope (RX50) was used to measure phytoplankton-related volume parameters at a magnification of 200× (or 400×). Each phytoplankton cell was measured 25-30 times, and the volume parameters were averaged to find the cell volume (Sun et al., 1999). Calculate the cell volume regarding Sun’s model and formula (Sun and Liu, 2003). Biomass calculation was based on Eppley (Eppley et al., 1970):

represents the single-cell volume (μm³); C represents the single-cell carbon biomass (pg).

The phytoplankton importance was calculated using the method of Sun (Sun, 2004):

where K is the importance of keystone species in the survey; B i is the total carbon biomass of species i (μg L^-1); B is the total carbon biomass of all phytoplankton (nanophytoplankton and microphytoplankton) in one survey (μg L^-1); f i is the frequency of occurrence of species i in the survey.

In this study, phytoplankton carbon biomass in the water column calculated using trapezoidal integration (Uitz et al., 2006):

where C is the average value of phytoplankton carbon biomass in water column (mg m^-3); n is the total number of layers sampled; C i is the carbon biomass of layer i (mg m^-3); D i is the sampling depth of layer i (m); D m a x represents the maximum sampling depth and D m i n represents the minimum sampling depth (m).

Calculation of the dominance of phytoplankton:

N is the total number of individuals; n i is the number of individuals of species i; f i is the frequency of occurrence of species i.

The Shannon index was calculated using the “Vegan” package (Oksanen et al., 2022) in R version 4.2.1, and significance was tested using the Wilcoxon rank-sum test. To account for the heterogeneous cell size distribution in the studied sea area, quantile regression was used to analyze cell size trends, as it provides a clearer understanding of cell volume. Quantile regressions of cell size were calculated using the “quantreg” package (Maniaci et al., 2022) in R version 4.2.1, and significance was determined using the P-test. Pearson’s correlation coefficient was used to assess the relationship between phytoplankton abundance, diversity of phytoplankton and environmental factors. Canonical Correspondence Analysis (CCA) was performed using Canoco 5.0 on the cell abundance of species and environmental factors, with both data sets being log10(x+1) transformed. The species data consisted of the abundance of the top 60 dominant species. Hierarchical Clustering was used for cluster analysis of phytoplankton, and the dissimilarity between different groups of phytoplankton was calculated using Primer 6 (6.1.12.0).

In 2019, the surface temperature distribution in the western Tropical Pacific Ocean (WTP) was characterized by high values, with a mean surface temperature of about 28°C, which is a typical feature of the western Pacific warm pool (Figures 2A-D). The water temperature decreased unevenly with increasing depth, and a sharp thermocline was observed at around 100 m at all sections. Salinity also increased unevenly with depth, due to the presence of North Pacific Tropical Water and South Pacific Tropical Water at the thermocline (Figures 2E–H). In section B (2°-15°N), both isothermals and isohalines were elevated. The subsurface water temperature and salinity were higher in sections C and D near the equator, owing to the warm pool and saline South Pacific Tropical Water (Figures 2C, D, G, H). Surface nutrients were deficient and only gradually increased at depths of 100 m (Figure 3). Similar to the vertical distribution of temperature and salinity, nutrients had a bump in the 2°-15°N section of section B (Figures 3B, F, J)

This investigation identified a total of 260 species from 4 phyla: Bacillariophyta, Dinophyta, Cyanophyta, and Chrysophyta, across 64 genera (Table S1). Diatoms were represented by 33 genera and 73 species, dinoflagellates by 28 genera and 181 species, cyanobacteria by 2 genera and 4 species, and chrysophyceae by 1 genus and 2 species. Dinoflagellates had the largest number of species, accounting for 69.92% of the total species with a cell abundance of 9698 cells L^-1. Diatoms made up 28.08% of the species with a cell abundance of 9026 cells L^-1. Cyanobacteria had the highest cell abundance of 121,398 cells L^-1, but their species accounted for only 1.54% of the total. The smallest proportion of both species and cell abundance was from chrysophyceae, at 0.77% and 896 cells L^-1, respectively. The top ten dominant species identified in the investigation included four dinoflagellates, three cyanobacteria, and three diatoms (Table 1).

Overall, the distribution of phytoplankton varied significantly among the four sections (Figure 4A). Dinoflagellates were found to be evenly distributed across all four sections. Section B had the highest proportion of cyanobacteria, while the proportion of diatoms sharply increased in sections B9-B15. Sections C and D had a similar phytoplankton composition, with higher levels of diatoms near land in section C. The dominant species of phytoplankton also differed across the sections (Figure 4B). Trichodesmium thiebautii was the dominant species in stations A1, A2, A8, and A9, while Scrippsiella trochoidea, Nitzschia sp., and Thalassiothrix longissima were the dominant species in other stations. Section B was mainly dominated by Trichodesmium thiebautii, Trichodesmium hildebrandtii, and Trichodesmium erythraeum, except for stations B9-B12 and B14, which had high levels of diatoms such as Nitzschia sp. and Synedra sp. Sections C and D had a similar phytoplankton composition, with dominant species including Synedra sp., Scrippsiella trochoidea, Nitzschia sp., Thalassiothrix longissima, Dictyocha fibula and Coscinodiscus granii.

For the diversity of phytoplankton, the Shannon diversity index of sites B1-B8 is lower, while sections C and D near the equator have higher diversity (Figure 5). Therefore, the difference in planktonic plant diversity between section B and sections C and D is very significant, while the difference relative to section A is relatively small.

When cell abundance was regressed against equivalent sphere diameter (ESD) using all data, a high degree of significance was found (p-value< 0.01), with 38.1% of the change in abundance being explained by ESD. The slope of the fitted straight line of cell abundance versus ESD increased with increasing cell volume, indicating that the decrease in cell abundance rate is faster when the cells are larger (Figure 6A). However, the slope hardly changes when the quantile is 0.01-0.3 (Figure 6C), suggesting that cell abundance is high and does not vary much in this ESD range of approximately 2 to 12 μm. At a quantile of approximately 0.3, the slope decreases, indicating that the abundance of cells with ESD greater than 12 μm begins to decrease. At quantile 0.8, the slope decreases again, and the ESD is about 25 μm, indicating that the abundance of cells with ESD greater than 25 μm is small, and the abundance of cells becomes smaller as the cell volume increases.

Along sections A and B, cell abundance was dominated by cyanobacteria, which were present only in the surface layer (Figures 7C, O). In section A, cell abundance distribution was similar for diatoms and dinoflagellates, but diatoms distributed in deeper layers (Figures 7A, B). The cell abundance of chrysophyceae was small and mainly distributed in the subsurface layer (Figure 7D). Compared to microphytoplankton, nanophytoplankton had a giant cell abundance and was mainly distributed in the upper layers, while microphytoplankton was mainly distributed in the lower layers (Figures 7E, F). Notably, cyanobacteria were more widespread in section B (Figure 7O) and phytoplankton were distributed in deeper layers compared to section A (Figures 7Q, R).

The cell abundance and carbon biomass of sections C and D are shown in Figure 8. In sections C and D, the phytoplankton abundance was dominated by dinoflagellates and diatoms, with little presence of cyanobacteria (Figures 8A–C, M–O). Dinoflagellates were primarily concentrated in the surface layer (Figure 8B), while the abundance of diatoms and chrysophyceae increased in section C, especially near land (Figures 8A, D). The cell abundance of microphytoplankton was greatest in section C, mainly distributed around 75m, and the cell abundance in section D was the smallest among all sections (Figures 8Q, R). The cell abundance of diatoms in section D decreased and was almost exclusively distributed in the subsurface layer (Figure 8A).

The carbon biomass in the survey area ranged from 6.597 mg m^-3 to 155.627 mg m^-3, with a mean value of 25.969 ± 24.752 mg m^-3 (Table S1). The carbon biomass of different species and cell sizes of phytoplankton were examined separately (Figures 9A, B). In terms of phytoplankton species, the carbon biomass of dinoflagellates, diatoms, cyanobacteria, and chrysophyceae was 674.266 mg m^-3, 338.329 mg m^-3, 125.266 mg m^-3, and 30.758 mg m^-3, respectively, accounting for 57.698%, 28.951%, 10.719%, and 2.632%, respectively (Figure 9A). Dinoflagellates contributed more than 50% of the carbon in each section. The carbon contribution of diatoms increased in sections C and D, with an abrupt increase at stations B9 to B17. The carbon contributed by cyanobacteria was mainly concentrated in stations A8, B1-B7, and B16. In terms of phytoplankton size, the total carbon biomass of nanophytoplankton and microphytoplankton was 376.324 mg m^-3 and 792.295 mg m^-3, respectively, accounting for 32.202% and 67.798% of the total carbon biomass (Figure 9B). The carbon contribution of microphytoplankton exceeded 60% in each section.

Analysis of the vertical distribution of carbon biomass revealed that diatoms contributed the most to the carbon biomass in sections B and C, and were primarily distributed in the subsurface layer (Figures 7S, 8G). In contrast, dinoflagellates and cyanobacteria showed higher carbon biomass in the upper layer (Figures 7H, T, I, U, 8H, T, I, U). Nanophytoplankton contributed more carbon biomass in sections A and B, and were primarily found in the surface layer (Figures 7K, W). The carbon biomass of microphytoplankton was highest in section B, and mainly concentrated in the subsurface layer (Figures 7X).

To better understand the role of low abundance and high volume cells in the ecosystem, this study utilized a combination of phytoplankton carbon biomass and dominance calculations to identify key species in the WTP phytoplankton community (Table 2). The importance of each species to carbon biomass was used to determine the keystone species, with dinoflagellates (3 species) and diatoms (2 species) being the primary contributors, followed by cyanobacteria (1 species). Scrippsiella trochoidea was frequently observed and had the highest importance among the identified keystone species, while Coscinodiscus granii had the highest single-cell carbon biomass (Table 2). Trichodesmium thiebautii was also considered a keystone species due to its high abundance despite its minimal single-cell carbon amount.

The carbon biomass of dinoflagellates and cyanobacteria peaked at depths of 5-50 m, with Trichodesmium thiebauti and Ceratium pulchellum having the highest mean carbon biomass at 5 m, at 0.430 μg L^-1 and 0.583 μg L^-1, respectively (Figures S2D, F). The peak carbon biomass of Gymnodinium lohmanni was observed at 50 m, with 0.395 μg L^-1 (Figure S2C). In contrast, diatoms had a deeper distribution, with the peak carbon biomass of Coscinodiscus granii at 100 m (0.280 μg L^-1) (Figure S2B), and Synedra sp. at 150 m (0.359 μg L^-1) (Figure S2E).

Cluster analysis was employed to classify phytoplankton into four groups. The results showed that group 1 and group 2 were highly similar, while group 3 and group 4 were also highly similar (Figure 10). Compared to group 2, group 1 had a higher diatom abundance, with Nitzschia sp. and Thalassiothrix longissima having the highest abundance among the four groups (Table 3). In group 2, the abundance of Trichodesmium thiebautii was the highest among the four groups, but Trichodesmium hildebrandtii and Trichodesmium erythraeum were not observed. Group 3 exhibited distinct characteristics of low temperature and low salinity, with the highest diatom abundance. Synedra sp. and Thalassiothrix longissima also had higher abundance in group 3.

The Pearson correlation reflects the relationship between phytoplankton and environmental factors (Figure S3). Cyanobacteria were highly significantly and positively correlated with temperature, suggesting that temperature may affect their distribution. Dinoflagellate abundance was highly significantly and positively correlated with temperature and negatively correlated with nutrition, reflecting that the growth of dinoflagellates is susceptible to temperature. At the same time, dinoflagellates may can survive better under nutrient-deficient conditions. Diatom abundance was positively correlated with DIN and DIP, indicating that nutrients can influence the distribution of diatoms. Microphytoplankton was extremely significantly and positively correlated with diatoms reflecting that it is mainly composed of diatoms. Several diatom species, such as Thalassiothrix longissimi and Synedra sp., were positively correlated with nitrate and negatively correlated with temperature (Figure 11). In contrast, the majority of cyanobacteria and dinoflagellate species showed a negative correlation with nitrate and a positive correlation with temperature, such as Trichodesmium thiebautii and Scrippsiella trochoidea.

We used cluster analysis to divide phytoplankton into four groups, and the differences in phytoplankton composition between stations reflected changes in ocean currents, which profoundly affected the diversity and distribution of phytoplankton. Although the changes in phytoplankton composition cannot be strictly distinguished by ocean currents, the variation in phytoplankton at different sections still showed certain patterns. For instance, in section A, the phytoplankton composition underwent a transition from Group 1 to Group 3 and then back to Group 1, respectively, which confirms that this section was impacted by the influence of distinct water masses. Kuroshio, known for its high temperature and low nutrient levels, provides suitable conditions for the growth of cyanobacteria, resulting in an increase in cyanobacteria abundance at stations A1 and A2 (Figures 4A, S3). Although station A9 is also classified under Group 1, it may be affected by the Mindanao Eddy, which brings up nutrient-rich cold water to the surface within the eddy. This causes the thermocline depth to become shallower, promoting the growth of diatoms, according to Zhang’s study (Zhang Q. et al., 2012). Unlike Group 1, Group 3 had the lowest temperature and salinity and higher nitrate concentration (Table 3). This may be due to the current-seamount interaction generated by the North Equatorial Current encountering seamounts at stations B9-B12 (Figures 1–3), consistent with the research by Ma (Ma et al., 2019). This effect creates a unique phytoplankton community structure in Group 3, which is dominated by diatoms that prefer low temperature and high nutrients (Figures 4B, 11). Except for the stations affected by current-seamount interaction, the phytoplankton in section B was dominated by cyanobacteria, with low species diversity, possibly due to minimal ocean current influence and distance from the land (Figure 1). In contrast to section B, the species diversity in sections C and D was high (Figure 5). On the one hand, the thermocline in the eastern equatorial Pacific is much lower than that in the western Tropical Pacific Ocean, and nutrient-rich seawater is transported by the South Equatorial Current to the Gulf of Papua, and nutrient levels are increasing due to the influence of the New Guinea continent (Gordon and Fine, 1996; Mackey et al., 2002). On the other hand, the New Guinea coastal upwelling, which flows southeast to northwest year-round, also carries high salinity and nutrient-rich South Pacific water northward along the coast (Christian et al., 2004). Therefore, the waters outside the New Guinea Islands have higher species diversity and diatom biomass (Figure 5, 9A), consistent with the research by Dong (Dong et al., 2012).

Different dominant species exhibit specific environmental preferences and can significantly influence the characteristics of entire ecological communities. Our study employs both cluster analysis and CCA analysis to identify two distinct types of phytoplankton: the high-temperature low-nitrate type, represented by Trichodesmium and Scrippsiella trochoidea, and the low-temperature high-nitrate type, represented by Thalassiothrix longissima and Synedra sp. (Figures 10, 11). Trichodesmium, a crucial nitrogen-fixing alga in the ocean, contributes substantially to primary production and biomass in our study area. However, factors such as wind, phosphorus, and unique metal characteristics can significantly impact the abundance of Trichodesmium (Chang et al., 2000; Nuester et al., 2012). Our survey has indicated that Trichodesmium erythraeum and Trichodesmium hildebrandtii solely appeared in Group 2, suggesting their affinity for living in undisturbed oligotrophic water bodies. Chen’s research also supports our findings, demonstrating that Trichodesmium thiebautii primarily dominates areas outside the New Guinea Islands with large ocean currents. Conversely, Trichodesmium erythraeum and Trichodesmium hildebrandtii have a greater abundance in sections unaffected by ocean currents (5°-35°N, 145°E). Scrippsiella trochoidea, a dominant and key species in the western Tropical Pacific Ocean, exhibits widespread distribution and high abundance across four sections, ranging from northern to tropical waters. This species has been reported to bloom in various regions, such as Japan, the Mediterranean, and the US coast (Montresor et al., 1998; Zinssmeister et al., 2011; Morozova et al., 2016). Our survey has also revealed that Scrippsiella trochoidea can maintain high density even in low-nutrient conditions, corroborating existing literature (Yin et al., 2008).

The dominant diatom species identified in this survey mostly belong to nanophytoplankton. The small cell size of these diatoms enables them to remain suspended and to rapidly absorb nutrients. Some diatom species have become elongated or flattened to increase their surface area-to-volume ratio for better nutrient absorption (Reynolds, 2006), such as Thalassiothrix longissima and Synedra sp. (Figure S1). Ceratium biceps is an important species that distinguishes Group 3 from Group 4 (Table S4). It prefers high-salinity conditions of Group 4 (Table 3), which is consistent with the results of CCA analysis (Figure 11). In the continental shelf area close to land, the abundance of Ceratium biceps is higher, which is consistent with the findings (Hallegraeff and Jeffrey, 1984).