Introduction

Planktonic ciliates taxonomically belong to phylum Ciliophora, class Spirotrichea, subclass Oligotrichia and Choreotrichia1, and they morphologically consist of tintinnids and aloricate ciliates. Marine planktonic ciliates are important components of microzooplankton as primary consumers of pico- (0.2–2 μm) and nano- (2–20 μm) sized plankton, and important food items of metazoans and fish larvae2,3,4. Therefore, they play an important role in material circulation and energy flow from the microbial food web into the classical food chain5,6,7. Owing to their rapid growth rates and sensitivity to environmental changes, ciliates, especially tintinnids, have been considered as effective bioindicators in different water masses because of distinctive species composition8,9,10.

Diel variations, which are common phenomenon in marine plankton, include variations in abundance, behavior, physiology, feeding and cell-division11,12,13,14,15. The diel behavior of phytoplankton was found to be affected by light-dependence of cell growth and continuous losses to grazing in the tropical and subtropical seas12,16,17,18, which eventually led to community diel variations. For example, in the northern South China Sea (nSCS) at night, the abundance and cell size of picophytoplankton (Prochlorococcus, Synechococcus, and picoeukaryotes) were respectively higher and smaller than during the day18. With respect to marine planktonic zooplankton, most studies dedicated to meso-/macro-zooplankton, which have higher abundance at night than in the day owing to their diel vertical migration (descending at dawn and ascending in late afternoon and evening)19,20,21,22,23,24,25,26.

In contrast, studies related to planktonic ciliate (microzooplankton) diel variations remain limited, even though several investigations on planktonic-ciliate diel variations were conducted in different habitats27,28,29,30,31,32,33,34,35. In oceanic waters, the mixotrophic ciliate Mesodinium rubrum was shown higher abundance at surface waters at daytime than at night in the Baltic Sea30,32. Some micro-sized heterotrophic ciliates at night were more abundant at surface water than in the day in the northwestern Mediterranean Sea33. Those above two phenomenon about diel variations owing to their different diel vertical migration behaviors. But in the shelf and slope waters of the Georges Bank (northwest Atlantic)29, and the Toyama Bay (Japan Sea)31, abundance of planktonic ciliates varied little during the day and night. In the eutrophic shallow waters of a Germany gravel pit lake characterised by stable water stratification, Rossberg and Wickham34 found that the abundances of several dominant ciliate species were significantly higher in the day than at night. Despite their important role in marine microbial food webs, our knowledge of ciliate assemblage diel variations in tropical oceanic waters are limited due to their inaccessibility for oceanographic surveys.

The South China Sea is the largest semi-enclosed basin in the western Pacific Ocean36, and the tropical Western Pacific (tWP) holds the largest warm pool area with sea-surface temperature > 28 °C throughout the year37. Many studies were conducted on ciliate communities in the northern slope of the South China Sea9,S2). As for dominant species in the nSCS, S. faurei and Epiplocylis acuminata prefer 50 m and DCM layers and their high abundance occurred at night more frequently. In contrast, in the surface layer, Dadayiella ganymedes and Steenstrupiella steenstrupii were present in higher abundance in the day than at night. The P. perpusilla prefer 25 m and 50 m layers at night. But at DCM and 100 m layers, its high abundance occurred in the day more frequently (Fig. 4; Supplementary Figs. S3 and S4). In the tWP, S. faurei, P. perpusilla, Ascampbelliella armilla, Acanthostomella minutissima and Metacylis sanyahensis prefer DCM layers and their high abundance occurred at night more frequently. While for Canthariella brevis and Protorhabdonella curta, their abundance were higher in the day than at night in surface layers. With regard to Eutintinnus hasleae, its abundance was higher at night than in the day in waters ranged from 50 to 200 m (except DCM) (Fig. 4; Supplementary Figs. S3 and S4).

Figure 4
figure 4

Diel variations of average abundance of tintinnid dominant species at each layer in the nSCS and tWP.

Full size image

Body size composition and abundance proportion of tintinnid species

Average abundance in tintinnid LOD (lorica oral diameter) size-classes had some differences throughout the day and night in both the nSCS and tWP (Fig. 5). Highest species richness and average abundance were in the 28–32 μm LOD size-class during the day and night in both the nSCS and tWP. For both day and night, the second highest species richness in the nSCS and tWP were 32–36 μm and 24–28 μm LOD size-class, respectively, while the second highest average abundance were 12–16 μm and 20–24 μm LOD size-class in the nSCS and tWP, respectively. Generally, average abundance of most tintinnid LOD size-classes were higher at night than in the day. However, these night and day values were similar in the tWP (Fig. 5).

Figure 5
figure 5

Day-night variations of tintinnid species richness, lorica oral diameter (LOD), lorica length, average abundance and abundance proportion in the nSCS and tWP.

Full size image

In the nSCS, abundance proportion of S. faurei (highest, 16.8%) and D. ganymedes (second highest, 15.7%) were lower in the day than at night (18.4% and 16.1%, respectively). Abundance proportion of S. steenstrupii (third highest, 9.3%) was higher in the day than at night (5.5%). In the tWP, S. faurei (9.2%), C. brevis (8.8%) and P. curta (7.1%) had the three highest abundance proportion in the day. At night, however, species with the three highest abundance proportion changed to A. minutissima (9.5%), S. faurei (9.0%) and P. perpusilla (6.2%) (Fig. 5). Additionally, tintinnid species with lorica length greater than 150 μm had higher abundance proportion in the day than at night in both the nSCS and tWP (Fig. 5).

Relationship between ciliate abundance and environmental factors

Temperature-salinity-plankton diagrams showed that aloricate ciliate size-fractions (small, medium, and large) and tintinnid dominant species behaved within different temperature and salinity ranges that varied from day and night in the nSCS and tWP (Fig. 6). Regarding differences between the two sites, the average temperature of each aloricate ciliate size-fraction with abundance > 100 ind. L−1 in the nSCS (23.1–24.8 °C, average 24.3 ± 0.5 °C) was lower than that in the tWP (24.8–29.8 °C, average 27.8 ± 1.9 °C) (Supplementary Fig. S5). As for tintinnids, all dominant species (except D. ganymedes) in the nSCS had temperature ranges wider at night than in the day, and their higher abundance was associated with salinity higher in the day (except E. acuminata) than at night (Fig. 6). In the tWP, all dominant species (except S. faurei) corresponded to wider salinity ranges at night than in the day (Fig. 6; Supplementary Fig. S6).

Figure 6
figure 6

Temperature-salinity-plankton diagrams for day-night variations of size-fractions (10–20 μm, 20–30 μm and > 30 μm) of aloricate ciliate and tintinnid dominant species in the nSCS and tWP.

Full size image

Relationships between ciliate abundances and environmental factors (depth, temperature, salinity, and Chl a) during day and night, differed in both the nSCS and tWP (Table 1; Supplementary Table S4). In the nSCS and tWP, Aloricate ciliates and total ciliates had strong significant negative and positive correlations with depth and temperature, respectively, whether in the day or at night. As for dominant tintinnids in the nSCS, S. faurei had significant positive correlation with Chl a at night, but no correlation with Chl a at daytime. P. perpusilla had significant positive correlation with Chl a in the day, but no correlation with Chl a at night (Supplementary Table S4). In the tWP, S. faurei, P. perpusilla, M. sanyahensis and total tintinnids were not correlated with Chl a in the day, but they exhibited significant correlations at night (Supplementary Table S4).

Table 1 Partial Mantel tests comparison between ciliate community (aloricate ciliate, tintinnid, and total ciliate) and environmental factors (depth, temperature, salinity, and Chl a).
Full size table

The partial Mantel test revealed that aloricate ciliate and total ciliate had similar significant correlations with their each environmental factors between day and night in both the nSCS and tWP, respectively. But different for tintinnid. In the nSCS, except Chl a, tintinnid had significant correlation with depth at night (P < 0.05). In the tWP, tintinnid had significant correlation with temperature in the day (P < 0.05), but changed into had strong significant correlation with Chl a at night (P < 0.01) (Table 1).

Discussion

Diel vertical distribution variations of ciliate community

In oceans, there were multiple factors that could influence diel variations of ciliate (diel-vertical-migration, food items concentration and quality, predator avoidance, light intensity, body metabolic rates, etc.) in various seas27,33,34,45,46,47. Previous studies found that most planktonic ciliates do not show perceivable vertical migration29,31, thus we speculate that diel-vertical-migration might not be an determining factor for ciliate diel variations. In the oligotrophic seas, the phytoplankton assemblage was dominated by Prochlorococcus, Synechococcus and picoeukaryotes, and they showed different diel variations12,48. As important food items of ciliate, heterotrophic bacteria also displayed clear daily oscillations in the oligotrophic Ionian Sea (Mediterranean)48. Thus we speculate that diel variation of food items was possibly the main reason in determining ciliate diel variation in the oligotrophic tropical seas.

The ciliate abundance was high in surface and DCM layers in both day and night of both the nSCS and tWP. These results were similar to previous ones established in the western Pacific Ocean40,42,43,44,45,49 and eastern Indian Ocean50. However, the studies that previously investigated the ciliate vertical distribution, did not assess potential differences between day and night in vertical direction. Therefore, our study provides more accurate data on ciliate diel variation in the nSCS and tWP. Additionally, our results in the upper 200 m provide evidence that ciliate abundance were higher at night than in the day in both the nSCS and tWP (Figs. 1 and 2). Zooplankton distribution in waters mainly depends on phytoplankton presence51,52. Thus, it is possible that the availability of more food items (flagellates, picoeukaryotes, Prochlorococcus, Synechococcus and heterotrophic bacteria) at night than in the day explains the higher ciliate abundance at night18,32,48.

Diel variations in aloricate ciliate size-fractions

Abundance proportions of different aloricate ciliate size-fractions have rarely been reported in the nSCS and tWP. In the tropical Pacific Ocean, average abundance proportions of small size-fraction (10–20 μm) of aloricate ciliates to total ciliates ranged from 38 to 50% (from surface to 200 m depth), and it belonged to the dominant group at each depth in most stations43,44,49. Our results for the small size-fraction of aloricate ciliates in the tWP are consistent with those of previous studies in both day and night. In the upper 100 m of both nSCS and tWP sites, the large size-fraction (> 30 μm) of aloricate ciliates had more pronounced diel variations than those of the small size-fraction (Fig. 3). We speculated that the large size-fraction of aloricate ciliates were migrating along distances longer than those crossed by the small size-fraction. This phenomenon may be similar to that observed in meso-/macro-zooplankton in the nSCS25, equatorial Pacific Ocean53, subtropical and subarctic North Pacific Ocean54, and northwest Mediterranean55.

Potential reason for tintinnid diel variations

The LOD of a tintinnid is closely related to the size of its preferred food item (approximately 25% of the LOD)56. Our results showed that tintinnid abundance was higher but biomass was lower at night than in the day in both the nSCS and tWP (Fig. 2). We also found that abundance and abundance proportion of the 12–16 μm LOD size-class of tintinnids was higher at night than in the day. These results suggest that both LOD size-classes of tintinnids and the size of their preferred food items were smaller at night than in the day. The night-dominant smaller cell sizes of food items (picoeukaryotes, Prochlorococcus, Synechococcus) at night than in the day18 may be coupled with the observed tintinnid diel variations.

For photosynthetic organisms, cell division generally occurs at night and/or in the late afternoon17,57, which eventually leads to higher abundance at night than in the day18. As for heterotrophic microzooplankton tintinnids, photosynthetic organisms, e.g., nanoplankton (nanoflagellates), are important food items influencing their abundance and composition in the oligotrophic seas58,59. Our study showed that tintinnid abundance at night was higher than in the day for two possible reasons: (1) oceanic tintinnid species have stronger cell division in midnight than in the day in tropical Pacific waters60; and (2) predation on picoplankton, nanoplankton and heterotrophic bacteria occurred primarily at night61,62,63. Further studies on growth rates and cell division of tintinnid species are needed to better characterizing their diel vertical migration in the Pacific Ocean.

Differences of ciliate community between the nSCS and tWP oceanic waters

Abundance peaks of planktonic ciliates occurred in surface and DCM layers in both the nSCS and tWP, but highest abundances occurred in surface layer of the nSCS, and DCM layer of the tWP (Fig. 1). Our results are consistent with Wang et al.40, which discovered this phenomenon and proposed a hypothesis to verify it. The nSCS is located at the convergence area of the shelf and slope waters where exchanges often occur with nutrient loaded waters originating from the Pearl River through surface current64,65,66,67,68,69,70,71. For example, the nutrient values in the Pearl River (~ 100 μM) were about 100 folds higher than in the nSCS slope (~ 1.2 μM)66,70,71. Nutrients are material basis for the growth of microphytoplankton and heterotrophic bacteria68. High nutrient concentrations always accompanied with high abundance of microphytoplankton and heterotrophic bacteria in surface waters in the oligotrophic tropical seas, which further affected and determined microzooplankton abundance and composition72,73,74,75,76. In contrast, the tWP is located at a tropical Pacific warm pool surrounded year-round by oligotrophic oceanic water. This may be the main reason for the surface layer ciliate abundance in the nSCS clearly higher than in the tWP.

Aloricate ciliates were dominant groups at each sampled depth of both sites (Supplementary Fig. S1), which was similar to previous observations in adjacent seas4,9,40,42,44,49. As for tintinnid assemblages, we identified more species in the nSCS (57 species) than in the tWP (51 species) (Supplementary Table S2), which was not consistent with previous investigations40,43,77, who found more species in adjacent seas. Low sampling frequency is often accompanied by low species richness78,79. The total samples in the tWP (45 samples) and nSCS (72 samples) were much lower than in previous studies (≥ 100 samples)40,43,77. Thus we speculate that low sampling frequency in our results could be the main reason for the disagreement. High tintinnid abundance and species richness mainly appeared at around DCM depths in both the nSCS and tWP. A high Chl a environment may be an important factor for influencing tintinnid distribution in oceanic waters80,81.

Methods

Study area and sample collection

The variation of ciliate vertical distribution was addressed by conducting two time-series sampling in the upper 500 m at two distinct sites, Station (St.) S1 in nSCS and St. P1 in tWP, during two different cruises (Fig. 7). St. S1 was visited from 29 to 31 March 2017 aboard R.V. “Nanfeng”, and St. P1 from 2 to 3 June 2019 aboard R.V. “Kexue”. During 48 h (St. S1) or 24 h (St. P1) sampling periods, seawater samples were collected by using a CTD (Sea-Bird Electronics, Bellevue, WA, USA)—rosette carrying 12 Niskin bottles of 12 L each (Supplementary Table S5). In the nSCS, the sampling depths were 3, 10, 25, 50, DCM (deep Chl a maximum layer), 100, 200 and 500 m; in the tWP, the sampling depths were surface (3), 30, 50, 75, DCM, 150, 200, 300 and 500 m. Casts were approximately launched every 6 h, the CTD determining vertical profiles of temperature, salinity and chlorophyll a in vivo fluorescence (Chl a). A total of 117 seawater samples were collected for planktonic ciliate community structure analysis. For each depth, 1 L seawater sample was fixed with acid Lugol’s (1% final concentration) and stored in darkness at 4 °C during the cruise.

Figure 7
figure 7

Survey stations in the northern South China Sea (nSCS) and tropical West Pacific (tWP).

Full size image

Sample analysis and species identification

In the laboratory, water samples were concentrated to approximately 200 mL by siphoning off the supernatant after the sample had settled for 60 h. This settling and siphoning process was repeated until a final concentrated volume of 50 mL was achieved, which was then settled in two Utermöhl counting chambers (25 mL per chamber)82 for at least 24 h. Planktonic ciliates were counted using an Olympus IX 73 inverted microscope (100 × or 400 ×) according to the process of Utermöhl82 and Lund et al.83.

For each species, size (length, width, according to shape) of the cell (aloricate ciliate) or lorica (tintinnid, especially length and oral diameter) were determined for at least 10 individuals if possible. Aloricate ciliates were categorized into small (10–20 μm), medium (20–30 μm) and large (> 30 μm) size-fractions for maximum body length of each individual following Wang et al.43. Tintinnid taxa were identified according to the size and shape of loricae following previous references1,9,40,44,84,85,97,98. The partial Mantel tests were performed between ciliate community and environmental factors in R4.1.1.