Introduction

Copepods are ubiquitous in the world's oceans and play a crucial role as major components of marine ecosystems (Turner 2004; Rombouts et al. 2009). They feed on bacteria, protists, including autotrophic, heterotrophic, and mixotrophic species, and larval fish (Mullin 1963; Turner et al. 1985; Gifford 1991; Turner and Tester 1992; Jeong 1994; Jeong et al. 2001; Calbet et al. 2007; Besiktepe and Dam 2020; Lee et al. 2023) and are suitable prey for diverse metazoans (Turner 1984; Barroso et al. 2013). As a key link in marine food webs, the population dynamics of copepods can thus affect those of diverse marine organisms (Turner 2004; Kim et al. 2013; Zeldis and Décima 2020). Therefore, elucidating the interactions between copepods and other marine organisms is crucial to understanding marine ecosystem balance and fishery production.

Phototrophic dinoflagellates are a core component of marine ecosystems and play diverse ecological roles as primary producers and prey for various predators such as mixotrophic and heterotrophic dinoflagellates, ciliates, copepods, and invertebrate larvae (Kleppel 1993; Taylor et al. 2008; Jeong et al. 2010, 2021; Ok et al. 2023a; You et al. 2023; Kang et al. 2023). However, certain phototrophic dinoflagellate species are harmful to diverse marine organisms; they often cause large-scale mortality in fish, shellfish, and mammals (Landsberg 2002; Tillmann and John 2002; Burkholder et al. 2006; Holmes et al. 2014; Turner 2014; Ok et al. 2017; Basti et al. 2018; Broadwater et al. 2018). Thus, although some dinoflagellates can be prey for copepods, they can kill or adversely affect copepods by decreasing their ingestion rates, fecundity, and egg-hatching success (Shin et al. 2003; Turner 2006, 2014). Therefore, the interactions between harmful dinoflagellates and copepods warrant closer inspection. Furthermore, some phototrophic dinoflagellates often cause harmful algal blooms (HABs), and their abundance varies largely in marine environments (e.g., Band-Schmidt et al. 2010; Karlson et al. 2021; Gu et al. 2022). To identify the critical abundances at which bloom-forming dinoflagellate species affect copepods, copepod survival and ingestion rates should be explored as a function of prey concentrations.

Many species of the dinoflagellate genus Karenia produce neurotoxins (Hort et al. 2021). There are 10 described species of Karenia (Guiry and Guiry 2023), all of which cause blooms (e.g., Chang 1999; Yang et al. 2001; Botes et al. 2003; Davidson et al. 2009; Steidinger 2009; Yamaguchi et al. 2016; Iwataki et al. 2022; Orlova et al. 2022). Several studies exist on the interactions between the common calanoid copepod Acartia and two Karenia species, K. brevis and K. mikimotoi, which have been shown to inhibit copepod survival, ingestion, and fecundity (Uye and Takamatsu 1990; Turner et al. 1998; Speekmann et al. 2006; Breier and Buskey 2007; Cohen et al. 2007; Waggett et al. 2012). However, no studies have reported the interactions between other bloom-forming Karenia species and Acartia.

Karenia selliformis is known to produce brevetoxins, gymnodimines, or brevenal (Miles et al. 2000, 2003; Haywood et al. 2004; Mardones et al. 2020), and its blooms have been associated with the death of vertebrates, including fish and sea birds, and invertebrates, including abalone, chitons, sea anemones, sea urchins, sea stars, snails, limpets, and octopuses, in many countries (Arzul et al. 1995; Mackenzie et al. 1996; Clément et al. 2001; Heil et al. 2001; Uribe and Ruiz 2001; Orlova et al. 2022). Although the lethality of K. selliformis has been reported in diverse marine organisms, such as phytoplankton, manila clams, oyster larvae, and juvenile kelp sporophytes (Hégaret et al. 2007; Natsuike et al. 2023; Ok et al. 2023b), no harmful effects have been reported for copepods. Moreover, Karenia bicuneiformis (= K. bidigitata) contains brevetoxins (Haywood et al. 2004); however, no harmful effects from its blooms have been reported to date when blooms by this species occurred in Gordon Bay, South Africa, and Benguelan current system (Botes et al. 2003; Trainer et al. 2010). Furthermore, the interactions between K. bicuneiformis and copepods have not yet been explored.

In this study, the survival and ingestion rates of Acartia hongi, feeding on K. bicuneiformis CAWD81 and K. selliformis NIES-4541, were determined as a function of prey concentration. These observed survival and ingestion rates were compared with those of Acartia spp. on other harmful dinoflagellate species based on literature data. Our results provide a basis for understanding the interactions between copepods and harmful Karenia species and HAB dynamics.

Materials and methods

Preparation of experimental organisms

A culture of K. bicuneiformis CAWD81, initially collected from Fouveaux Strait, New Zealand, and a culture of K. selliformis NIES-4541, initially collected from Katsurakoi Port, Japan, were obtained from the Cawthron Institute Culture Collection of Microalgae and the Microbial Culture Collection at the National Institute for Environmental Studies, Japan, respectively (Table 1). All Karenia cultures were maintained in 250-mL flat culture flasks containing L1 medium at 20 °C under a 14:10 h light–dark cycle (50 μmol photons m−2 s−1 produced by a cool-white fluorescent light). Every 3 weeks, they were transferred into identical flasks containing fresh L1 medium (Guillard and Hargraves 1993). The cell volumes of K. bicuneiformis (ellipsoid) and K. selliformis (prolate spheroid) were calculated following Hillebrand et al. (1999). The carbon content of each culture was estimated using the measured cell volumes following Menden-Deuer and Lessard (2000).

Table 1 Information on the experimental organisms used in this study
Full size table

Using a 303-µm mesh net, copepods were collected from Shiwha Bay, Korea, in June 2023, when the water temperature and salinity were 18.2 °C and 32.1, respectively. They were transported to the laboratory within a few hours (Table 1). The copepods were acclimatized in a temperature-controlled room at 20 °C. They were fed with the dinoflagellate Prorocentrum cordatum (approximately 200 cells mL−1) as prey for 1 day for the K. selliformis experiment and additional 2 days for the K. bicuneiformis experiment. Adult female copepods collected during the same sampling event were used throughout this study.

Survival and ingestion rates

In Experiments 1–2, the survival and ingestion rates of A. hongi, feeding on either K. bicuneiformis (Experiment 1) or K. selliformis (Experiment 2), were measured as a function of the prey concentration (Table 2). The initial concentrations of K. bicuneiformis (or K. selliformis) were set up using autopipettes, and the initial copepod densities (10 female copepods per 500-mL bottle) were established by transferring the copepods individually using a Pasteur pipette (Table 2). Triplicate 500-mL polycarbonate experimental (predator plus prey) and prey control bottles (prey only) were established at each predator–prey density combination. Moreover, triplicate 500-mL predator control bottles (predator only) containing 10 female copepod individuals without prey were also established. Subsequently, a 10-mL aliquot was removed from each experimental and prey control bottle, transferred to a scintillation vial, and fixed with 2% Lugol's solution at the beginning of the experiment to determine the actual initial prey concentrations. The bottles were then filled with freshly filtered seawater, capped, placed on plankton wheels rotating at 0.9 rpm (0.00017 g), and incubated at 20 °C under 20 µmol photons m−2 s−1 illumination with a 14:10 h light–dark cycle. After incubation for 24 h, 10-mL aliquots were extracted from each experimental and prey control bottle and fixed with 2% Lugol's solution. The number of living copepods was enumerated by careful observation through the sides of the capped bottles. The bottles were then refilled to capacity with freshly filtered seawater, capped, placed on rotating plankton wheels, and incubated under the same conditions mentioned above. After the bottles were incubated for 48 h, 10-mL aliquots were extracted from each experimental and prey control bottle and fixed with 2% Lugol's solution. The water in the experimental bottles was filtered through a 100-µm mesh, and the retained copepods were quickly placed into 50-mL beakers. All living and dead copepods were then enumerated under a dissecting microscope (SZX2-ILLB, Olympus, Tokyo, Japan). The prey concentration was measured by counting all the cells, or > 200 cells in 1-mL Sedgewick Rafter chambers under a compound microscope (BX53; Olympus). When Experiment 2 was conducted, the survival and ingestion rates of A. hongi feeding on Prorocentrum cordatum as a non-toxic prey control were measured at a single high prey concentration (10,000 cells mL−1) in the same manner as described above.

Table 2 Experimental design and the actual initial densities of prey and predator species
Full size table

Using the equations described by Frost (1972), the ingestion and clearance rates of the copepods for each target Karenia species at day 1 (from day 0 to day 1) and day 2 (from day 1 to day 2) were calculated. The dilution of the cultures associated with refilling the bottles on days 0 and 1 was considered in calculating the ingestion rates. The data on the K. bicuneiformis ingestion rates were fitted to a Michaelis–Menten equation:

$$IR = \frac{{\text{I} _{{max}} (x)}}{{K_{{IR}} + (x)}}$$
(1)

where x = mean prey concentration (cells mL−1 or ng C mL−1); Imax = the maximum ingestion rate (cells predator−1 day−1 or ng C predator−1 day−1); and KIR = the prey concentration sustaining ½ Imax (cells mL−1 or ng C mL−1).

Molecular identification of Acartia species used in this study

To identify the copepods used in this study, each of the 10 copepods in the predator control bottle in Experiment 1 (with K. bicuneiformis) was transferred to an individual 1.5-mL tube after the experiment had ended. The genomic DNA of the copepods was extracted using the AccuPrep® genomic DNA extraction kit (Bioneer, Daejeon, Korea). For the amplification of small subunit ribosomal DNA (SSU rDNA), polymerase chain reaction (PCR) was conducted using F-Star Taq DNA Polymerase (BIOFACT Co., Daejeon, Korea) in a total volume of 50 µL containing 5 μL of 10 × Taq buffer, 1 μL of 10 mM dNTP mix, 0.25 μL of DNA polymerase, 5 μL of 5 × Band Helper™, 2 μL each of the forward and reverse primers, and 1 μL of the DNA template. The forward and reverse primers used in this study were 18s1075H (5′-CGA AGG CGM TCA GAT ACC GCC CTA G-3′) and 18s1871Hr (5′-CAC CTA CGG AAA CCT TGT TAC GAC-3′), respectively (Easton and Thistle 2014). The thermocycler (AllInOneCycler™, Bioneer) was run under the following conditions: 95 °C for 2 min, followed by 40 cycles of 95 °C for 20 s, 59 °C for 30 s, and 72 °C for 1 min, followed by a final extension at 72 °C for 5 min. The PCR products were purified using the Wizard® SV Gel and PCR Clean-Up System (Promega, Madison, WI, USA) following the manufacturer’s instructions. They were then sequenced using an ABI PRISM® 3700 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). The obtained sequences were assembled using ContigExpress software (Infomax, Frederick, MD, USA) and compared with the SSU rDNA sequences of Acartia species in the NCBI GenBank database based on the BLASTn algorithm. Sequences with > 90% query coverage are shown in the sequence difference comparisons in this study.

For the phylogenetic tree, SSU rDNA sequences of the Acartia species were obtained from GenBank and were aligned using the MEGA v4 program (Tamura et al. 2007). A phylogenetic tree based on the Bayesian and maximum likelihood analyses was established following Kang et al. (2010).

Swimming speed of Karenia

To explore whether the ingestion rates of Acartia species feeding on dinoflagellates were affected by the swimming speeds of the dinoflagellates, the swimming speeds of K. bicuneiformis CAWD81 and K. selliformis NIES-4541 were measured. A dense clonal culture of K. bicuneiformis (or K. selliformis) in a 50-mL cell culture flask, incubated at 20 °C under 20 µmol photons m−2 s−1, was placed under a dissecting microscope (SZX10, Olympus). The field in which the flask was placed was leveled using a magnetic torpedo level. After 30-min acclimation, the swimming Karenia species were recorded in each flask using a video analyzing system (SRD-1673DN; Samsung Techwin, Seongnam, Korea) at a magnification of 20×. The swimming speeds of K. bicuneiformis (or K. selliformis) were calculated by measuring the travel distance and time of 20 swimming Karenia cells showing straight, linear paths.

Statistical analysis

The non-parametric Kruskal–Wallis test was used to examine the effect of Karenia concentration on A. hongi survival. Pearson’s correlations were conducted to examine the relationships between the ingestion rates of A. hongi feeding on dinoflagellates, the mortality (%) of A. hongi incubated with the dinoflagellates, and the dinoflagellate cell sizes and swimming speeds. The software SPSS 26.0 (IBM-SPSS Inc., NY, USA) was used for statistical analyses. The criterion for statistical significance was set at a P value of 0.05.

Results

Molecular identification of A. hongi

The obtained SSU rDNA sequences of the 10 copepod individuals used in this study were 100% identical to each other and to those reported for A. hongi (Accession numbers GU969195 and MZ413971; Fig. 1). The A. hongi sequences obtained in this study were 13.2–15.8% different from those of other Acartia species such as A. omorii, A. clausii, A. tonsa, A. bifilosa, A. negligens, A. danae, A. steueri, A. pacifica, and A. erythraea. The SSU rDNA sequence of A. hongi used in this study was deposited in GenBank under accession number OR356145.

Fig. 1
figure 1

Bayesian tree showing genetic relationships among the calanoid copepod Acartia based on the small subunit ribosomal DNA region (707 bp), using a GTR + G + I model. Numbers near branches represent the Bayesian posterior probability (left) and maximum likelihood bootstrap values (right). A black box represents A. hongi, which was used in this study. The assumed empirical nucleotide frequencies of SSU rDNA comprised a substitution rate matrix with A–C substitutions = 0.1152, A–G = 0.2406, A–T = 0.1173, C–G = 0.1114, C–T = 0.2874, and G–T = 0.1281. Rates were assumed to follow a gamma distribution with a shape parameter of 0.3661 for variable sites. The proportion of sites assumed to be invariable was 0.2416

Full size image

Survival of A. hongi

On day 1, the survival of A. hongi incubated with K. bicuneiformis at mean prey concentrations of 0–1352 ng C mL−1 (0–1988 cells mL−1) was 87–100% (Fig. 2a). On day 2, they were 90–100% at mean prey concentrations of 0–1394 ng C mL−1 (0–2050 cells mL−1) (Fig. 2b). The survival was not significantly affected by K. bicuneiformis concentration (Kruskal–Wallis test; H6 = 8.97, P = 0.175 for day 1; H6 = 9.12, P = 0.167 for day 2).

Fig. 2
figure 2

Survival (%) of Acartia hongi as a function of mean Karenia bicuneiformis concentration on day 1 (a) and day 2 (b). Symbols represent treatment means ± standard error

Full size image

On day 1, the survival of A. hongi incubated with K. selliformis at mean prey concentrations of 0–2400 ng C mL−1 (0–1690 cells mL−1) was 87–97% (Fig. 3a). On day 2, the survival decreased as a function of mean prey concentrations (Fig. 3b); all the copepods had died at a mean prey concentration of 2682 ng C mL−1 (1889 cells mL−1). The survival was not significantly affected by K. selliformis concentrations on day 1 (Kruskal–Wallis test, H6 = 4.03, P = 0.673); however, they were significantly affected on day 2 (Kruskal–Wallis test, H6 = 18.23, P = 0.006).

Fig. 3
figure 3

Survival (%) of Acartia hongi as a function of mean Karenia selliformis concentration on day 1 (a) and day 2 (b). Symbols represent treatment means ± standard error

Full size image

The survival of A. hongi incubated with P. cordatum was 95% at a single mean prey concentration of 1452 ng C mL−1 (9680 cells mL−1) for 2 days.

Ingestion rate of A. hongi

On day 1, the ingestion rates of A. hongi feeding on K. bicuneiformis continuously increased with increasing mean prey concentrations (Fig. 4a). The highest ingestion rate of A. hongi on K. bicuneiformis, 3813 ng C predator−1 day−1 (5608 cells predator−1 day−1), was achieved at a mean prey concentration of 1352 ng C mL−1 (1988 cells mL−1). When Eq. (1) was used, the Imax of A. hongi feeding on K. bicuneiformis was 9286 ng C predator−1 day−1 (13,656 cells predator−1 day−1), and the KIR was 1710 ng C mL−1 (2515 cells mL−1). The maximum clearance rate of A. hongi feeding on K. bicuneiformis was 338 µL predator−1 h−1 at a mean prey concentration of 13 ng C mL−1 (19 cells mL−1). On day 2, the ingestion rates of A. hongi feeding on K. bicuneiformis increased at mean prey concentrations ≤ 116 ng C mL−1 (171 cells mL−1); however, they decreased to zero at higher prey concentrations, indicating that ingestion was inhibited at high prey concentrations (Fig. 4b).

Fig. 4
figure 4

Ingestion rates of Acartia hongi feeding on Karenia bicuneiformis as a function of mean prey concentration (x) on day 1 (a) and day 2 (b). The curve in (a) is fitted using the Michaelis–Menten equation. Ingestion rate (ng C predator−1 day−1) = 9286 [x/(1710 + x)], r2 = 0.525. Symbols represent treatment means ± standard error

Full size image

On day 1, the ingestion rates of A. hongi feeding on K. selliformis were positive at mean prey concentrations ≤ 1159 ng C mL−1 (816 cells mL−1). However, the rate decreased to zero at a higher prey concentration (Fig. 5a). The highest ingestion rate of A. hongi on K. selliformis, 1373 ng C predator−1 day−1 (967 cells predator−1 day−1), was achieved at a mean prey concentration of 1159 ng C mL−1 (816 cells mL−1). The maximum clearance rate of A. hongi feeding on K. selliformis was 1612 µL predator−1 h−1 at a mean prey concentration of 18 ng C mL−1 (13 cells mL−1). On day 2, the ingestion rates of A. hongi feeding on K. selliformis were zero at all prey concentrations (Fig. 5b).

Fig. 5
figure 5

Ingestion rates of Acartia hongi feeding on Karenia selliformis as a function of mean prey concentration on day 1 (a) and day 2 (b). Symbols represent treatment means ± standard error

Full size image

The ingestion rate of A. hongi feeding on P. cordatum was 4541 ng C predator−1 day−1 (30,273 cells predator−1 day−1) at a single mean prey concentration of 1452 ng C mL−1 (9680 cells mL−1) for 2 days.

Swimming speed of Karenia

Karenia bicuneiformis and K. selliformis showed linear and helical swimming behaviors; however, jumping behaviors were not observed. The average (± standard error) and maximum swimming speeds of K. bicuneiformis (n = 20) were 257 (± 17) and 480 µm s−1, respectively. The average and maximum swimming speeds of K. selliformis (n = 20) were 232 (± 13) and 340 µm s−1, respectively.

Discussions

The present study reports, for the first time, the interactions between Acartia species and the dinoflagellates K. bicuneiformis CAWD81 and K. selliformis NIES-4541. Previous studies on the interactions between Acartia species and K. brevis, Acartia species and K. mikimotoi, and Temora longicornis and K. selliformis K-1319 have been explored (Uye and Takamatsu 1990; Turner et al. 1998; Breier and Buskey 2007; Cohen et al. 2007; Waggett et al. 2012; Xu and Kiørboe 2018).

This study highlights that A. hongi feeds on K. bicuneiformis and K. selliformis; however, it stops feeding on them depending on the mean prey concentration and incubation time. The highest ingestion rate of A. hongi when feeding on K. bicuneiformis was higher than that when feeding on K. selliformis. Furthermore, A. hongi died when exposed to K. selliformis but not when exposed to K. bicuneiformis. Therefore, K. bicuneiformis and K. selliformis interact with A. hongi differently.

Effects of two Karenia species on the survival of A. hongi

Acartia hongi was killed by K. selliformis NIES-4541 but not by K. bicuneiformis CAWD81. Before this study, in the family Kareniaceae, K. brevis CCMP2228, the Nagasaki University strain of K. mikimotoi, Karlodinium armiger K-0668, and the Alfacs Bay strain of Karlodinium corsicum were known to kill Acartia spp., whereas Karlodinium veneficum K-1385, K-1635, K-1640, and K-1386 does not kill them (Uye and Takamatsu 1990; Delgado and Alcaraz 1999; Waggett et al. 2012; Berge et al. 2012). At similar Kareniacean dinoflagellate concentrations (1100–1500 ng C mL−1) after 2 days of incubation, the mortality (the inversion of the survival) of A. hongi incubated with K. selliformis NIES-4541 was slightly lower than those of Acartia spp. incubated with Kl. armiger K-0668 and the Alfacs Bay strain of Kl. corsicum; however, much higher than those of Acartia spp. incubated with K. brevis CCMP2228, K. bicuneiformis CAWD81, and Kl. corsicum GCORS1 (Table 3; Fig. 6). These comparisons indicate that K. selliformis is more harmful to A. hongi than K. bicuneiformis. To date, there have been no reports on the mortality of marine animals during K. bicuneiformis blooms (e.g., Botes et al. 2003; Trainer et al. 2010). However, reports from various countries, such as Chile, Kuwait, New Zealand, and Russia, have detailed the mortality of marine animals during K. selliformis blooms (Mackenzie et al. 1996; Clément et al. 2001; Heil et al. 2001; Uribe and Ruiz 2001; Orlova et al. 2022). Thus, the field reports on the relationships between marine animal mortality and K. bicuneiformis and K. selliformis blooms are similar to the results of this study, which shows that K. bicuneiformis and K. selliformis have different effects on the survival of A. hongi.

Table 3 Mortality (%) of Acartia incubated with harmful dinoflagellates at the prey concentrations of 1100–1500 ng C mL−1 for 2 days, calculated ingestion rate (ng C predator−1 day−1) of Acartia species feeding on harmful dinoflagellates at 1000 ng C mL−1, and the cell size (ESD, equivalent spherical diameter, µm) and maximum swimming speed (MSS, µm s−1) of the dinoflagellates
Full size table
Fig. 6
figure 6

Mortality (%) of Acartia species incubated with harmful dinoflagellates at 1100–1500 ng C mL−1 concentrations for 2 days. Red bars indicate the mortality measured in this study. An asterisk on each bar indicates that the value in this figure was significantly or ≥ 20% different from the mortality of Acartia species without the target dinoflagellate species (i.e., predator-only control), with the lowest cell density, or with non-toxic prey species in the reference. See Table 3 for details

Full size image

Some strains in the genus Karenia have diverse toxin compositions such as brevetoxins, brevenal, gymnodimines, and gymnocins (Baden 1989; Seki et al. 1995; Miles et al. 2000, 2003; Satake et al. 2002, 2005; Bourdelais et al. 2005; Rasmussen et al. 2017; Mardones et al. 2020). Therefore, it is difficult to compare the lethality of these Kareniacean dinoflagellate species and strains using the amount of toxins, and a different method of comparing their lethality is needed. The results of this study suggest that comparing the mortality of Acartia at similar concentrations of Kareniacean dinoflagellates can be a useful method for understanding the relative harmfulness of these organisms. There may be intraspecific variability in the amount of secondary metabolites produced by each dinoflagellate species (e.g., Burkholder and Glibert 2006), and this amount may also be affected by the growth phase and culture conditions. Thus, with careful consideration of intraspecific variability and the growth phase of the culture, the relative harm to strains or species should be compared.

Among the dinoflagellates causing mortality of Acartia spp., the mortality of Acartia spp. incubated with cells of Kl. armiger K-0668, the Alfacs Bay strain of Kl. corsicum, K. selliformis NIES-4541, Margalefidinium polykrikoides CP1, K. brevis CCMP2228, and Gymnodinium catenatum GC19V at the dinoflagellate concentrations of 1100–1500 ng C mL−1 for 2 d was significantly higher than that in controls such as being incubated without the dinoflagellate species, with their lowest cell density, or with non-toxic prey species (Fig. 6). Physical contact and chemicals (toxins or inhibitory substances) can cause mortality of Acartia spp. incubated with these dinoflagellates. Delgado and Alcaraz (1999) reported that the death of A. grani was caused by direct physical contact with the Alfacs Bay strain of Kl. corsicum. Berge et al. (2012) also reported that Kl. armiger K-0668 attacked, immobilized, and ingested A. tonsa. In addition, Kl. armiger K-0668, K. brevis CCMP2228, and G. catenatum GC19V contain karmitoxin, brevetoxins, and gonyautoxins and saxitoxin, respectively, which cause mortality of Acartia spp. (da Costa et al. 2012; Wagget et al. 2012; Rasmussen et al. 2017). Moreover, M. polykrikoides CP1 has harmful effects on copepods and contains reactive oxygen species-like chemicals (Jiang et al. 2009; Tang and Gobler 2009). Therefore, toxins or inhibitory substances of Kl. armiger K-0668, M. polykrikoides CP1, K. brevis CCMP2228, and G. catenatum GC19V are likely to cause mortality in Acartia spp. Some strains of K. selliformis are known to contain toxins such as gymnodimines or brevenal (Miles et al. 2000, 2003; Mardones et al. 2020), and thus K. selliformis NIES-4541 may potentially possess toxins leading to mortality of A. hongi. However, the presence of toxins of K. selliformis NIES-4541 has not been reported. The Alfacs Bay strain of Kl. corsicum at 1100–1500 cells mL−1 for 2 days of incubation resulted in high mortality of A. grani, while Kl. corsicum GCORS1 did not cause significant mortality of A. grani at similar Kl. corsicum concentrations, implying intraspecific variability (Fig. 6). Therefore, considering potential intraspecific variability, further examination is needed to confirm the presence of toxins of K. selliformis NIES-4541.

Effects of two Karenia species on the ingestion rate of A. hongi

The calculated ingestion rates of Acartia spp. on seven reported harmful dinoflagellate species (nine strains) at 1000 ng C mL−1 varied, ranging from 1241 to 7658 ng C predator−1 day−1 (Table 3; Fig. 7a). The calculated ingestion rate of A. hongi feeding on K. bicuneiformis at 1000 ng C mL−1 was lower than those of Acartia spp. feeding on Alexandrium catenalla, Alexandrium minutum, and K. brevis; however, it was higher than those of Acartia spp. feeding on M. polykrikoides, G. catenatum, and K. selliformis. Thus, the ingestion rate of A. hongi feeding on K. bicuneiformis was close to the median of those of Acartia spp. feeding on harmful dinoflagellates. The calculated ingestion rate of A. hongi feeding on K. selliformis was lower than that of Acartia spp. feeding on the other six harmful dinoflagellates (eight strains) at a prey concentration of 1000 ng C mL−1. The mortality of A. hongi incubated with K. selliformis for 2 days was higher than those of Acartia spp. incubated with M. polykrikoides, G. catenatum, and K. bicuneiformis (Fig. 7b). Thus, this high A. hongi mortality may be partially responsible for the low ingestion rate of A. hongi feeding on K. selliformis.

Fig. 7
figure 7

Calculated ingestion rates (ng C predator−1 day−1) of Acartia species feeding on harmful dinoflagellates at the prey concentration of 1000 ng C mL−1 (IR; a). Correlations of IR with mortality (%) of Acartia species incubated with the dinoflagellates at 1100–1500 ng C mL−1 (b) for 2 days, equivalent spherical diameters (ESD) of the dinoflagellates (c), and maximum swimming speeds (MSS) of the dinoflagellates (d). Red bars or circles indicate the results of this study. See Table 3 for details. Linear equation in (c): IR =  − 292 (ESD) + 10,900, r2 = 0.529

Full size image

The calculated ingestion rates of Acartia spp. feeding on the nine harmful dinoflagellate strains at a prey concentration of 1000 ng C mL−1 showed a negative correlation with the equivalent spherical diameters of the harmful dinoflagellates (19.2–33.9 µm; Pearson’s correlation; r =  − 0.727, P = 0.026; Fig. 7c). The calculated ingestion rate of A. hongi feeding on K. selliformis at 1000 ng C mL−1 was the lowest, while its equivalent spherical diameter was the second largest. Nival and Nival (1976) reported that modified filtration efficiency could decrease when particle sizes are ≥ 26.2 µm. Thus, the large size of K. selliformis may be partially responsible for the relatively low ingestion rate of A. hongi feeding on K. selliformis.

The calculated ingestion rates of Acartia spp. feeding on harmful dinoflagellates at a prey concentration of 1000 ng C mL−1 were not significantly correlated with the maximum swimming speeds of the nine harmful dinoflagellate strains (Pearson’s correlation; r =  − 0.407, P = 0.278; Fig. 7d). The calculated ingestion rate of A. hongi feeding on K. bicuneiformis was higher than that of A. hongi feeding on K. selliformis, and the maximum swimming speed of K. bicuneiformis was higher than that of K. selliformis. Contrary to this trend, the calculated ingestion rate of A. tonsa feeding on M. polykrikoides was relatively low, but M. polykrikoides has the highest maximum swimming speed among the harmful dinoflagellates (Jiang et al. 2009; Jeong et al. 2015). Thus, the fast swimming speed of M. polykrikoides may lower the ingestion rate of Acartia feeding on it; however, the maximum swimming speed of K. selliformis may not be high enough to lower the ingestion rate of Acartia.

Difference in the survival and ingestion rates on days 1 and 2

The survival and ingestion rates of A. hongi feeding on K. selliformis NIES-4541 on day 2 were much lower than those on day 1. Furthermore, the ingestion rates of A. hongi feeding on K. bicuneiformis CAWD81 at high mean prey concentrations on day 2 were much lower than those on day 1, although the survival on day 2 was not different from those on day 1. Therefore, differences in the survival or ingestion rates of A. hongi feeding on K. bicuneiformis and K. selliformis were observed as the incubation time increased. Thus, if the target dinoflagellate species are harmful, it is reasonable to measure the survival and ingestion rates of Acartia on days 1 and 2, with subsampling on days 0, 1, and 2.

Ecological implications

Blooms of K. selliformis have been reported in many countries, such as Chile, Japan, Kuwait, New Zealand, Russia, and Tunisia (Mackenzie et al. 1996; Clément et al. 2001; Heil et al. 2001; Elleuch et al. 2021; Iwataki et al. 2022; Orlova et al. 2022; Boudriga et al. 2023). During these blooms, the reported K. selliformis abundance ranges from 43 to 60,000 cells mL−1; however, it is usually ≥ 140 cells mL−1 (Table 4). In the present study, half of A. hongi were killed at a K. selliformis concentration of 149 cells mL−1 on day 2. Moreover, the surviving A. hongi did not feed on K. selliformis on day 2. Therefore, during K. selliformis blooms, Acartia spp. may be killed by the inhibiting substances from K. selliformis or starvation. Exploring the absence of Acartia spp. during K. selliformis blooms will be beneficial.

Table 4 Locations and times of Karenia selliformis blooms and the highest abundance (cells mL−1) during each bloom
Full size table

Karenia selliformis is known to lyse several phytoplankton, such as the prasinophyte Pyramimonas sp. and the cryptophytes Teleaulax amphioxeia and Storeatula major (Ok et al. 2023b). Therefore, K. selliformis may eliminate alternative prey for the copepods. During K. selliformis blooms, Acartia spp. may have difficulty maintaining their populations due to the direct (i.e., production of inhibiting substances) and indirect effects of the dinoflagellate (i.e., removal of alternative prey).

The results of this study show that K. bicuneifomis was predated by A. hongi at prey concentrations of < 500 cells mL−1 but not at higher prey concentrations on day 2. Thus, K. bicuneifomis cells may be eaten by A. hongi at the initial stage of a K. bicuneifomis bloom but not during the main bloom.

The interactions between copepods and six of the ten described Karenia species (i.e., K. asterichroma, K. brevisulcata, K. concordia, K. cristata, K. longicanalis, and K. papilionacea) have not yet been explored. These Karenia species are known to cause HABs (Chang 1999; Yang et al. 2001; Botes et al. 2003; de Salas et al. 2004; Chang and Mullan 2012; Yamaguchi et al. 2016). Therefore, to better understand the interactions between copepods and harmful dinoflagellates in marine ecosystems, the survival and ingestion rates of copepods feeding on the six Karenia species as a function of the Karenia concentration should be explored.