Specimens of Tigriopus inhabiting intertidal rock pools were collected with a spuit at seven different coastal areas of Korea (Figure 1 and Table 1) and brought to the laboratory in plastic container with seawater. In the absence of other animals, specimens of Tigriopus were kept alive in a 20°C incubator under a photoperiod of 12 h light: 12 h dark, for 2–3 months after being brought to the laboratory. As food source for copepods, a small amount of Tetra Bits was provided.

Several individuals of Tigriopus were sorted under a dissecting microscope and stored in 99% ethyl alcohol. The descriptive terminology of Huys et al. (1996) was adopted. Abbreviations used in the text are as follows: Mxp, maxilliped; exp, exopod; enp, endopod; P1–P5, first to fifth thoracopod; exp (enp)-1 (2, 3) to denote the proximal (middle, distal) segment of a three-segmented ramus. BN, Baengnyeong island; DB, Daebu island; BR, Boryeong; SGP, Seogwipo; BS, Busan; JMJ, Jumunjin; GS, Goseong.

For morphological observation, slides were prepared using the sandwich method of 20 adult females of each species (total 60 individuals). Figure works for morphological comparison of each appendage (Mxp, P1 enp-1, P1 enp-2, P4 exp-3, P5 exp) were completed through observation at 400x magnification with an Olympus BX51 differential interference contrast microscope equipped with a drawing tube. First, length of each segment was measured to calculate the length/width ratio. Based on the ratio, analyses were conducted to compare the differences between the mean of the three species in SPSS. Normal distribution was tested, and while some Mxp’s data were below significance, because the number of samples was sufficient, the data were kept for analysis. In addition, Bonferroni correction was performed, and ANOVA test determined which features showed significant morphological differences between species (p < 0.05).

In total, 18 specimens from BS were prepared for Scanning electron microscopy (SEM). For SEM preparation, specimens were transferred into hexamethyldisilazane (HMDS) for drying (Shively and Miller, 2009), mounted on stubs, and finally coated with gold in an ion coater, SPT-20 (COXEM, Korea). Coated stub was observed under SEM on the in-lens detector at an accelerating voltage of 15.0 Kv. The specimens were photographed using a COXEM EM-30 SEM at the Biodiversity laboratory in Hanyang University. Digital photographs were processed and combined into plates using Adobe Photoshop CS6.

Partial mtCOI sequences were successfully obtained from five specimens of the three putative different species (JMJ; BN, BR, SGP; BS). As a result of calculating the average pairwise distance with the mtCOI sequences obtained from each cluster, the mtCOI distances within each species are shown in Table 2. In the case of T. californicus, the average intra-species distance was 0.222, which is considered to reflect the fact that it is a cryptic species complex.

Busan population differs from the West coast and Jeju population by about 5–8% (mean = 0.063, S.E. = 0.011) and differs from the East population by about 31–38% (mean = 0.341, S.E. = 0.037) (Supplementary Tables 2, 3). These divergence values exceed the range of mean distances within T. east, T. japonicus, and T. west (0.014–0.037) (Table 2).

Mean distance within Tigriopus was 0.298 (S.E. = 0.022) (Supplementary Table 2). The topologies of the Bayesian inference tree and maximum likelihood tree (Figure 7), maximum parsimony tree (Supplementary Figure 1), and neighbor joining tree (Supplementary Figure 2) were different.

Family Harpacticidae Dana, 1846

Genus Tigriopus Norman, 1869

Tigriopus koreanus sp. nov.

(Figures 8, 9)

urn:lsid:zoobank.org:act:9F568FAE-D7A0-4439-8BED-0B72056C494C

Type locality – Rock pool, Gijang, Busan, South Korea, 35°11′40.6″N 129°13′29.2″E

Type material – Holotype; 1 female on slide (NIBRIV0000880626). Allotype; 1 male on slide (NIBRIV0000880627). Paratypes; 3 females and 3 males on a SEM stub (NIBRIV0000880628), 2 females and 2 males in a vial (NIBRIV0000880629).

Other material examined − 17 females on slides (for morphological observation). 12 individuals on a SEM stub

Description – Total body length of female 1000.06 μm (n = 10) (Figures 8A,B), male 720.12 μm (n = 9) (Figures 8D,E); Habitus, segmentation, armature, and general ornamentation of somites and appendages as given by Ito (1969) for the redescription of T. japonicus.

Based on female, Mxp basis (Figure 9A) has a maximum length of 63.9–80.6 μm (mean = 73.5; SD = 4.2), a maximum width of 34.7–45.8 μm (mean = 40.6; SD = 2.9), and the ratio average between length and width is 1.8. P1 enp-1 (Figure 9B) has a maximum length of 66.7−91.7 μm (mean = 80.2; SD = 7.6), a maximum width of 25.0−38.9 μm (mean = 33.1; SD = 3.5), and the ratio average between length and width is 2.4. The maximum length of P1 exp-1 is 94.4−121.0 μm (mean = 106.5; SD = 7.5), and the maximum length of P1 exp-2 is 65.3−88.9 μm (mean = 78.5; SD = 6.3). P4 exp-3 (Figure 9C) has a maximum length of 55.6−68.1 μm (mean = 61.5; SD = 3.7), a maximum width of 18.1−26.4 μm (mean = 22.4; SD = 1.7), and the ratio average between length and width is 2.7. Female P5 exp (Figure 9D) has a maximum length of 45.8−58.3 μm (mean = 51.7; SD = 3.1), a maximum width of 27.8−34.7 μm (mean = 31.1; SD = 2.0), and the ratio average between length and width is 1.7 (Supplementary Table 4).

Sexual dimorphism of male expressed in A1 (Figure 8F), P2 enp (Figure 8G), P5, and P6 (Figure 8H).

Remarks – From measuring and comparing the morphological features of 20 females of each species, subtle morphological differences in the length:width ratio (l/w) of the Mxp basis, l/w ratio of the P1 enp-1, P4 exp-3, and P5 exp (Table 3 and Figures 9, 10) among the Korean species of Tigriopus were revealed.

T. koreanus sp. nov. and T. west showed significant differences in Mxp basis l/w (P < 0.001), P1 enp-1 l/w (P < 0.001). The difference between T. koreanus sp. nov. and T. east showed significant differences in all four morphological features (P < 0.001). The Mxp basis l/w of T. koreanus sp. nov. was about 1.8, which was stouter than the other two species (l/w = 2). For P1 enp-1 l/w, T. west is about 2.8, the longest, T. east is 2.6, and T. koreanus sp. nov. is 2.4, which is different in all three species. On the contrary, P4 exp-3 l/w showed the narrowest trend at T. east at about 2.9, followed by T. koreanus sp. nov. with 2.7 and T. west with 2.6. P5 exp l/w was similar in T. west and T. koreanus sp. nov., but both species showed significant differences from T. east.

According to DDA, compared to the other two species, T. koreanus sp. nov. showed strong intensity at 2,986−2,988, 3,066−3,067, and 3,026 m/z, and these peaks can be regarded as representative peaks of this species (Figure 5). In the case of T. west, ten representative peaks were identified, including 2,958, 3,036−3,037, 3,617−3,618, and 4,242. Tigriopus east showed distinct differences from the other two species in 22 mass, including 2,950, 3,310−3,312, and 3,287−3,288.

Computing the distance by comparing the mtCO1 partial sequences of three species, the mtCO1 distance between each species is as follows: T. west—T. koreanus sp. nov.; 0.063 (0.045–0.081), T. west—T. east; 0.301 (0.261−0.346), T. east—T. koreanus sp. nov.; 0.341 (0.308−0.379) (Supplementary Tables 2, 3).

Etymology – Species name refers to the type locality (i.e., Republic of Korea).

There was no difference in raw spectra (Figure 2A) when an individual specimen was cut in half by prosome and urosome, processed in separate tubes by body parts. The protein spectra of Schizopera sp. (Miraciidae) were also obtained in order to observe spectra variation from species not belonging to Tigriopus (Figure 2F). The protein spectra of Schizopera sp. were significantly different from Tigriopus spp. (Figures 2A–E).

Figure 5 shows relative difference between peaks when the three species are compared together, and it is sufficient to find the difference between the three Korean species of Tigriopus. Until now, various factors which could affect difference in spectra such as fixative solution and storage period after fixation have been investigated (Rossel and Martínez Arbizu, 2018). Usefulness of MALDI-TOF MS in cryptic species identification such as bacteria, mosquitoes, fish and sandflies have also been documented (Müller et al., 2013; Dieme et al., 2014; Paauw et al., 2015; Chavy et al., 2019; Maasz et al., 2020). This study further expands on the usefulness of MALDI-TOF MS, confirming that higher-resolution species identification (identification of cryptic species) is possible for copepods as well.

Despite its ability to identify cryptic species, there are several limitations. First, closely related species have similar protein composition, so the aspects of mass spectra patterns may appear similar. In such case, using only the overall pattern of mass spectra may lead to insignificant results. We expect that increasing the SNR to focus specifically on peaks with strong signal difference help detect cryptic species better. Just as DNA barcoding varies in resolution by different gene markers, and markers are chosen depending on the specific study purpose, parameters such as SNR, halfwindow, minfrequency for MALDI-TOF methodology must be adjusted accordingly.

Secondly, MADLI-TOF MS does not provide insight on the phylogenetic relationships between species. Because this method simply compares the overall protein pattern for species identification, protein molecules, composition, or function is not considered. Granted, if there is a difference in genetic distance, it is reflected on protein composition, resulting in mass spectra difference. However, this does not reflect proportional difference in the genetic distance. As such, to gain insight on phylogenetic relationship whilst discerning species complex, MALDI-TOF MS should be used complementally with DNA barcoding method.

Thirdly, it can be misleading to delimit species solely using MALDI-TOF without any prior knowledge about the species being investigated. In such case, cluster result obtained from the dendrogram may be difficult to interpret, since the number of clusters (or in this case, species) differ depending on the criterion setting of the y-axis. However, since this study presented three pieces of evidence (morphology, molecule, and protein) together, other evidences allowed us to present species classification criterion in MALDI-TOF. What this study can suggest is that Korean species of Tigriopus can be classified based on the height value of 3 in the dendrogram through protein mass comparison. Since criteria vary depending on the taxa, in order to use MALDI-TOF for species identification, a library must be established through further studies dealing with many taxa. To achieve this, role of taxonomists to confirm morphology while building on curated spectra, or molecular data would be essential.

Finally, MALDI-TOF MS can be affected by other factors, such as environmental factors (Karger et al., 2019; Chac et al., 2020). Karger et al. (2019) suggest that subtle differences within mass spectra can occur depending on habitat type, season, and region. This means environmental factors can directly influence differences in clusters, resulting differentiation of species. To improve the aforementioned limitations, sufficient accumulation of data from various species and habitats is required.

As is the case with other methods, it is still difficult to accurately identify a species with MALDI-TOF MS alone, so it is necessary to accumulate data library by morphological taxonomists, and in the beginning, it will be necessary to verify the species identification through other methods.

Based solely on this result (Supplementary Table 3), it could be confirmed that the Busan community differs from the West coast and Jeju communities by about 6%. According to a study on copepod DNA barcoding (Baek et al., 2016), the mean interspecific distance of harpacticoida was 1.6%, which is similar to the results of the three Korean Tigriopus species obtained in this study (Table 2). Also, in other copepod orders, there were cases where the mtCOI sequences between congeners differed only by 0.26−1.1%. Therefore, the difference of 6% is considered to be one of the supporting evidences for the species classification. As such, it is evident from existing studies, that range of intraspecific and interspecific distance varies for each copepoda order. The fact that the range of distance always vary among taxa is an indication that classification of species using only fragment sequence of one molecular marker is with limitations.

With DNA templates, researchers can obtain sequences of different lengths and locations depending on the primers used. Therefore, distinguishing cryptic species with just mtCOI barcode may prove to be an ambiguous task, and there is a possibility that they will result in the same species, although they are in fact different species. In accordance with this, in several studies (e.g., Huys et al., 2007; Thum and Harrison, 2009; Wyngaard et al., 2010; Blanco-Bercial et al., 2011; Marrone et al., 2013; Vakati et al., 2019) suggest that multigene analyses are required for more accurate results when identifying species through molecular information. Therefore, rather than using only one marker such as mtCOI for species identification, providing protein analysis alongside can strengthen the support for the overall results.

In a previous study, Rossel and Martínez Arbizu (2019) showed that Leptastacus cf. laticaudatus can be separated through MALDI TOF-MS experiments in three putative cryptic species. The mtCOI distances of the three cryptic species included in this L. cf. laticaudatus complex ranged from 15.32 to 31.15%. Although mtCOI distance between T. koreanus sp. nov. and T. west was less (6%) compared to distances reported by Rossel and Martínez Arbizu (2019), results from protein analysis showed that they were indeed different clusters.

Clusters observed from the protein data are different from clusters from phylogenetic trees based on mtCOI data. Therefore, it is difficult to extract evolutionary and systematic relationships based solely from these results. It is also challenging to interpret whether T. east and T. west clusters are closer to the T. koreanus sp. nov. clusters.

It is predicted that the reason that a Jeju individual had a high similarity to Busan community is because Busan and Jeju are relatively closer than other West Sea regions. In the same vein, when comparing the mtCOI data uploaded to the NCBI, there was only 4.5−6.3% difference from the data known as T. west on the South Sea coast (Yeosu, Sacheon, Tongyeong) or on the northeast coast of Jeju (Gimnyeong, Seongsan), which are adjacent to Busan.

Although the East sea community is relatively close to Busan in geographic distance, the molecular distance differs as much as the distance from the Japanese species. A probable explanation for this is that T. east in the East Sea region have been influenced by the North Korea Cold Current coming from the north, rather than the Kuroshio Current coming from the south. If samples of Tigriopus from South Sea coast and North Korea are obtained in the future, it would be possible to verify this hypothesis through protein spectra measurements. In the BI and ML tree (Figure 7), the topology supports this hypothesis. Because the origin of T. east is different, it appears as a clade different from the species in the region related to the Kuroshio Current.

On the other hand, according to the MP tree (Supplementary Figure 1) and NJ tree (Supplementary Figure 2), the Korean species form one clade (bootstrap; 23, 67%, respectively). This supports the possibility that individuals originating from the same species by the Kuroshio Current were geographically divided, resulting in speciation. Considering the geological aspect, according to a previous study (Knowlton and Weigt, 1998), it was estimated that millions of years would have passed per 1.5−2.6% of mtCOI mutation rates in crustaceans. According to this result, the estimated timing of speciation of T. east and T. west from their most recent common ancestor would be between 20 and 11.6 mya (Miocene).

The two trees obtained in this study were similar to the topologies of the trees suggested in previous studies. The topologies of MP tree and NJ tree in this study were similar to that of Karanovic et al. (2018), which also used mtCOI gene marker. In these cases, T. east and T. west formed the same clade. On the other hand, the topologies of the BI tree and ML tree in this study (Figure 7) were similar to those of the BI tree in Ki et al. (2009) and Karanovic et al. (2018). Although the markers used were different in Ki et al. (2009), based on ITS-5.8S rDNA, the topology was corresponding.

As can be seen from different topologies depending on the marker, method, and model, in fact, no one knows what evolutionary histories remain veiled. Naturally, there are only a few hypotheses that we can infer from these results.

However, the second hypothesis is difficult to support since Tigriopus is a fast-evolving genus, and Miocene may be too old as time for speciation. Therefore, in this study, topological results of the BI tree and ML tree were deemed more reliable, and we suggest that T. east had different origin from T. koreanus and T. west.

Nevertheless, it is evident that Tigriopus species from Busan is different from the previously known T. west and T. east. The mtCOI barcodes further support that the population of Tigriopus inhabiting Korea consists of at least three different species.

When comparing the morphological differences, the microscopic morphological features that distinguish the new species and other two species of Tigriopus were found in the length-width ratio of the P1 enp-1 and Mxp. Karanovic et al. (2018) found the differences between T. west and T. east by geometric morphometrics in P4 exp-3, female P5 exp, and male P2 enp. In this study, differences in P4 exp-3 l/w and P5 exp l/w of the two species were also confirmed, but when comparing the P5 exp l/w of the three species together, it was confirmed that it was difficult to clearly distinguish the three species by these morphological features as T. koreanus shows overlap, the mid-level characteristics of the other two species. Geometric morphometrics is a good tool to find significant differences in morphological characteristics, but it requires a lot of time and effort. Therefore, based on the results of this study, we recommend MALDI-TOF MS as a simple and efficient method for species identification rather than these time-consuming methods.

Furthermore, studies that are likely to be conducted in the future should include experiments to confirm the difference between nauplii and copepodites of the species of Tigriopus. It would also be meaningful to apply this method to other sibling species complexes such as T. californicus (Table 2; mtCO1 distance within genus; 0.22, S.E.; 0.02) to clear its convoluted taxonomic status.