The detection of three species complexes similar to Diacyclops galbinus, D. versutus and D. improcerus (Copepoda: Cyclopoida) from Lake Baikal

Full Text

1. Introduction

Cyclopoids are one of the most abundant and diverse groups of arthropods surpassed only by amphipods, ostracods, and harpacticoids in Lake Baikal (Timoshkin, 2001). Diacyclops Kiefer, 1927 and Acanthocyclops Kiefer, 1927 are the most species-rich genera. Both genera are taxonomically complex and unstable because of their close relation and large number of diverse species. There are several morphological groups in Diacyclops that are recognized as species complexes (Pesce, 1994; Karanovic, Krajicek, 2012; Reid and Strayer, 1994). Diacyclops in Lake Baikal is represented by 17 species, 15 of which are endemic (Mazepova, 1978; Sheveleva et al., 2012; Flössner, 1984). Three Diacyclops species from Lake Baikal, of which D. talievi (Mazepova, 1970) is endemic, belong to the bicuspidatus-group. Two endemic species, D. eulithoralis Arov, Alekseev, 1986 and D. biceri Boxshall, Evstigneeva and Clark, 1993, belong to another virginianus-group according to Pesce (Pesce, 1994) or to group 2 (languidoides) according to Reid (1994). Twelve endemic Diacyclops species have an 11-segmented antennule, segmentation formula (exopod/endopod) of swimming legs: 2.2/3.2/3.3/3.3 and the presence of an exopodal seta on the antenna and belong to the languidoides-group. Diacyclops inhabit from interstitial to maximal depths year-round, however, are diverse and abundant in the littoral zone of Lake Baikal. The endemic Diacyclops is presented by interstitial, benthic and sponge-associated species (Timoshkin, 2001; Alekseev and Arov, 1986).

The majority of endemic Diacyclops species were described by Mazepova G.F in the 1950s and 1960s, who discovered this abundant and highly endemic group (Mazepova, 1978). According to Rylov’s taxonomy system (Rylov, 1948), the author classified all of them as Acanthocyclops. In the subsequent years, only four new species of Diacyclops were described and D. arenosus (Mazepova, 1950) was redescribed from Lake Baikal (Flössner, 1984; Boxshall et al., 1993; Sheveleva et al., 2010; Sheveleva and Mirabdullaev, 2017).

Three endemic species, D. versutus (Mazepova, 1962), D. improcerus (Mazepova, 1950), D. konstantini (Mazepova, 1962) and a subendemic D. galbinus (Mazepova, 1962), are found together in samples and have a similar morphology. D. improcerus, D. konstantini, and D. galbinus, are widespread, while D. versutus is rare for the baikalian littoral zone (Mazepova, 1978). D. galbinus and D. improcerus also inhabit the interstitial of Lake Baikal. D. galbinus was found outside Baikal in Lake Shartlinskoye, located in the northwestern part of baikalian coastline (Sheveleva et al., 2013; Shaburova, 2010).

There are significant variations in the morphological characters of D. improcerus, D. galbinus and D. versutus. Thus, the question arises as to whether all specimens named as versutus, galbinus or improcerus belong to the same species. Mazepova G.F. explained significant individual morphological variability of D. galbinus and D. improcerus due to active speciation (Mazepova, 1978).

The first of our result revealed a discordance between molecular phylogeny and taxonomic identification for morphologically similar cyclopoidsto D. versutus, D. improcerus, and D. galbinus from Lake Baikal (Mayor et al., 2017). According to molecular data, analyzed specimens formed phylogroups with unclear taxonomic status. Each phylogroup contained sequences of D. versutus, D. improcerus and D. galbinus identified by their morphology. The following integrative analysis of the cyclopoids from one of these phylogroups, which inhabited the South Baikal, showed that all specimens were closely related based on morphological and molecular data. Despite their morphological similarity to D. galbinus D. improcerus, and D. versutus, they demonstrated some differences in morphological characters making them representatives of a new species. We have referred to this species as D. sp. (VIG2) (Mayor et al., 2019).

In this study, we continues the study of Diacyclops from Lake Baikal, which are similar to the versutus-, improcerus-, galbinus-groups and D. konstantini, using both morphological and molecular methods.

2. Material and methods

2.1. Sample collection and Taxonomic Identification

Copepods were collected from the South, Central, and Northern basins of Lake Baikal in 2018- 2023. Samples were collected from depths in 2-30 m by scuba divers in 2021-2023. Other samples were collected using a scoop-net with a mesh size of 100 µm from depths in 0.3-0.5 m. Some copepods were kindly provided by Sukhanova L.V and Luchnev A.G (LIN SB RAS). All of the cyclopoids were preserved in 96% ethanol and stored at - 20°C.

The taxonomic identification of the cyclopoids was performed using the identification table by G.F. Mazepova (1978). Only specimens morphologically similar to D. versutus, D. improcerus, D. galbinus, and D. konstantini were used in the study. Sampling characteristics are listed in Table 1.

 

Table 1. Sampling locations, depths and NCBI accession numbers of obtained sequences

ID	Sex	Date	Locality	Coordinates	Depth, m	Substrate	Morphological data	ITS length	NCBI accession numbers
									COI mtDNA	12S rRNA	ITS1 rRNA
BG5**	♀	28.05.2018	Bolshoye Goloustnoye	52°01.352’ N 105°23.514’ E	0.3-0.5	stones	+	318	MK207031	MT020872	MK207045
84**	♀	06.2008	Kurma	53°10.114’ N 106°58.424’ E	0.3-0.5	stones	-	-	GU055755	-	-
BG2**	♀	28.05.2018	Bolshoye Goloustnoye	52°01.352’ N 105°23.514’ E	0.3-0.5	stones, sand	+	-	MK207029	-	-
BG4**	♀	28.05.2018	Bolshoye Goloustnoye	52°01.352’ N 105°23.514’ E	0.3-0.5	stones, sand	+	-	MK207030	-	-
D10**	♀	04.2018	Listvyanka	51°52.022’ N 104°49.567’ E	0.3-0.5	stones	+	-	MK207035	-	-
D12**	♀	04.2018	Listvyanka	51°52.022’ N 104°49.567’ E	0.3-0.5	stones	+	-	MK207037	-	MK207051
BK4**	♀	04.06.2018	Bolshiye Koty	51°54.114’ N 105°04.267’ E	0.3-0.5	stones, sand	-	318	MK207027	MT020873	MK207049
266*	♀	-	Baikal	-	-	-	+	338	MT176788	MT020868	MT010631
270*	♀	-	Baikal	-	-	-	+	338	MT176789	-	MT010632
366*	♀	17.06.2019	Bolshiye Koty	51°54.111’ N 105°04.061’ E	1.2	stone with lichen	+	314	MT176791	-	MT010628
369*	♀	17.06.2019	Bolshiye Koty	51°54.111’ N 105°04.061’ E	1.2	stone with lichen	-	314	MT176792	MT020870	MT010629
397*	♀	03.2014	Bolshiye Koty	-	-	-	+	-	MT176790	-	-
D16	♀	19.05.2018	Sludyanka	51°40.017’ N 103°42.532’ E	0.3-0.5	stones	+	314	MK207039	MT020874	MT010630
BG14	♀	28.05.2018	Bolshoye Goloustnoye	52°01.352’ N 105°23.514’ E	0.3-0.5	stones, sand	+	-	MT176787	-	MK591137
BG15	♀	28.05.2018	Bolshoye Goloustnoye	52°01.352’ N 105°23.514’ E	0.3-0.5	stones, sand	+	320	MT176793	MT020871	MK207047
BK21	♀	04.06.2018	Bolshiye Koty	51°54.114’ N 105°04.267’ E	0.3-0.5	stones, sand	+	333	MK207028	-	MK207050
MM1	♀	23.12.2018	Malye Olchonskiye Vorota	53°01.073’ N 106°54.008’ E	0.5	-	+	-	MT176794	MT020875	MK591138
MM2	♀	23.12.2018	Malye Olchonskiye Vorota	53°01.073’ N 106°54.008’ E	0.5	-	+	-	MT176795	MT020876	-
MM3	♀	23.12.2018	Malye Olchonskiye Vorota	53°01.073’ N 106°54.008’ E	0.5	-	+	-	MT176796	-	-
F2	♀	31.05.2021	Bolshiye Koty	51°54.128’ N 105°06.168’ E	6	stones	+	314	-	-	OR502812
F96	♀	05.06.2021	Nemnyanka	55°32.32’ N 109°48.57’ E	6	sponge	+	-	-	-	OR502819
F107	♀	19.09.2022	Katkov cape	53°11.3512’ N 108°25.679’ E	10	stones	+	314	OR506695	-	OR502813
F112-1	♀	26.09.2022	Arul cape	53°27.855’ N 107°33.192’ E	10	stones, sand	-	-	-	-	OR502820
F112-3	♀	26.09.2022	Arul cape	53°27.855’ N 107°33.192’ E	10	stones, sand	+	-	-	-	OR502821
F112-5	♀	26.09.2022	Arul cape	53°27.855’ N 107°33.192’ E	10	stones, sand	+	-	-	-	OR502822
F112-6	♀	26.09.2022	Arul cape	53°27.855’ N 107°33.192’ E	10	stones, sand	+	-	-	-	OR502823
F112-7	♀	26.09.2022	Arul cape	53°27.855’ N 107°33.192’ E	10	stones, sand	-	-	-	-	OR502814
F116-1	♀	17.09.2022	Angasolka	51°43.6301’ N 103°46.486’ E	20	sand	+	-	-	-	OR502800
F116-2	♀	17.09.2022	Angasolka	51°43.6301’ N 103°46.486’ E	20	sand	+	-	OR501221	-	OR502801
F120	♀	25.09.2022	Nizhniy Kedroviy	54°21.47’ N 108°30.4’ E	15	stone	+	330	-	-	OR502803
F130	♀	26.09.2022	Arul cape	53°27.962’ N 107°33.999’E	30	-	+	300	-	-	OR502804
F135-1	♀	16.09.2022	Ulanovo	51°47.7913’ N 104°31.515’ E	30	sand	+	303	-	-	OR502805
F135-2	♀	16.09.2022	Ulanovo	51°47.7913’ N 104°31.515’ E	30	sand	+	325	-	-	OR502806
F136-1	♀	17.09.2022	Angasolka	51°43.6301’ N 103°46.486’ E	5	sand	+	320	-	-	OR502807
F136-2	♀	17.09.2022	Angasolka	51°43.6301’ N 103°46.486’ E	5	sand	+	-	-	-	OR502815
F144	♀	27.09.2022	Chertov most	51°56.133’ N 105°15.672’ E	15	stones, sand	+	-	PP280626	-	OR502808
F156-1	♀	16.09.2022	Ulanovo	51°47.7913’ N 104°31.515’ E	10	sand	+	-	OR506692	-	OR502809
F156-2	♀	16.09.2022	Ulanovo	51°47.7913’ N 104°31.515’ E	10	sand	+	328	OR506693	-	OR502810
F156-3	♀	16.09.2022	Ulanovo	51°47.7913’ N 104°31.515’ E	10	sand	+	-	OR506691	-	OR502817
F191-1	♀	27.04.2023	Bolshiye Koty	51°53.5859’ N 105°03.502’ E	1	stones, sand	-	333	-	-	OR502811
F191-3	♀	27.04.2023	Bolshiye Koty	51°53.5859’ N 105°03.502’ E	1	stones, sand	+	-	-	-	OR502802
F191-4	♀	27.04.2023	Bolshiye Koty	51°53.5859’ N 105°03.502’ E	1	stones, sand	+	-	-	-	-
F191-5	♂	27.04.2023	Bolshiye Koty	51°53.5859’ N 105°03.502’ E	1	stones, sand	+	-	OR506696	-	OR502818
F193	♀	27.04.2023	Bolshiye Koty	51°53.5859’ N 105°03.502’ E	1	stones, sand	+	-	-	-	-

Note:

«*» – copepods were kindely provided by L.V. Sukhanova and A.G. Luchnev

«**» – data for these copepods were published early (Mayor et al., 2019)

«+» – available data

«-» – no data available

 

2.2. Morphological Analysis

The morphological analysis was performed using a stereomicroscope MSP-1 (Lomo, Russia) and the Olympus CX 41 (Olympus, Japan). The specimens were rehydrated in water, photographed, and measured using a Levenhuk M 800 Plus camera attached to the Olympus CX 41 and LevenhukLite (Levenhuk, Inc., USA) software. The U-DA Olympus Drawing Attachment for Olympus CX 41 (Olympus, Japan) was used to draw the morphological characteristics. The length of the body was obtained by summing the lengths of the cephalothorax, thorax segments, and abdomen segments.

 

Morphological abbreviations:

Ti	innermost terminal seta
Te	outermost terminal seta
Td	dorsal seta
Tl	lateral seta
Tmi	median inner terminal seta
Tme	median outer terminal seta
Lf	caudal rami length
Wf	caudal rami width
Enp3	third endopodal segment
P4, P5	fourth, fifth legs
LP5	length of the distal segment of P5
IAS	internal apical spine of the third endopodal segment of P4
EAS	external apical spine of the third endopodal segment of P4
L	length
W	width
A1	antennule
A2	antenna
Cphth	cephalotorax
sp	spine
se	seta

 

For confocal laser scanning microscopy (CLSM), the female specimen was stained with Congo Red overnight and mounted on a slide, following the procedure outlined by Michels and Büntzow (Michels and Büntzow, 2010). The material was scanned using a Carl Zeiss LSM 710 laser confocal microscope (Zeiss, Germany) with Plan-Apochromat 20×/0.8 and 63×/1.40 Oil DIC M27 lens; 570 – 670 nm filters; 561 nm: 3.0 % lasers. The most variable morphological indices were evaluated using the Correspondence Analysis and used to construct a Principal Component Analysis (PCA) plot. All statistical analyses were performed using Past 4.11 (Hammer et al., 2001).

2.3. DNA Extraction, PCR, and Sequencing

Total DNA was extracted from egg sacs or somatic tissue as follows: ethanol-preserved specimens were rehydrated in mQ water for 20 minutes. The biological material was then incubated in a 2x PCR Encyclo buffer (without Mg2+) (Evrogen, Russia) containing 0.1 mg/ml Proteinase K at 65°C for 1 hour. After that, it was incubated at 94°C for 5 minutes to deactivate Proteinase K. The resulting solution containing total DNA was stored at –20°C and used in a 10-fold dilution for PCR as a DNA template.

PCR was performed using universal primers LCO-1490 and HCO-2198 to amplify the СОI fragment (Folmer et al., 1994), KP2 (5’-AAAAAGCTTCCGTAGGTGAACCTGCG-3’) and 5.8S (5’-AGCTTGGTGCGTTCTTCATCGA-3’) to amplify ITS1 (Phillips et al., 2000), and H13845-12S (5′-GTGCCAGCAGCTGCGTTA-3’) and L13337-12S (5′-YCTACTWTGYTACGACTTATCTC-3′) to amplify 12S (Machida et al., 2002). The amplification was carried out in a T100TM thermal cycler (BioRad, USA) using PCR reagents from Evrogen (Russia). The reaction was performed in a 20µl mixture: 1x Encyclo buffer, 3.5 mM magnesium, 0.5 µM of each primer, 0.2 mM of each dNTP, 0.5 units of Encyclo DNA polymerase, and 2 μl of DNA solution. The amplification program included the stage of heating the mixture to 94 °C for 4 min, 35-40 cycles consisting of the following steps: 94 °C for 15 s, 48 °C or 57 °C (for COI and ITS1, 12S fragments, respectively) for 20 s, 72 °C for 1 min, and the final elongation stage at 72 °C for 4 min. The amplicons were separated and isolated for sequencing from the agarose gel using the protocol described previously (Mayor et al., 2010). The nucleotide sequences of the target fragments were determined using the ABI PRISM BigDye Terminator v. 3.1 sequencing kit in an ABI 3500 8-capillary genetic analyzer (Thermo Fisher Scientific, USA) and in a Nanophor 05 genetic analyzer (Sintol, Russia).

2.4. Molecular-Phylogenetic Analysis

The sequences obtained were deposited in the GenBank database, and their NCBI accession numbers are listed in Table 1. Alignment of the nucleotide sequences and calculation of genetic distances were performed using the MegaX program (Kumar et al., 2018). We evaluated DNA polymorphism using the DnaSP 5.10.01 program (Rozas et al., 2003). Intragenomic polymorphism was detected in some ITS1 sequences, and we encoded sites with double peaks according to IUPAC. To analyze the COI, 12S, and ITS1 datasets, we selected GTR + G, HKY+G, and TN93 + G models, respectively, based on the Akaike information criterion and the Bayesian information criterion, as determinated by jModelTest 2.1.6 (Darriba et al., 2012). Maximum Likelihood trees were constructed using IQ-TREE2 (Minh et al., 2020) and MegaX software. Nodal support for the resulting branches was estimated with 1000 bootstrap replications. Additionally, we include sequences of Diacyclops species and other cyclopoids in our analysis. The accession numbers of the sequences used in the GenBank database are displayed on the phylogenetic trees. The trees were visualized and edited using Interactive Tree Of Life (iTOL) version 6.8.1 (https://itol.embl.de, accessed on September 16, 2023 and October 3, 2023) (Letunic and Bork, 2021).

To propose species partitions from COI data sets, we used species-delimitation methods that employ pairwise genetic distances and tree-based methodologies. We used the Assemble Species by Automatic Partitioning (ASAP) tool with the p-distances and default settings (https://bioinfo.mnhn.fr/abi/public/asap/ accessed on October 3, 2023) (Puillandre et al., 2021). The study employed the Poisson Tree Processes (PTP) and Bayesian Poisson Tree Processes (bPTP) (Zhang et al., 2013) with default settings and the ML-tree obtained in the study. Additionally, the Generalized Mixed Yule Coalescent method (GMYC) was used with a single threshold and an ultrametric tree obtained in the Beast 2.5.2 program (Bouckaert et al., 2014). The source links for PTP and GMYC are https://species.h-its.org/ptp/ and https://species.h-its.org/gmyc/, respectively, accessed on October 20, 2023.

3. Results

3.1. Taxonomic identification

This study selected 43 adult females and one adult male of Diacyclops from Lake Baikal, collected from varios substrates and depths up to 30 m, based on their morphological characteristics (Table 1). Seven of these specimens belong to D. sp. (VIG2), for which we previously published molecular and morphological data (Mayor et al., 2019).

We photographed 38 specimens and measured their morphological characteristics, including the antenule, caudal rami and their setae, P4, P5, and body length. Many estimated morphometric parameters are used in cyclopoid taxonomy, including those used by Mazepova G.F. to describe endemic Diacyclops. The majority of crustacean photos were taken before DNA extraction. Unfortunately, three cyclopids exoskeletons were lost after DNA extraction, leaving only molecular data for these specimens. Two of them are the smallest Diacyclops analyzed.

We found 10 specimens that are similar to D. improcerus, but they differ from it in having a shorter cephalothorax and a longer antennule or armature of the Enp3P4, which contains three spines and two setae instead of two spines and three setae. Out of the 10 specimens that are similar to D. galbinus, 6 specimens have longer caudal rami, shorter lateral seta in relation to the caudal rami width, and longer Te. One specimen (MM1) has a shorter spine of P5 and cephalothorax. Two specimens (BG14, F144) have a shorter cephalothorax (Table 2). For one specimen (F112-7), similar to D. galbinus by EnpP4, we have only molecular data.

 

Table 2. Morphometric characteristics of analyzed cyclopoids

Species, specimen	Lf/Wf	Te/Ti	Tmi/Tme	Td/Te	Tl/Wf	Lenp3P4/ Wenp3P4	IAS/EAS	LA1/Lcphth	Lcphth/LPed2-Ped5**	LspP5/LP5
D. versutus*	1.8-3.4 (2.5)	0.5-1.8 (1.2)	1.7-2.2 (1.8)	0.9-1.5 (1.2)	0.3-0.6 (0.5)	1.0-1.8 (1.4)	0.6-1.2 (1.0)	short, hardly reach the middle of the cephalothorax	-	the length of P5 spine varies significantly
D. galbinus*	2.9- 5.4 (3.7)	0.6- 1.8 (1.2)	1.3-1.9 (1.6)	0.8-1.2 (1.00)	1.4-2.0 (1.6)	1.6-3.0 (2.0)	1.1-1.8 (1.3)	reach the posterior margin of the cephalothorax	the length proportion of the cephalothorax and remained segments of the thorax is about 2	the length of P5 spine is equal or slightly short than the length of P5 distal segment
D. improcerus*	1.8- 4.1 (3.0)	1.0- 2.1 (1.6)	-	-	-	1.0-1.4 (1.2)	1.0-1.9 (1.35)	short, usually reach the middle of the cephalothorax	the length proportion of the cephalothorax and remained segments of the thorax varies from 1.6 to 2.3 (1.86)	the length of P5 spine is equal, slightly short or long than the length of P5 distal segment
MM3	2.63	1.67	1.93	1.00	1.17	1.43	1.29	0.78	0.90	0.68
MM2	2.35	1.39	1.81	-	1.14	1.54	0.83	-	1.15	-
BG14	1.86	1.71	1.82	1.04	0.93	1.21	1.18	0.92	0.84	1.13
BK21	2.50	1.49	2.20	-	1.16	1.28	1.17	0.73	0.98	0.77
F193	1.80	1.20	1.70	1.20	1.20	1.40	1.20	0.79	0.86	0.89
F191-3	2.37	1.57	1.64	1.05	1.18	1.43	1.22	0.77	1.59	1.54
F191-4	2.34	1.73	1.45	1.02	1.25	1.43	1.17	0.95	0.85	1.41
F156-3	1.40	1.69	1.83	-	0.72	1.38	1.05	0.77	0.98	0.81
F156-1	1.44	1.43	1.91	0.47	0.84	1.22	1.00	0.67	0.84	0.75
F120	2.60	1.60	1.90	0.94	1.33	1.46	1.12	0.71	0.98	0.71
F156-2	2.51	2.06	-	1.05	1.13	1.02	1.20	-	0.76	1.15
F96	3.08	1.68	1.73	0.91	1.22	1.57	1.20	0.73	0.77	1.01
F-130	2.70	1.56	1.54	1.10	0.94	1.15	1.32	0.66	2.10	0.80
D16	2.89	1.17	1.48	1.04	1.07	1.24	1.24	1.12	0.93	1.01
366	2.03	2.03	1.53	1.39	1.11	1.16	1.21	0.87	1.40	1.37
F112-5	3.41	1.72	1.59	1.25	1.01	1.27	1.26	0.81	1.16	0.70
F112-6	3.79	1.33	1.86	1.70	1.25	1.18	1.23	0.88	1.00	0.60
F112-3	3.25	1.80	-	1.62	0.87	1.08	1.15	0.79	1.18	0.80
F-107	3.57	-	-	-	0.97	1.14	1.41	-	0.99	1.00
F2	3.69	1.41	1.21	1.57	1.23	1.25	1.13	0.92	1.10	0.80
MM1	3.40	1.26	NA	0.95	1.62	2.16	1.26	0.98	1.14	0.52
BG15	3.45	1.82	1.68	1.00	1.36	1.99	1.19	0.94	1.26	0.51
F144	3.80	1.39	1.60	1.06	1.38	2.10	1.16	1.24	0.98	0.82
F116-1	5.82	0.94	1.53	-	1.83	1.90	1.23	1.23	0.98	0.60
F116-2	4.00	1.17	-	1.28	1.32	2.11	1.33	0.94	1.13	0.89
F136-1	3.76	1.96	1.71	0.81	1.36	2.23	1.33	-	1.20	0.81
F136-2	5.05	2.29	-	0.95	1.71	2.57	1.25	1.3	1.15	0.96
F135-1	3.78	1.75	-	-	1.00	2.50	1.25	1.2	1.56	0.75
F135-2	3.95	1.63	1.52	-	1.35	2.56	1.21	1.2	1.17	0.98

Note:

«*» – data of G.F. Mazepova (1978), the brackets indicate the average value of parameters

«**» – combined length of the segments pediger 2 to pediger 5

«-» – no data available

 

All 14 specimens that are similar to D. versutus have a longer lateral seta relative to the caudal rami width. Two of the specimens (F156-1, F156-3) differ from D. versutus in having shorter caudal rami and a smaller proportion of dorsal seta to Te. Additional, specimen F156-2 differs from D. versutus in having a smaller proportion of Ti and Te. A male specimen (F191-5) was found to be similar to the females (F193, F191-1 – F191-4) and D. versutus in the armature of the enp3P4 and was included in this study. Three specimens belong to D. konstantini.

3.2. Molecular phylogeny and Species delimitation

The study amplified molecular markers under the same conditions from specimens that were morphologically similar and collected together. However, the amplification of DNA fragments varied in specificity and yield (Table 1). For instance, F193 and F191-4, which were collected together and closely resembled F-191-1, F191-3, and F191-5, were not successfully amplified. The study obtained a total of 10, 35, and 21 sequences for the 12S, ITS1, and COI gene fragments, respectively (Table 1). Intragenomic polymorphism of ITS1 was detected in 10 specimens. Six sequences (266, BG15, BK21, F116-1, F120, and F135-1) have double peaks at one site, while three sequences (BG14, 369, and F130) have double peaks at two sites. One sequence (366) has double peaks at seven sites. There were 14 transitions and three transversions. The positions of ITS1 sites with double peaks coincided in the sequences of different specimens (BG14/366, BG14/BK21, and F116-1/266).

Three short ITS1 sequences (F96, F116-2 and F136-2) were deposited in GenBank, but were excluded from the phylogenetic analysis. The analysis used an alignment of 31 sequences (461 bp), revealing 245 sites (excluding sites with gaps/missing data), 73 polymorphic (segregating) sites, and 69 parsimony informative sites.

The ML ITS1 tree divided the sequences of the selected specimens, which were similar to D. galbinus, D. improcerus, and D. versutus, into three clusters: the galbinus-group, the versutus-group, and the improcerus-group (Fig. 1). Each cluster contains several genetic lineages. The galbinus-group comprises four genetic lineages (I-IV), the versutus-group comprises three genetic lineages (VI-VIII), and the improcerus-group comprises three genetic lineages (IX-XI). Sequence F156-2 represents a distinct genetic lineage (V) and was not included in any of the three groups.

 

Fig.1. (a) Distribution of sampling locations (Table 1) of Diacyclops lineages in Lake Baikal, color coded as in the tree; (b) Phylogenetic tree constructed on the base of the ITS1 by the maximum likelihood method (ML, TN93 + G). The number in the node is the bootstrap value of the branching node support. Roman numerals indicate the genetic lineages. The circles near with the ID specimen indicate a sample location.

 

The versutus-group, F156-2 (V), D. konstantini, D. sp. (VIG2), and the improcerus-group formed a separate large cluster. The galbinus-group is genetically distant from this cluster.

The analysis used a total alignment of 38 COI sequences (693 bp), including 22 sequences of Diacyclops from Lake Baikal, 11 sequences from three Diacyclops species of the bicuspidatus-group (D. bisetosus, D. crassicaudis, and D. bicuspidatus) from Poland and Korea, and 4 sequences of D. sp. from Australia. The alignment comprised 134 sites, excluding sites with gaps/missing data, and 57 parsimony informative sites. All Diacyclops sequences from Lake Baikal form a monophyletic group, that is distinct from other Diacyclops. The COI and ITS1 tree topologies are consistent. The COI tree distinguish the versutus-, galbinus- and improcerus-group, each containing several genetic lineages (Fig. 2).

 

Fig.2. (a) Distribution of sampling locations (Table 1) of Diacyclops lineages in Lake Baikal, color coded as in the tree; (b) Phylogenetic tree constructed on the base of the COI gene fragment by the maximum likelihood method (ML, GTR + G). The number in the node is the bootstrap value of the branching node support. Roman numerals indicate the genetic lineages. The bars near with the tree indicate the ‘species’ delimited by ASAP, PTP, bPTP, GMYC methods. The circles near with the ID specimen indicate a sampling location.

 

Two genetic lineages of the versutus-group (VI, VIII) are sister to F156-2 (V) and D. konstantini with high bootstrap support values, which together form a separate cluster. D. sp. (VIG) with two genetic lineages of the galbinus-group (I, II) and two genetic lineages of the improcerus-group (IX, X) form three distinct clusters.

The ASAP and PTP delimitation methods applied to the COI data for Diacyclops from Lake Baikal estimated 10 ‘species’. The results are generally congruent with the COI tree topology, but BG15 (I) and 84 (I), MM1 (I) are distinguished as separate two ‘species’. The bPTP and GMYK methods delimited 11 ‘species’ among the analyzed Diacyclops from Lake Baikal. These methods are incongruent with the ASAP/PTP methods only in the versutus-groupdivision. The bPTP delimited F156-1 (VI), F156-3 (VI), and the VIII genetic lineage (MM2, MM3, BG14, F191-5, and BK21) as three separate ‘species’. The GMYK delimited F156-1 (VI) and F156-3 (VI) as ‘species’, but divided VIII genetic lineage into two ‘species’: MM2/MM3 and BG14, F191-5, and BK21.

The analysis used of 17 sequences for 12S alignment (497 bp). Of these, 9 sequences belong to Diacyclops, from Lake Baikal, 9 sequences belong to 5 Diacyclops species from Australia, Japan, and Ukraine. The alignment comprises 327 sites, excluding sites with gaps and missing data, with 186 polymorphic sites and 162 parsimony-informative sites. The 12S phylogenetic tree showed a monophyletic group of all Diacyclops from Lake Baikal consistent with the COI tree. This group and representatives of the bicuspidatus-group, D. bisetosus from Japan and D. bicuspidatus from Ukraine, are separated into a large cluster (Fig. 3). Three Australian endemic species of the aticola-group, D. scaloni, D. sobeprolatus, and D. humphreysi, are distant from this cluster. The tree topology is congruent with COI and ITS1 topologies for Diacyclops from Lake Baikal. BG15 (I) and MM1 (I) of the galbinus-group form a distant cluster, and D16 (X) and 369 (XI) of the improcerus-group form another cluster. D. sp. (VIG2) is a sister taxon to the X and XI lineages of the improcerus-group.

 

Fig.3. Phylogenetic tree constructed on the base of the 12S fragment of mtDNA by the maximum likelihood method (ML, HKY + G). The number in the node is the bootstrap value of the branching node support. Roman numerals indicate the genetic lineages.

 

3.3. Genetic distances

The length of the rDNA ITS1 region ranges from 300 to 338 bp and is specific to each genetic lineage (Table 1), except for the improcerus-group, where the ITS1 length of the XI and X genetic lineages is identical at 314 bp. Model-corrected genetic distances (TN93+G) between all genetic lineages of Diacyclops from Lake Baikal are very similar to p-distances of ITS1, ranging from 0.7% to 20.1% (Supplementary, Table S1-S3). The maximum genetic distance within genetic lineages is 0.4%. The COI p-distances among all studied genetic lineages of Diacyclops from Lake Baikal range from 9.1% to 20.9%. The largest intra-genetic lineage p-distances for COI are found in the I genetic lineage (5.6%) and the II genetic lineage (2%). The minimum distances among the genetic lineages for both molecular markers are between the sequences of the versutus-group, ranging from 0.7% to 1.6% in ITS1 and 9.1% in COI datasets. The ITS1 and COI genetic distances among genetic lineages of the galbinus-group are 3.5-4.0% and 17.7%, respectively. The ITS1 and COI genetic distances among genetic lineages of the improcerus-group are 3.3-7.2% and 16.1%, respectively. The maximal ITS1 and COI genetic distances between the galbinus-group and the improcerus-group genetic lineages are 12.2-20.1% and 20.1-20.9%, respectively. The 12S p-distances between all studied Diacyclops genetic lineages from Lake Baikal range from 11.8% to 23.1%. The X and XI lineages of the improcerus-group are the closest. The XI lineage of the improcerus-group and the VIII lineage of the versutus-group are the most distant. Similar genetic distances are estimated between the versutus-group and the galbinus-group (20.5%) and between the improcerus-group and the galbinus-group (21.6% and 21.7 %).

3.4. Distribution of the genetic lineages

The representatives of the galbinus-, improcerus-, and versutus-groups have a sympatric distribution (Fig. 1a). Representatives of D. konstantini, D. sp. (VIG2), the VIII genetic lineage of the versutus-group, and the XI genetic lineage of the improcerus-group were found in Bolshiye Koty on different dates. Representatives of D. sp. (VIG2), the VIII, and I genetic lineages were found in one sample from Bolshoye Goloustnoye (Table 1). Additionally, in one sample we found representatives of the versutus-group (F156-1, F156-3 - VI) and F156-2 (V), which is genetically sister to them; representatives of the galbinus-group (F112-7 - II) and the improcerus-group (XI) or representatives of the galbinus-group (MM1 – I) and the versutus-group (MM2, MM3 - VIII).

Each group has one predominant genetic lineage, which has been found in various locations and periods. These are the I genetic lineage of the galbinus-group, the XI genetic lineage of the improcerus-group, and the VIII genetic lineage of the versutus-group.

3.5. Morphometric analysis

PCA was performed on 37 individuals. Molecular data were obtained for most of them. Among them, 9 specimens belong to the galbinus-group, 12 specimens belong to the versutus-group, 8 to the improcerus-group according to molecular data, 3 specimens of D. konstantini, and 5 specimens of D. sp. (VIG2) (BG2, BG4, BG5, D10, D12) were included.

The CA resulted in 8 morphometric characteristics. Tmi/Tme, Td/Te, Ti/Tmi, Ti/Tme, Ti/Td, LA1/Lcpht, Lbody, and LseP5/LspP5 together explain 5% of the Axis1 variability (total variability - 25.6%) and were excluded from the next PCA. More variable parameters such as Lf/Wf, Te/Ti, Tl/Wf, Lenp3P4/Wenp3P4, IAS/EAS, Tl position, IAS/Lenp3P4, IAS/Wenp3P4, Ti/Lf, Tl/Te, Ti/Td, Ti/Te, LP5/WP5, LspP5/LP5, and WP5/LspP5 were used in the PCA. The first two principal components explain 76.1% of the variation in our morphometric data. PC1 and PC2 explain 61.5% and 14.6% of variance, respectively, with eigenvalues of 2.1 and 0.5. The first principal component has the strongest positive correlation with the length and width proportion of the caudal rami (Lf/Wf), less positive correlatina with the length and width proportion of the third endopodal segment length (enp3P4) of P4 (Lenp3P4/Wenp3P4) and the length of the lateral setae with the caudal rami width proportion (Tl/Wf) and negative correlations with ratio of the outer to inner terminal caudal setae (Te/Ti) and with the ratio of the length of the internal apical spine to the length of the P4 third endopodal segment (IAS/Lenp3P4). The second component has strong positive correlations with the ratio of the internal apical spine length to width of the P4 third endopodal segment (IAS/Wenp3P4) and with Lenp3P4/Wenp3P4, the less positive correlation with ratio of the length to width of the P5 (LP5/WP5) and the negative correlation with ratio the P5 internal spine to length of the distal P5 segment (LspP5/LP5) (Fig. 4). The results of the PCA analysis, which is based on morphometric indices, are consistent with the molecular phylogeny in distinguishing genetically distant species or species complexes of Diacyclops. Along the PC1, the D. konstantini and the versutus-group are located, while the galbinus-group and the improcerus-group with D. sp. (VIG2) are located along PC2. The specimens of D. sp. (VIG2) are close to the improcerus- and the versutus-group. Specimens F130 (IX) of the improcerus-group, F156-2 (V) overlapped with specimens of the versutus-group. Specimen F116-1 (I) of the galbinus-group is closer to D. konstantini than to the galbinus-group. The PCA did not clearly separate all closely related genetic lineages. However, specimens F156-3 and F156-1 (VI) of one ITS1 and COI genetic lineage differ from other representatives of the versutus-group in their smaller length and width proportion of the caudal rami, and are located extremely along PC1.

 

Fig.4. Principal component analysis of Diacyclops based on morphometric indices. The numbers indicate specimens ID. Green rows indicate morphometric indices. The colour indicates the species complex according to the phylogenetic trees.

 

3.5. Morphology diversity of the versutus- and improcerus-groups

Fig. 5-7 depict micrographs and drawings of P4 and an antenna of specimens belonging to the versutus- and improcerus-groups. All analyzed specimens of the versutus-group have Enp3P4, armed with three spines and two setae. MM3 and BG14 closer to each other based on COI and were attributed to the VIII genetic lineage. They were delimited as one ‘species’ by ASAP and as two ‘species’ by GMYC. They differ in P4 intercoxal sclerite ornamentation (Fig. 5). BG14 and F193 (VIII) have naked the caudal surface of the sclerite, while MM3 has two rows of short spinules in the middle and on the distal margin. Specimen F156-3 of the VI genetic lineage has the P4 intercoxal sclerite with two rows of short and long hair-like spinules in the middle and on the distal margin. Among MM3 (VIII), F193 (VIII), and F156-3 (VI), the latter has the most groups of setae and spinules on the P4 coxopodite. F156-3 differs from F193 in the number of setae on the second endopodal segment of A2 and has nine setae versus eight setae of F193. The A2 basipodite of F156-3 is ornamented with two diagonal rows of spinules in the middle and on proximal part of the frontal surface, a diagonal row of spinules on proximal part, and a row of the longest spinules in the middle of the caudal surface, a row of long spinules along lateral margin, and a row of hair-like spinules along opposite lateral margin, armed with two setae, one of which is naked and one has short setules and exopodal seta with short setules. The A2 basipodite of F193 are ornamented with three rows of spinules on the proximal and central margins, and two rows of hair-like spinules along both lateral margins, armed with two setae, one of them is a naked and one has short setae and naked exopodal seta (Fig. 6).

 

Fig.5. Confocal laser (396, BG14) micrographs and drawings of P4, caudal of the versutus-group, the V genetic lineage, and D. konstantini (396). Numbers indicate ID specimens. BG14 – P4, intercoxal sclerite, caudal. Roman numerals indicate the genetic lineages. Scale bars: BG14, F156-2 = 20 µm; F156-3, MM3, F193, 396 = 10 µm.

 

Fig.7. Confocal laser micrographs (a VIG2, b VIG2, 366) and drawings of P4, caudal of the improcerus-group. Numbers indicate ID specimens, Roman numerals indicate the genetic lineage. bVIG2 – P4, coxopodite, intercoxal sclerite, caudal. Scale bars: F130 = 10 µm; aVIG2, bVIG2, D16, 366 = 50 µm.

 

Рис.7. КЛСМ микрофотографии (a VIG2, b VIG2, 366) и рисунки каудальной стороны P4 improcerus-группы и D. sp (VIG2). Арабские числа означают номер особи, римские числа – номер генетической линии. b VIG2 – P4, коксоподит, соединительная пластинка с каудальной стороны. Шкала: F130 = 10 µm; a VIG2, b VIG2, D16, 366 = 50 µm.

 

D. konstantini and F156-2 (V) are genetically closely related and differ from the versutus-group in the armature of Enp3P4. They have two spines and three setae. F156-2 P4 intercoxal sclerite with two rows of long spinules. The A2 basipodite of F156-2 is ornamented with two rows of spinules and a group of small spinules on the caudal surface, two rows of hair-like spinules on the proximal part of the frontal surface and hair-like spinules along both lateral margins armed with two naked setae and exopodal seta with short setules (Fig. 5, 6). Second endopodal segment armed with seven setae.

D16 (X), F130 (IX), and 366 (XI) are related to different genetic lineages of the improcerus-group and D. sp (VIG2) is closeely related to them. All analyzed specimens of D. sp. (VIG2) and the improcerus-group, except D16 and F130, have Enp3P4 with two apical spines (IAS and EAS) and three setae. Specimens D16 and F130 have Enp3P4 with three spines – IAS, EAS, and a spine on the outer margin and two setae. D. sp (VIG2) differs from the improcerus-group in the apical position, near the IAS, by having one internal seta on Enp3P4 (Fig. 7). F130 (XI) and D. sp (VIG2) have a P4 intercoxal sclerite with two rows of spinules in the middle and on the distal margin. Specimen 366 (XI) also has a P4 intercoxal sclerite with two rows of spinules, but the rows are interrupted in the middle of the sclerite. D16 (X) has a P4 intercoxal sclerite with two groups of a few short spinules on both sides. All these specimens have similar ornamentation characters on the P4 coxopodite. The A2 basipodite of D. sp (VIG2) and 366 (XI) are generally similarly ornamented with two rows of spinules on the caudal surface and a row of setae on the proximal part of lateral margins, but VIG2 has more spinules in each row (8 and 9) than specimen 366 (3 and 4) (Fig. 6).

4. Discussion

4.1. Three species complexes similar to D. improcerus, D. galbinus and D. versutus based on molecuclar and morphological data

Although we found Diacyclops specimens with similar morphology to three species of interest, namely endemic D. improcerus, D. versutus, and subendemic D. galbinus, we failed to identify them taxonomically at the species level. The phylogenetic analysis of Diacyclops, based on data from two mtDNA and one nuclear molecular markers, revealed three clusters according to the division of specimens into three groups based on morphological characters: the improcerus-group, the galbinus-group, and the versutus-group. Each of them comprises multiple genetic lineages. We assume that the improcerus-, galbinus-, and versutus-groups are closely related species complexes. The existence of these complexes likely explains the significant variations in diagnostic characters for D. improcerus, D. versutus, and D. galbinus, as described.

The inter-lineages genetic distances of the galbinus- and improcerus-groups correspond to known interspecies genetic distances for these molecular markers in Copepoda in general and Cyclopoida in particular (Zagoskin et al., 2014; Kochanova et al., 2021; Karanovic and Bláha, 2019; Sukhikh et al., 2020). The genetic lineages of the versutus-group are the closest to each other. Initially, the identification of F120 as a separate VII lineage from the VIII lineage was questionable. However, we identified it as distinct from the VIII lineage based on the entire length of the rDNA ITS1 (330/333 bp), although its taxonomic status remains unclear. The entire length of ITS1 can be a species-specific feature, as demonstrated in the genus Culicoides (Diptera) (Li et al., 2003). Only in the versutus-group, different delimitation methods give varying results. The GMYK method divided the VIII genetic lineage into two ‘species’ based on COI data. On the COI tree, sequences BG14, F191-5, BK21 are genetically distant from MM2, MM3, and at the same time there is no such clustering on the ITS1 tree. On the one hand, it is possible that GMYK oversplit the data (Pentinsaari et al., 2017; Luo et al., 2018). On the other hand, we note the different character of the ornamentation of the intercoxa of P4 in MM3 and BG14. This, along with the ornamentation of the basipodite of the antenna, is a crucial taxonomic feature in Cyclopoida and is mentioned in their description (Karanovic et al., 2013; Hołyńska et al., 2021). Lineage VI (F156-1, F156-3) of the versutus-group may be considered a distinct species due to its unic pattern of microcharacters in A2 basipodite and P4, as well as its formation of a separate cluster based on both molecular markers. Additionally, specimens F156-1 and F156-3 differ from all other specimens of the versutus-group in having the smallest proportion of caudal rami length and width. Although these indices, as noted for the Baikal cyclopoids, vary at the intraspecific level and may not be taxonomically significant, as in other cyclopoid genera such as Eucyclops (Rylov, 1948; Mazepova, 1978; Flössner, 1984), they enable the distinction between genetically sister D. konstantini and the versutus-group, according to the results of PCA analysis. It is also possible to use the number of setae on the second endopodal segment of A2 to distinguish specimens of the VI lineage from those of the VIII lineage.

We suggest that individual F156-2 (V) belongs to a new species that is closely related to the versutus-group species. Although PCA results based on morphometric characters do not differentiate this potentially new species from the versutus-group, it has meristic differences from the versutus-group in the armature of the third endopodite of P4, the second endopodal segment of A2, the ornamentation of the coxopodite of P4 and the A2 basipodite. All three methods used for species delimitation also separate F156-2 as a distinct species. Additionally, the entire length of ITS1 (328 bp) of F156-2 differs from the versutus-group sequences.

The three supposed species of the improcerus-group and they genetic sister species D. sp. (VIG2) overlap in PCA results, but clearly differ in both molecular markers. They are delimited unanimously by all species delimitation methods used and have differences in the ornamentation of the coxopodite of P4. Additionally they have meristic differences in the armature of the P4, which in most copepods are diagnostic of different genera (Boxshall and Halsey, 2004).

Karanovic noted similarities in the short proportion of caudal rami in the endemic D. ishidai Karanovic, Grygier & Lee, 2013, which inhabits interstitial water near ancient Lake Biwa, as well as D. improcerus and D. versutus (Karanovic et al., 2013). However, he also identified several of quantitative differences between D. ishidai and the latter two species, even based on the limited set of available morphological characters. The morphological data obtained in this study allows us to supplement these differences with meristic characters. Only lineage XI of the improcerus-group exhibits a similar armature to the third endopodite of P4, with 3 setae and 2 spines. The other lineages of this group and the versutus-group (VI-IX) differ in this feature, with 2 setae and 3 spines. In this study, all genetic lineages, including the improcerus-group and the versutus-group, were found to have the exopodal seta on the A2 basipodite, while it was absent in D. ishidae.

Most specimens of the improcerus-group, which belong to the XI genetic lineage, and some specimens of the galbinus-group, which belong to the I genetic lineage, differ from the descriptions of D. improcerus and D. galbinus only in having a smaller proportion of cephalothorax and other thoracic segments or length of antennule. Due to the telescopic nature of these features, variations in specimen fixation and storage may have affected the accuracy of their measurements, as noted for Cyclopoida (Huys and Boxshall, 1991; Karanovic and Krajicek, 2012).

Additionally, we calculated the ratio of cephalothorax length to the sum of pediger 2-5 segment length, but it is unclear if G.F. Mazepova used the same method, as she only may have only used the lengths of pediger 2-4 segments. Thus, these lineages may represent D. improcerus and D. galbinus, and the genetic and morphological data obtained could provide a basis for an integrative redescription of these species in the future. This is especially important since there is no type material available for the endemic D. improcerus, D. versutus, and D. galbinus from Lake Baikal.

PCA of morphometric indices based on linear measurements is limited in separating closely related species within species complexes. Despite their widespread use in Cyclopoida taxonomy and inclusion in species descriptions. More detailed morphometric analyses may allow for the distinction of more closely related species within each of the three complexes found. This has been demonstrated in the separation of morphologically similar species from the genus Acanthocyclops, which were considered cryptic species (Karanovic and Bláha, 2019), and in populations of Harpacticoida Bryocamptus pygmaeus (G.O. Sars, 1863) (Novikov and Fefilova, 2021). Additionally, we are confident that a detailed morphological analysis will reveal new meristic characters that can distinguish closely related species. Differences were found between specimens from different genetic lineages in the basipoditeof the antenna and the caudal side of P4. However, a large number of informative characters on the cephalothoracic appendages, legs, and caudal rami appendages in both female and male specimens remain unanalyzed. The results suggest that the diversity of the endemic Cyclopida fauna in Lake Baikal is higher than previously belived. Molecular methods have increased understanding of the species diversity of crustaceans, such as amphipods and ostracods, inhabiting Lake Baikal by detecting of cryptic species (Schön et al., 2017; Väinölä and Kamaltynov, 1999). Our study of Diacyclops found species with genetic and morphological differences, which can be used to identify them. This group requires an integrative detailed redescription, along with the description of new species.

Representatives from different genetic lineages across all three groups occur in the same samples and belong to sympatric species. The closest genetically related species that occur sympatrically are sister lineages V and VI. G.F. Mazepova also collected samples of D. galbinus, D. improcerus and D. versutus and emphasized their sympatric distribution. In each group, we observed one predominant genetic lineage, which we found in various locations and at different times. It is interesting to note that specimens from the predominant lineages (I and XI) are more similar in morphology to D. galbinus and D. improcerus. The remaining genetic lineages represented by only one or two specimens in our analysis, are rare. The abundance of three lineages in our study is probably due to their occurrence at shallow depths, starting from the water’s edge where most of our samples were taken. Specimens of rare lineages were collected from depths of 10 to 30 m.

4.2. Phylogeny of Diacyclops from Lake Baikal

Although Diacyclops is highly diverse and currently comprises over 100 species, genetic databases for this genus are extremely limited. We included sequences for non-baikalian Diacyclops in the analysis, and both molecular markers (COI and 12S) indicate that Baikal Diacyclops form a monophyletic group. Previous studies of Baikal Diacyclops, using a conserved 18S fragment, have demonstrated the monophyly of endemic Diacyclops species from Lake Baikal (Mayor et al., 2010). On the one hand, this could indicate a common ancestor for all Baikal Diacyclops analyzed, which diverged and gave rise to the observed species complexes. This scenario is possible due to the unique nature of Baikal. For millions of years, it has remained a refuge and a center of speciation for many groups of animals, including crustaceans, such as amphipods, ostracods, and harpacticoids (Timoshkin, 2001; Schön et al., 2017; Moskalenko et al., 2020). On the other hand, our dataset only includes Diacyclops species from different morphological groups (species complexes), and only Baikal species belong to the languidoides-group. The genus Diacyclops may be polyphyletic or paraphyletic, and the indentified morphological groups may represent distinct genera (Monchenko, 2000; Karanovic, 2006). These groups are distinguished based on the segmentation of swimming legs and antennules. Our 12S tree includes sequences of species from three groups: the languidoides-, the bicuspidatus- and the aticola-group. It is believed that the evolution of Cyclopoida has led to the oligomerization of appendages. As a result, the bicuspidatus- and aticola-groups are considered the most primitive groups, while the languidoides-group is more evolutionarily advanced and successful, and is one of the most specious groups (Pesce, 1994; Monchenko and Klein, 1999).

The languidoides- group and the aticola-group are represented by endemic species from Lake Baikal and Australia, respectively. The bicuspidatus-group is represented by Palaearctic species. Species from the three groups predictably formed three monophyletic groups. One interesting finding was that the bicuspidatus-group is genetically closer to the languidoides-group than to the aticola-group, despite their similar morphology. To fully comprehend the relationships of species complexes within Diacyclops, further research is necessary. This should include the study of conserved nuclear molecular markers and the inclusion of representatives from all morphological groups.

Within the monophyletic group of the analyzed Baikal endemic Diacyclops, two clusters of sister taxa were identified. The first cluster comprises D. sp. (VIG2) and the improcerus-group species complex, while the second cluster consist of D. konstantini, the V genetic lineage, and the versutus-group. Specimens of the improcerus-group belong to the smallest Diacyclops in this study. In general, among baikalian Diacyclops, they are larger than only three species: littoral D. zhimulevi, interstitial D. biceri, and D. eulithoralis. The reduction in body size likely contributes to the ecological success of the improcerus-group and is related to their adaptation to feeding on small animals, algae, or detritus. The forager group includes small littoral and interstitial cyclopoids that feed on detritus, water plants stems, epiphytic algae, protozoa, small invertebrate’s corpses (Monakov, 1998). V.I. Monchenko, used Diacyclops as a model genus to demonstrate that during the morphological-evolutionary development of Cyclopoida, there was a decrease in body size and oligomerization of thoracic appendages. The reduction in body size was the primary process. The author linked both processes to a significant reduced in energy expenditure (Monchenko, 2003). Initially, an ancestor of the improcerus-group with ancestors of the versutus- and galbinus-groups likely diverged sympatrically. Within the improcerus-group, there may have been instances of peripatric speciation. In this study, we collected specimens of genetic lineage IX of the improcerus group from the same geographical location and period as some of the specimens of lineage XI. However, the specimens of lineage IX were collected from a depth of 30 meters, while specimens of lineage XI were collected from a depth of 10 meters. It is important to note that this assumption of parapatric speciation of the IX and XI lineages requires further verification.

The most recent of the Diacyclops considered are the potential species of the versutus-group. They have the closest genetic distances according to COI and ITS1. Further investigation of this group should employ more polymorphic genetic markers, such as the nad2 gene of mtDNA, which has a higher evolutionary rate than COI in Copepoda (He et al., 2023). This group characterized by shortened, thickened antennules that do not reach the posterior margin of the cephalothoracic shield. The antennules are armed with a large number of setae and the third endopodite of P4 is armed with three spines and two setae. According to G.F. Mazepova, the presence of a large number of setae on A1 in D. versutus is associated with its habitation on soft soils, including silts. Specimens of the versutus-group were collected from various substrates, including hard substrates, sand, stones, and stones with sand. The change in antennule and the third endopodite of P4 morphology in the ancestral form of the versutus-group may have contributed to ecological success, similar to the body reduction in the improcerus-group, and provided a predatory feeding advantage. It is possible to assume that the group underwent peripatric speciation along the depth gradient. Specimens of genetic lineage VIII were found at minimum depths of up to 1 meter, while specimens of lineages VII and VI were found at greater depths of 10 and 15 meters.

The galbinus-group is the most perplexing. The validity of D. galbinus has been questioned since its description. Monchenko V.I. considered this species to be a synonym of D. moravicus (Sterba, 1956), which inhabits the karst waters of Moravia (Monchenko, 1974). Unfortunately, D. moravicus sequences are not available in the databases to compare it with the galbinus-group. Additionally, D. galbinus is the only species of the endemic Diacyclops found outside Lake Baikal, in Lake Shartlinskoye, making it subendemic. The benthic copepods of Lake Baikal include another one subendemic species - Harpacticella inopinata, which also found in the Yenisei River. The molecular study of this species from the Yenisei River has revealed its relatively recent Baikal origin, likely due to anthropogenic introduction (Fefilova et al., 2023). To assess the dispersal routes of the galbinus-group, it is necessary to analyze molecular genetic data from specimens of D. galbinus collected from Lake Shartlinskoye. Furthmore, our study shows that while the galbinus-group is a monophyletic group with the rest of the Baikal Diacyclops, it is genetically distant. It is possible that ancestral form of the galbinus-group diverged during the relatively early stages of Cyclopoida evolution in Lake Baikal and may have had a different ancestral form of Diacyclops than the other endemic Diacyclops analyzed. The study revealed the existence of four genetic lineages among specimens with similar morphology to D. galbinus. It is possible that each lineage represents a potentially new species.

Acknowledgments

We thank A.P. Fedotov (LIN SB RAS), L.V. Sukhanova (LIN SB RAS), S.V. Usov (LIN SB RAS), A.G. Luchnev, and Е. Safonov for sample collection. We are grateful to E.B. Fefilova (Institute of Biology, Komi Scientific Centre UB RAS), to M. Hołynska (Museum and Institute of Polish Academy of Sciences) and to J.P. Sapozhnikova (LIN SB RAS) for their valuable consultations and translation assistance. This study was performed in the Framework of The State Assignment No. 0279-2021-0005 (121032300224-8). The nucleotide sequences using a Nanophor 05 genetic analyzer (Sintol, Russia) and CLSM study was performed at the Instrumentation Center “Electronic Microscopy” of the Collective Instrumental Center “Ultramicroanalysis” (LIN SB RAS).

Conflict of interest

The authors declare no conflicts of interest.

About the authors

T. Y. Mayor

Limnological Institute Siberian Branch of the Russian Academy of Sciences

Author for correspondence.
Email: tatyanabfo@mail.ru
ORCID iD: 0000-0003-4425-1330
Russian Federation, Ulan-Batorskaya Str., 3, Irkutsk, 664033

I. Y. Zaidykov

Limnological Institute Siberian Branch of the Russian Academy of Sciences

Email: tatyanabfo@mail.ru
ORCID iD: 0000-0001-6669-682X
Russian Federation, Ulan-Batorskaya Str., 3, Irkutsk, 664033

S. V. Kirilchik

Limnological Institute Siberian Branch of the Russian Academy of Sciences

Email: tatyanabfo@mail.ru
ORCID iD: 0000-0002-9997-6294
Russian Federation, Ulan-Batorskaya Str., 3, Irkutsk, 664033

References

Supplementary files