A contribution to the earthworm diversity (Clitellata, Moniligastridae) of Kerala, a component of the Western Ghats biodiversity hotspot, India, using integrated taxonomy

A Abstract A contribution to the earthworm diversity (Clitellata, Moniligastridae) of Kerala, a component of the Western Ghats biodiversity hotspot, India, using integrated taxonomy. Earthworms (Clitellata, Moniligastridae) of Chaliyar River Malappuram, Eravikulam National Park, Neyyar Wildlife Sanctuary, Parambikulam Tiger Reserve, Wildlife Sanctuary, Kerala, a component of the of Ghats, India, were studied by the standard method of taxonomy, and their DNA barcode signatures using the mitochondrial gene cytochrome c oxidase I ( COI ) were generated for the first time. This study represents eleven species of earthworms of the family Moniligastridae: Drawida brunnea Stephenson, Drawida circumpapillata Aiyer, Drawida ghatensis Michaelsen, Drawida impertusa Stephenson, Drawida nilamburensis (Bourne), Drawida robusta (Bourne), Drawida scandens Rao, Drawida travancorense Michaelsen, Moniligaster aiyeri Gates, Moniligaster deshayesi Perrier, and Moniligaster gravelyi (Stephenson). In the phylogenetic analysis all the species were recovered in both neighbour–joining (NJ) and maximum likelihood (ML) trees with high clade support. The average K2P distance within and between species was 1.2 % and 22 %, whereas the clear barcode gap of 2–5 % was suggested by barcode gap analysis (BGA) of studied species, reflecting the accuracy of characterization. The study presents the first step in the molecular characterization of the native earthworm family Moniligastridae of India.


Introduction
Moniligastridae is a family of earthworms indigenous to southeast and eastern Asia. It is believed that the family Moniligastridae originated in the Malaya Archipelago's geographical region (Gates, 1972;Blakemore, 2014), but later Jamieson (1977) suggested an origin near Myanmar. Its natural range encompasses south, southeast and east Asia, from peninsular India to Japan through Myanmar, China, the extreme southern portion of far Eastern Russia, Korea, the Philippines, Borneo, and Sumatra (Gates, 1972). Moniligastrids are dominant members of the earthworms fauna in India especially in the South and North East Regions. Three genera Desmogaster Rosa, 1890; Drawida Michaelsen 1900; and Moniligaster Perrier, 1872 are known from India (www.earthwormsofindia.com). Among them Drawida is most diverse with 73 species in India. Earthworms of this family have drawn the attention of earthworm biologists as they retain the single layered clitellum characteristic of Clitellata other than earthworms (Crassiclitellata) yet function ecologically as do the crassiclitellate earthworms. The Moniligastridae have a broad size range, just like earthworms sensu stricto. The family is characterized by simple pointed setae, four pairs per segment, a clitellum beginning on segment 9 or 10 and extending over 3 to 10 segments, including those bearing genital pores; male pores one pair (Drawida, Moniligaster) or two pairs (Desmogaster) in or near grooves 10/11, 11/12 or 12/13; female pores one pair in 11/12 or XII or XIV. The spermathecal pores are one or two pairs in 7/8 or 8/9 or 7/8 and 8/9; the oesophagus with two gizzards anterior to X or two to ten gizzards at the beginning of the intestine. The last hearts are two segments in front of the ovarian segment; they are holonephridial. Testes and funnels one or two pairs enclosed in one or two pairs of testis sacs. Vasa deferentia opening into prostate glands. One pair of ovaries in the segment immediately in front of the groove or segment on which the female pores are situated, one pair of ovisacs extending backwards from the ovarian segment. One or two pairs of spermathecae with long tubular ducts. Without typhlosole, calciferous glands, supra-intestinal glands and seminal vesicles.
The use of these morpho-anatomical characteristics has often been a barrier to the identification of these earthworms, leading to imprecise identification of taxa. The present study is the first attempt to provide a means for rapid assessment of some moniligastrids by using molecular data. The study was performed in the Kerala state, as half of its area falls under the Western Ghats, one of the world's eight most important biodiversity hotspots (Myers et al., 2000;Mittermeier et al., 2011). Therefore, the present study aimed to assess the earthworm diversity and the phylogenetic relationship of some moniligastrids with the use of DNA barcodes as a standard genetic marker for identification of earthworm species of Kerala.

Study site
Kerala is a small state in the south-western tip of India. It is a narrow strip of coastal plain that borders the Arabian Sea from the north to south, next to the neighbouring states of Karnataka and Tamil Nadu. The state is recognized for its lush greenery, highly dense forests, diversified ecological habitats, topography, and the high biodiversity. It is bounded by the thickly wooded and forested hills of the Western Ghats to the east and the Arabian Sea to the west. Kerala occupies 38,863 sq. km and comprises approximately 1.18 % of India's landmass (Sreedharan, 2004). Out of the total length of the Western Ghats, Kerala covers around 600 km. Nearly 56 % of the total geographical area of the state has an annual average temperature ranging between 31 to 37 o C and annual rainfall of 3,500 mm, mainly due to the windward location to the Ghats (Rao, 1976). Due to the integration and combination of different climatic conditions, like warmer climate, altitudinal variations, two different rainfall patterns and seasons (Southwest monsoon and North-East monsoon), several soil types and agro-ecological zones, Kerala has a variety of macro environments that vary from tropical rain forests to hot dry deciduous forests. These diversified habitats and local ecological niches contributed to a variety of macro and micro environments conducive for a variety of flora and fauna requiring contrasting environment. Of the biota of India, the state sustains over 24 % of the plant species, 30 % of the animal species, and 35 % of the freshwater fish species (Sreedharan, 2004).

Collection of earthworm samples
Earthworm samples analysed in the present study were collected from different sampling sites in Kerala ( fig. 1; see also the dataset published through GBIF (Doi: 10.15470/l2nlhz). The locations, species names, coordinates, and their BOLD accession numbers are provided in table 1. Samples were collected by digging and hand-sorting according to the method described by Satchell (1969). The specimens were anesthetized in 30 % (v/v) ethanol. Small pieces of muscle tissue from the tail region were then cut and preserved in 100 % (v/v) ethanol solution for molecular investigation. Next the earthworm samples were fixed in 10 % (w/v) formalin for morphological identification. 100 % ethanol preserved tissue of each sample was placed in the Museum of Dr. Harisingh Gour Vishwavidyalaya (A Central University), Sagar, Madhya Pradesh, India as reference.

Sample management and morphological classification
Prior to applying the molecular technique for evaluation, we identified earthworms on the basis of specific diagnostic morphological characters under a stereoscopic zoom microscope (Leica Model No. M60) using the available literature (Stephenson, 1923;Aiyer, Fig. 1. Study area with distribution of Moniligastrids in Kerala. Kerala. 1929;Gates, 1972;Julka, 1988;Narayanan et al., 2016Narayanan et al., , 2017. A camera lucida was used for drawings and abbreviations: spp, spermathecal pore; mp, male pore; atr, atrium; spd, spermathecal duct; amp, ampulla; tss, testis sac; vd, vas deferens; prs, prostate; atrgl, atrial gland were used in the figures. Voucher specimens are housed in the Museum, Department of Zoology, Dr. Harisingh Gour Vishwavidyalaya (A Central University), Sagar, Madhya Pradesh, India.

DNA sequencing
For DNA sequencing the small pieces of muscle tissue from tail region were used. A total of 28 samples of moniligastrids were sent to Barcode of Life Data Systems (BOLD System), Biodiversity Institute of Ontario, University of Guelph, Canada (Ratnasingham and Hebert, 2007) following appropriate protocol to obtain DNA sequences, accession numbers and Barcode. All the data used in present study is available on BOLD website under the project entitled 'Diversity studies in earthworms of India' (IEW). In addition, 28 COI sequences were retrieved from the NCBI and BOLD free public domain for molecular analysis (see table 2 for more details).

Sequence alignment and data analysis
The alignment of 56 COI data matrix [28 COI from this study (table 1) and 28 additional from NCBI and free public domain of BOLD (table 2) were analysed in MEGA X (Kumar et al., 2018). For intraspecific and interspecific genetic distances within and between species, the best substitution model i.e., Kimura two-parameter (K2P) (Kimura, 1980) was used. For phylogenetic analysis we used the COI matrix to generate the neighbour-joining (NJ) and maximum likelihood (ML) trees using MEGA X, with a graphic depiction for the evolutionary distances between species. Node robustness was inferred with 1,000 bootstrap replicates. The barcoding gap analysis was also inferred in the barcode gap analysis (BGA) tool available on BOLD, but only 28 COI sequences generated in this study were used. Similarly, the calculation of sequence composition was computed in MEGA X. Deep pigmentation or without pigmentation, clitellum on X-XIII segments, one pair of male pores in 10/11, female pores in 11/12, spermathecal pores in 7/8. Two to eight gizzards at the beginning of the intestine. Last heart in IX. Dorsal pores usually absent. One pair of tes-

Results
The earthworm species collected and identified from the study area are arranged in alphabetical order. Each entry gives the information in sequence: earthworms' scientific name, material examined, sample ID with accession number(s), collection site, description of species. Brief descriptions of the genera are also given.
tes and funnels enclosed within setal sac which project from septum 9-10 into segment X or segment IX and X. Prostate of various forms, ovaries XI, this segment may be reduced to a special ovarian chamber of characteristic form, one pair of ovisac projecting backwards from septum 11-12. Spermathecae with or without atrium like dilation (diverticulum, function unknown) at the ectal end, without stalked glands. Penial setae and copulatory setae absent. Genital markings rarely found. Stephenson, 1915 Material examined: KERL269A2 Sample ID with BOLD accession number: KER-L0269A2 (ADH0514) Collection site: Peppara Wildlife Sanctuary (8º 37' 14.7'' N, 77º 10' 03.3'' E), Thiruvananthapuram, Kerala, India. Date of collection: 25 X 2015 Diagnostic features: length 70 mm; diameter 5 mm, segments 180; body short and relatively broad, dorsoventrally flattened. Prostomium prolobus. Setae small, closely paired, aa < bc and dd = ½ circumference. Male pores eye shaped bordered by prominent shape midway between b and c. Female pores nearer b. Spermathecal pores in cd, close to c. Septa 5/6-8/9 thickened. Three gizzards in XIII-XV, the first less firms than others. Testis sacs large kidney-shaped occupies in 9/10 more into X, vas deferens short without coiling; prostate opaque white, spherical/ovoid with short moderately thick stalk, smooth but no muscular iridescence. Ovarian chamber with its roof at the dorsal parietes, funnel extends upwards on each side of the gut nearly to mid-dorsal line, ovisacs in XII. Spermathecal ampulla ovoid, atrium mammillary in shape sessile on parietes joined by the slightly coiled duct at base ( fig. 2).

Drawida brunnea
Drawida circumpapillata Aiyer, 1929 Material examined: KERL273A5 Sample ID with BOLD accession number: KER-L0273A5 (ADH2327) Collection site: Eravikulam National Park (10º 06' 56.9'' N, 77º 05' 20.5'' E), Kannan Devan Hills, Kerala, India. Date of collection: 29 X 2015 Diagnostic features: length 40-60 mm; diameter 3-5 mm, segments 127-150. Colour light grey. Dorsal pores absent. Setae closely paired; aa < bc; dd equal or lesser to ½ circumference visible from III segment. Clitellum dark brown saddle-shaped in X-XIII. Nephridiopores in cd visible from IV. Male pores between b and c nearer to b, minute on a conical elevation in the centre of a large circular papillae, sometimes almost touch each other in the mid-ventral line. Each papillae extends outwards about half bc of setal zones both in X and XI segments. Female pores in ab. Spermathecal pores in d (sometimes in ab) slit-like aperture. Septa 5/6-8/9 thickened. Three gizzards in XII-XIV. Testis sacs large ovoid sacs in X, tapered towards posterior end. Vas deferens lies on the anterior face of septum 9/10 in segment IX, where it twines round the heart and enters the prostate on its anterior side near the ectal end. Prostate large club-shaped, furry or papillose, densely covered with large granulated gland cells. Ovarian chamber present, ovisacs very long extending backwards through eight and ten segments. Spermathecal ampulla sac-like in VIII, atrium digitiform in VII, duct loosely colled entering its ectal end ( fig. 3).   Drawida impertusa Stephenson, 1920 Material examined: KERL269A6, KERL269A8, KER-L274A2, KERL274A4, KERL274A6 Sample ID with BOLD accession number: KER-      Setae closely paired, large and prominent especially in the ventral bundles of III-XII; aa = bc or in the anterior part of the body is rather greater; dd = ½ circumference Male pores two pairs, the anterior in 9/10, rather outside the line of setae b, on a median transverse somewhat dumbbell-shaped cushion, extending on each side to between the line of b and c; posterior male pores on 10/11, in line/just outside of setae b, in the antero-lateral angles of a thickened median patch which occupies the ventral surface (almost entire) of XI. Female pores in 11/12 between the lines a and b. Spermathecal pores in ab or slightly lateral to b. Septa 6/7-8/9 considerably thickened, 5/6 thin, 9/10 and a few following slightly thickened. Two gizzards in XIII and XIV, sometimes three in XIII-XV. Testis sacs extending into IX and X. Prostate two pairs in IX and X, elongated, cylindrical or pear-shaped, surface soft, minutely papillated. No ovarian chamber, spermathecal atrium ovoid and sac like, duct entering towards the ectal end ( fig. 8).  . 9).     from 11/12. Spermathecae with a bifid muscular atrial chamber ( fig. 10), each horn of which bears a lobulated glandular mass. Genus differs from Drawida in the presence of two horns of atrial chamber (one pair of glands discharging by its own canal into a common duct; fig. 10) and mostly found in South India.
Moniligaster aiyeri Gates, 1940  Spermathecal pores transversely placed slits in cd or d. Septa 6/7 is thickly muscular, 7/8-8/9 very thickly muscular. Three-five gizzards in XVI-XXII. Testis sacs smaller than the cluster of vas deferens which is much thicker, very long, loop of slender portion in IX and X, thickened portion in a cluster of loops in IX which is larger than the testis. Prostate large mushroom-shaped, muscular, thick, opaque and lined internally with white material which is raised into irregular ridges. Ovarian chamber present, ovisacs with thick wall extending into XVI. Spermathecal duct is 20 mm long and passes into the base of posterior atrial gland. Two atrial glands with the usual mammillated surface. Ectally two glands appear united but can be separated after removal of the investing tissues, the common duct of atrial glands is thick, short, lateromesially flattened ( fig. 11).    extending back through several segments. Spermathecae with ovoid ampulla and coiled duct, which joins a bifurcation of the atrial gland in VII, which is large, bifid, each half compact and rounded with a yellowish mammillated surface, the stalk of the two halves unite to form a common duct ( fig. 12).

Moniligaster deshayesi
Moniligaster gravelyi Stephenson, 1915 fig. 14). The phylogenetic analysis showed full support for each taxon and there were minor differences in topologies of NJ ( fig. 15) and ML ( fig. 16) trees. In addition, all the species represented separate clades and were fully recovered on both NJ and BI trees.

Discussion
Earthworms are one of the most valuable soil animals but their characteristics and burrowing nature makes their taxonomy difficult, resulting in a massive underestimation of the true level of earthworm diversity (Sket, 1999). The DNA barcodes combined with the classical morpho-anatomic screening, the integrative approach, would aid in better estimate earthworm diversity (Dayrat, 2005;Lone et al., 2020). The reliability of the DNA barcode as a data source for species delimitation depends on the barcode gap, which is a marked discontinuity between the values of intraspecific and interspecific divergences. A clear break between intraspecific and interspecific divergences improves confidence of species delimitation and identification (Hebert et al., 2004;Meier et al., 2008). In our study, BGA suggested a barcode gap of 2-5 % with no overlap, which support the accuracy of DNA barcoding to delimit moniligastrid taxa. Furthermore, in DNA barcoding, species are distinct from their nearest neighbours (NN) if their maximum intraspecific distance is less than the distance to their nearest neighbour sequence (Ashfaq et al., 2014). Similar observations were seen in our report (table 4) where all the intraspecific distances were less than to their nearest neighbour (NN), which further support DNA barcoding for species delimitation of earthworms. The interspecific K2P genetic distances among the moniligastrid species ranged between 5.6 % and 34.0 % ( by D. japonica and D. gracilis 33.9 %. The highest K2P intraspecific distance was found in D. ghilarovi. This could be explained by the presence of different morphs and therefore a species complex (Blakemore et al., 2014). Moreover, except for D. japonica and D. ghilarovi, all the studied moniligastrids had low intraspecific genetic distances. Moreover, the NJ and ML phylogenetic trees' topologies further confirmed that these eleven species were distinct evolutionary units. All species were recovered as clade. However, it appears that Moniligaster may need revision, as the representatives used here nested within Drawida. In recent times, the use of molecular identification has increased, along with morphological characteristics, to differentiate earthworm species and to detect cryptic species (Decaëns et al., 2013). Progress in molecular taxonomy of earthworms of India has been ad hoc. The fact that we do not know the full systematic inventory the soil biota generally, and especially the key earthworms, is an oversight that urgently needs to be redressed. Each ecosystem is composed of populations of species and each species has its particular ecological requirements and responses. Earthworms in general are keystone animals in nutrient cycling processes due to their role as major detritivores (i.e., feeding on dead and decaying matter, including dung). Correct species identification is important to accurate ecological study that we need as a first step to fully understanding, appreciating and utilizing this natural and national resource appropriately. DNA barcoding is an indispensable tool for its speed and accuracy, but only after the initial vouchers are correctly identified. Thus, the present study has been taken to provide a baseline status survey of ear-thworm diversity using molecular tools. Identification using molecular data helps to morphologically identify variable individuals of the same species, juvenile specimens and cocoons. In the case of earthworms, the classical standards of taxonomy are mainly based on genital structures, while the collection of the sexually developed worms is uncertain. Therefore, the generation of a DNA database of sequences of earthworms is advantageous and inevitable. The study presents eleven earthworm species belonging to two genera of the family Moniligastridae, represents about 55 % of this family (20 species) (Blakemore, 2007;Narayanan et al., 2016) and about 2.5 % (435 species/subspecies) of the country (Thakur and Yadav, 2018). It is noteworthy that Kerala possesses about 50.5 % of the earthworm diversity found in the Western Ghats biodiversity hotspot (Julka et al., 2009;Nair et al., 2010). The giant earthworm species Drawida nilamburensis about 75 cm body length, native to Kerala, is reported in this study.
The study explored eight different sites in the State of Kerala. Other regions, especially western and south-western parts of the state remain unexplored and need to be investigated further. All recorded species were endemic, indicating undisturbed habitats within the protected areas of Kerala. Moreover, of the eleven earthworm species reported in this study, COI sequences of only three species (D. ghatensis, D. impertusa and D. travancorense) are available in the NCBI database, which indicates limited availability of molecular analysis and assessment of earthworm diversity in Kerala. Some recent reports on earthworm diversity are available, however, based on morphological observations (Narayanan et al., 2016(Narayanan et al., , 2017. The present study fills the gap of molecular taxonomy of earthworms, with 28 COI datasets of endemic earthworms of Kerala having been generated. The findings may serve as a reference library of genomic signatures of earthworms in the study area, providing data for taxonomic assessments, phylogeny, molecular identification, dispersal, and population dynamics. Yet, in total of 83 species of moniligastrids are known from India, but none ones given here are barcoded and published. An integrative taxonomic classification, incorporating morphological classification, DNA barcoding and phylogeny is especially useful when working on a huge variety of endemic fauna, where records are scanty. Furthermore, threats to biodiversity have already raised alarms with respect to the conservation of biodiversity and much emphasis has been given to endemic species. Considering the high diversity of earthworms (e.g. Blakemore, 2000;Chang et al., 2008), biomass (e.g. Brockie and Moeed, 1986) and ecosystem functioning (Boyer and Wratten, 2010) are likely to be threatened with extinction. The genomic signature of these species may not only delimit earthworm species but could be used in various other fields such as conservation strategies, toxicological research, and bioremediations.