Insects belonging to the Odonata order have been studied from the perspective of ecological indicators, and many studies show that certain species exhibit high association with particular habitats (Smith et al. 2007, Gómez-Anaya and Novelo-Gutiérrez 2010). Presence or absence of certain groups of species can be looked as a proxy for habitat variables or to assess quality of freshwater habitats (Subramanian and Sivaramkrishnan 2005, Subramanian et al. 2008). These insectivorous insects have been looked also as bio-control agents against mosquitoes (Andrew et al. 2008).The order Odonata includes dragonflies and damselflies, separated into two suborders, namely Anisoptera (dragonfly) with 12 families and Zygoptera (damselfly) with 24 families. Anisozygoptera, a previously recognized third suborder, has been merged with Anisoptera in recent revision of the taxonomy of Odonata (Bybee et al. 2008, Dumont et al. 2010). About 6,000 species of Odonata and subspecies belonging to 652 genera have been documented world-wide (Schorr and Paulson 2014). India harbors 474 species and 50 subspecies belonging to 142 genera spread across 18 families (Subramanian 2014). The taxonomy of Indian Odonata is well worked out and descriptions are available for almost all the reported species (Fraser 1933a, 1934, 1936; Davies and Tobin 1984, 1985; Prasad and Varshney 1995; Subramanian 2014).

After Fraser's seminal work (Fraser 1933a, 1934, 1936) on Odonata of India, there was a gap of almost 50 yr in studies on Odonata across the country. After establishing the Zoological Survey of India in 1916, trained taxonomists started collecting data and publishing lists of Odonata of localities or regions. In spite of this effort, Odonata were largely neglected due to the lack of awareness and difficulty in field identification. Introduction of field-guides (such as Emiliyamma et al. 2005, Subramanian 2005, Andrew et al. 2008, Nair 2011, Kiran and Raju 2013) has recently accelerated process of data collection on Odonata. Advent of open access public forums and websites (such as Asia-Dragonfly 2014, DragonflyIndia 2014, IBP 2014) has further increased data deposition in public domain and authentication of records. As species records often vary on spatiotemporal scale, creating checklists and updating them become crucial to understand species distribution dynamics and possible threats to them. Maharashtra State lies in central-west India. It is one of the Odonata species rich states, given the variety of habitats it supports, owing to its unique geographical position. Researchers from the Zoological Survey of India and other academic institutes have often surveyed various parts of Maharashtra. Their studies have been mostly localized to a small area. Most of the endemic Odonata of Western Ghats of India are habitat sensitive, restricted to a narrow ecological space (Subramanian 2007; Subramanian et al. 2011; Koparde et al. 2014, 2015).Therefore, updating species presence data is continuously needed to devise specific conservation strategies. Only locality based information by Fraser (1924, 1933a, 1934, 1936) was available for Maharashtra until the first checklist of the state, including 46 species was published by Prasad (1996). This list was further updated by Kulkarni et al. (2012) increasing the total count of species to 99, including subspecies. Although Kulkarni et al. (2012) took into consideration a larger collection sampled across most of the state; they failed to incorporate recent field studies and authenticated data in public domain. Moreover, most of their studies were scattered throughout the state with very few studies in Western Ghats, which is a hotspot for endemic Odonata (Subramanian 2007). This article updates the current checklist of the state based on recent field studies, published literature and authentic data mined through social media; and discusses Odonata distribution and geographical gaps in Odonata surveys in Maharashtra.

Materials and Methods

Results

The compilation of field studies and data from literature resulted in 134 species of Odonata, including 87 species of Anisoptera and 47 species of Zygoptera from Maharashtra. They belong to 70 genera representing 11 families (Table 2). Libellulidae was represented by the highest number of species (48), followed by Coenagrioniidae (27) and Gomphidae (22). Both Euphaeidae and Cordulidae were represented by a single species. Twenty-two species were Data Deficient, and 11 were not listed in International Union of Conservation for Nature (IUCN) red-list of threatened species. Two species viz. Indothemis carnatica and Heliogomphus promelas were listed as Near Threatened. Around 50% of Data Deficient species in the list belong to Gomphidae (11 species), followed by Macromiidae and Platycnemididae (three species each). Twenty five species of Anisoptera and 10 species of Zygoptera were added to the State checklist. The new Anisoptera species belong mainly to Gomphidae (nine species), Libelluliidae (eight species), and Aeshnidae (four species). The 10 new species of Zygoptera, belong to Coenagrionidae (four species), Platycnemididae (four species), Euphaeidae (one species), and Lestidae (one species). In addition, 16 species have been recently recorded in single studies. Thirteen of them belong to Gomphidae (9) and Macromiidae (4). During the field-studies, two species recorded were not known from the State: Anaciaeschna jaspidea and Anax imperator.

Discussion

Maharashtra supports high faunal diversity owing to its geographic position and the biogeographic zones it covers. Given the variety of macro- and micro-habitat types, it was expected that the State supports high number of Odonata species. From recent surveys and data mining, we have added 35 species to the previous list by Kulkarni et al. (2012) that included 99 species. Kulkarni et al. (2012) had counted some of the subspecies, such as Libellago lineate lineate and Libellago lineata indica or Aciagrion hisopa hisopa and Aciagrion hisopa krishna as different species in their species checklist. However, we confined our identification till species level. Recording subspecies on field is highly difficult. Hence, including subspecies in the checklist may introduce error. This difference of 35 species between previous (Kulkarni et al. 2012) and current checklist is mainly due to incomplete on-field sampling by previous researchers, which was scattered throughout the State. They undersampled various biogeographic provinces in Maharashtra. Moreover, difficulty in collection, unapproachable terrain in certain areas such as Western Ghats (mountains) or Central highlands, and limitations due to resources and expertise might have resulted in such a gap. Also Kulkarni et al. (2012) failed to incorporate records other than those published by the Zoological Survey of India. However, this work along with Prasad (1996) has been instrumental in providing the first exhaustive checklist of Odonata of Maharashtra State. Our sampling in addition to sampling done by previous researchers almost spanned the State. The sampling was not systematic and spread across seasons, because concentrated mainly in postmonsoon season when Odonata activity is at peak (Kulkarni and Subramanian 2013). The data collection was a collective effort, and sampling was highly limited due to resources and expertise. Therefore, although the current checklist significantly updates the previous ones by Prasad (1996) and Kulkarni et al. (2012), it may not be interpreted as a complete checklist of Odonata of Maharashtra.

Certain biogeographical areas such as theWest coast or the Malabar plains of Western Ghats have been underrepresented in this and previous studies. The Deccan south region was not sampled at all. The Deccan peninsula-central plateau which represents the largest area of Maharashtra was fairly well-represented in all the studies, except central Maharashtra, dominated by scrub-forest and dry-deciduous forest, for which there is a serious lack of data. Similarly, the Western Ghats (mountains) have not been sampled exhaustively during the study period, despite the fact that they are also areas of high endemism (Myers et al. 2000; Subramanian 2007; Subramanian et al. 2011). This region which is rich in evergreen and semi-evergreen forest patches, even though fragmented, has been highly underrepresented in samples. Out of 74 localities from where data were compiled, only four localities represent evergreen forest areas (Table 1). The northern part of Western Ghats of Maharashtra has been also undersampled. This undersampling might be the root-cause of lack of data on species numbers and distributions (Koparde et al. 2014). It seems that most of the data on Odonata diversity from Maharashtra comes either from West Maharashtra or East and North-east Maharashtra (Fig. 1). From Central-north and North-east Maharashtra, Satpuda mountain ranges have been undersampled, even if biogeographically important areas (Hora 1949, 1953; Auden 1949; Daniels 2001; Karanth 2003). Species distribution data from these areas should be important in answering questions related to the biogeography of Indian peninsula and/or the Indian subcontinent. Such studies have been carried out using Odonata as model systems (Dijkstra 2007, Shah et al. 2012), underscoring the importance of spatial data from these regions.

Table 1.

Details of the study localities

Continued

Fig. 1.

Map of the surveyed localities (Hijmans et al. 2005b).

Table 2.

Checklist of Odonata of Maharashtra state.

Continued

Continued

Table 3.

Species records thatare doubtful in terms of either taxonomic identity or spatial distribution

Sathe and Bhusnar (2010) have listed many species, especially Gomphidae family members, which are not included in the previous literature by Fraser (1933, 1934, 1936), Prasad (1996) and Kulkarni et al. (2012) (Table 3). Recent studies by Tiple (2012a,b), Tiple et al. (2013), Kulkarni and Subramanian (2013), Koparde et al. (2014, 2015) from Maharashtra or even those from Western Ghats of Goa (Rangnekar et al. 2010, Subramanian et al. 2013, Rangnekar and Naik 2014), which is a neighbouring State, failed to record these species. Sathe and Bhusnar (2010) have listed Microgomphus longistigma (Table 3) which is most probably Merogomphus longistigma. Similarly, Orthetrum caledonicum recorded by Wankhede et al. (2012) in Pune district is not a valid species. Recent studies by Kulkarni and Subramanian (2013) and Koparde (P. K., unpublished data) in the same district did not record this species. The species list of Amba RF, that lies in Western Ghats of Maharashtra, by Aland et al. (2012) also includes four species which have not been recorded by earlier researchers or during recent studies (Table 3). This might be because these areas were not surveyed earlier. However, authors in these articles do not mention anything specific about these species, i.e., new records to Maharashtra State, unusual sightings or taxonomic uncertainties. Koparde et al. (2014, 2015) had done a short-survey in areas around Kolhapur and Amba RF, however, they did not record species mentioned by Sathe and Bhusnar (2010) and Aland et al. (2012). Koparde et al. (2014, 2015) studies were shorttermed, specifically in postmonsoon season, which might be a reason that they could not detect many Gomphidae. However, this study was more extensive than other studies in the same region. We failed to retrieve these species even from public data. Although it is difficult to assess their authenticity, while compiling the state checklist we have retained the species which have been recorded by Sathe and Bhusnar (2010) and Aland et al. (2012), considering that they had been probably undersampled by other researchers.

Most of our additions belong to the family Gomphidae. This family is also represented by the highest number of Data Deficient species as well as species for which information is not available in the IUCN redlist of threatened species (Table 1). The members of this family are fastmoving insects and may have crepuscular habits. These insects are difficult to observe or collect. Many Gomphidae are already rare. Therefore, there are high chances of not detecting them during surveys. Microgomphus verticalis and Cyclogomphus vesiculosus are the only two species of Gomphidae that have not been recorded recently; whereas 10 species have been recorded only recently (Table 2). This explains the huge gap in knowledge on the distribution of Gomphidae. Although the spatial distribution may vary in time, addition of only 10 species over almost 17 yr indicates slow rate of data acquisition on Gomphidae. Lack of recent records and systematic information on population occurrences has been discussed as a major fallout in assigning conservation status to Gomphidae by many IUCN red-list assessors (Dow 2009a,b,c; Sharma 2010; Kakkasery 2011a,b; Subramanian 2011). This trend can also be observed in Platycnemididae. Ten species of Platycnemididae are known from Maharashtra, five of which have been recorded recently. The cases of the Macromiidae and Corduliidae families are similar. These insects are highly habitat sensitive, localized, and fast moving (Subramanian 2005, Koparde et al. 2015). Despite the records of Idionyx species from Goa (Rangnekar et al. 2010, Subramanian et al. 2013, Rangnekar and Naik 2014), there is a complete lack of data on these species from Maharashtra except one unidentified species recorded by Koparde et al. (2015). Species from Euphaeidae and Platycnemididae (Zygoptera) are habitat sensitive and localized in small areas (Koparde et al. 2015). There is considerable lack of knowledge on the distribution and ecology of these species. This lack of data can be attributed to the lack of recent records and incomplete sampling (Koparde et al. 2014). Although the current checklist provides presence data, the absence of species from the old checklists (Prasad 1996, Kulkarni et al. 2012) can not be considered as total absence, given spatiotemporal variation and bias in data collection. These species breed in—and are closely associated with—unpolluted rivers and streams (Subramanian 2011). They have been observed to be associated with dense forests (Koparde et al. 2015). Species which are known from old literature and have not been detected recently might have even gone locally extinct due to habitat degradation and loss. The probable major causes of this extinction could be loss of habitats by expanding urbanization along with large scale climatic changes. Urban development is expected to have a deleterious impact on Odonata by reducing the area of natural habitats. The quality of residual habitats may also be adversely affected by various forms of pollutants (Subramanian et al. 2011,Tiple et al. 2013).

There are few species, such as Protosticta graveleyi, which are known from old literature and from records in public domain. It seems that, because sampling of Odonata has been sporadic overtime; the possibility of species gone extinct could be attributed to sampling artifact. An interesting case is Onychothemis testacea and Zygonyx Iris. Both species inhabit fast-flowing streams, water-falls and probably areas surrounded by dense forest (Subramanian 2005, Nair 2011). It is difficult to collect or even photograph these species in such areas. There is a record of O. testacea by Prasad (1996), after which Koparde et al. (2014, 2015) have recorded it from two different localities. Similarly, Z. iris has been recorded from Sindhudurg (Koparde et al., unpublished data) and Chandoli NP (Koparde et al. 2015). Another interesting finding is that of Elattoneura nigerrima. This species was underrecorded after Prasad (1996) study. We retrieved many other spatial records of this species across Maharashtra from public domain. It seems that this underrecorded species is widespread, but patchily distributed in Maharashtra (Koparde et al. 2015; Koparde et al., unpublished data). This probably explains artifact of sampling and usefulness of crowdsourcing in data collection.

Advent of field-guides and public forums has driven to the next level. Continuous data sharing among researchers through social networks has led to free flow of information and site and/or speciesspecific studies. However, such public forums often suffer from deposition of non-authenticated records and false presence data. If checklists of regionsmade by experts are referred alongwith records in public domain, they may result into usable species data. Field data collected by experts is of primary importance to understand changing species distributions and the causes of this change. Additional systematic field-studies across Maharashtra State covering all possible microhabitats, will provide insights into species richness and threats to them. Establishing the current checklist of Odonata needs an effort to gather species presence data across an extensive landscape. This checklist is highly likely to get modified asmore data flow in across the State.