Diversity of scorpions (Arachnida: Scorpiones) in Polonnaruwa Archaeological Reserve, Sri Lanka

their productive comments and suggestions. Abstract: Sri Lanka harbours 20 scorpion species belonging to four families, of which 15 are endemic. The distribution and ecology of scorpion fauna in Sri Lanka is poorly known. In this study, we surveyed the diversity of scorpions in the Polonnaruwa Archaeological Reserve in the dry zone of Sri Lanka. Microhabitats were thoroughly observed using the direct visual encounter method and UV lights from July to November 2018 for about seven hours (19.00–02.00 h) by two to three observers. Species, abundance, age/sex, and microhabitat features were recorded. Diversity indices, including α-diversity and β-diversity, were calculated. Heterometrus swammerdami was the most abundant species recorded, while Isometrus thwaitesi was the rarest. Reddyanus loebli and R . besucheti were common in both open and forest habitat types. Charmus laneus was recorded for the first time in Polonnaruwa. The highest Shannon Index and Margalef Diversity Index values were recorded in open habitats, but species evenness was low compared to forest habitats. Sørensen index values showed a 58% species similarity between two habitats. The results presented here contribute to the knowledge of the diversity of scorpions in these historically significant sites. This can serve as a basis for future research on the impact of habitat modification and fragmentation on populations, distribution and ecology of scorpions.

The spatial distribution of scorpions is influenced by a range of climatic and environmental variables such as temperature, rainfall, elevation, slope, soil properties, vegetation type and land cover (Polis 1990;Prendini 2005). Sri Lanka has distinct types of habitats, including rain forest, dry mixed evergreen forest, montane forest, and shrub forest, which support scorpions (Ashton et al. 1997). Most scorpion species are distributed through the dry zone, and few are found in the wet zone of Sri Lanka (Kovařík et al. 2016). The objective of this study was to assess the diversity of scorpions in an archaeological reserve located in the ancient city of Polonnaruwa, in North-central Province, Sri Lanka, as a conservation initiative for scorpions. Additionally, the study aimed to provide important information on population structure (age/sex ratio), microhabitat preference, and community-level characteristics (species richness and diversity in two selected habitats). Since Polonnaruwa is a well-preserved historic site and tourist attraction, this study is relevant to the impact of tourism on the conservation of biological diversity.

Study site
This study was carried out in the archaeological reserve in Polonnaruwa ancient city (7.9584N & 81.0027E) located in North-central Province, Sri Lanka, from early July to late November 2018. The selected study site with an area of 7.9km 2 was an isolated secondary forest patch consisting ancient monuments dating back to King Parakramabahu in the 12 th Century, and surrounded by human settlements. We have divided the study area into two habitat types: open habitat and secondary forest (Image 1). Open habitat predominantly consists of ancient monuments maintained by the Central Cultural Fund, Sri Lanka, with scattered trees. Some parts of the open habitat encompass exposed bedrock with boulders, and the soil type is sand to gravel particle-sized soil with low/no leaf litter (Image 2a). Secondary forest habitat consists of a dry mixed evergreen forest dominating by Cassia marginata (Fabaceae), Manilkara hexandra (Sapotaceae), Drypetes sepiaria (Putranjivaceae), and Ficus sp. (Fabaceae), tree species (Abeynayake et al. 1993) and scattered amidst shrubs and herbs (Image 2b).

Survey
A pilot study was carried for two days in early July for habitat selection and species identification before the survey. All possible microhabitats, including both terrestrial and arboreal, were thoroughly observed using the direct visual encounter method with the aid of UV lights. Sampling was carried out by two to three observers and lasted for about seven hours (19.00-02.00 h). A total of 78 human hours were spent equally for open and forest habitats (39 human hours per each habitat). Abundance and age-sex classes were recorded as male, female, or juvenile. But burrowing scorpions were not classified into age/sex categories due to difficulties in excavating their burrows and habitat disruptions. Tree barks were observed up to 3m in height from the ground level. Tree heights were categorized into five height classes as 1: 0-60 cm, 2: 61-120 cm, 3: 121-180 cm, 4: 181-240 cm and 5: 241-300 cm. Tree diameter at breast height (DBH) was measured using a DBH tape. Tree DBH measures were categorized into five classes as 1: 0-120 cm, 2: 121-240 cm, 3: 241-360 cm, 4: 361-480 cm and 5: 481-600 cm. Photographs were taken using a Canon 750D camera with Canon EF 100mm f/2.8L Macro IS USM lens with an external flashlight. Identifications of the species were based on Kovařík et al. (2016).

Statistical analysis
The α-diversity of scorpion species across open and forest habitat was calculated using the Shannon diversity index (H') separately for two habitats (Magurran 1988). Shannon evenness (E) was calculated to analyse the evenness of species across the forest and open habitats (Magurran 1988). Margalef's species richness index (D Mg ) was used to compare species richness across microhabitats (Magurran 1988). Bootstrap sampling using the means of each data set was carried out to assess 95% confidence intervals of Shannon Index (H'), Shannon Evenness (E) and Margalef Diversity Index (D Mg ) using R version 6.3.
The β-diversity, which represents unshared species, was measured by finding similarity or overlap between scorpion species composition across microhabitats, using Sørensen index. We employed chi-squared tests of independence to test the significant difference in the microhabitat preference (height and DBH) of scorpions between open and forest habitat types.

RESULTS
During the survey, five species of scorpions belonging to four genera in two families were recorded (Image 3). Of which, 28% of individuals belonged to family Buthidae, and 72% of individuals belonged to family Scorpionidae (Table 1). Observed four species of scorpions were terrestrial, and only one species, Reddyanus loebli, was arboreal.
Tree height and DBH preference of arboreal R. loebli were varied. The highest occurrence height was recorded as 300 cm in a Manilkara hexandra tree, whereas the lowest occurrence height was 15 cm in a Drypetes sepiaria tree. Importantly, the highest number of individuals was recorded in height class 3, while the lowest number of individuals was recorded in height class 5 (Figure 1a). The average DBH was recorded as 330cm. The highest number of individuals was recorded in DBH class 4, whereas, at least was recorded in DBH class 2 (Figure 1b). However, there was no significant difference among tree height preference and habitats (χ 2 = 2.947, DF = 4, p= 0.5667). Nevertheless, there was a significant difference in DBH preference and habitat type (χ 2 = 18.041, DF = 4, p= 0.0012).

DISCUSSION
The equatorial location of Sri Lanka and the complex topography of the island produce several distinct climatic zones and diversified habitats. The dry zone (60% of the island), intermediate zone (15%), and the wet zone (25%) are the major climatic zones. Though climatic and environmental factors and vegetation vary among these climatic zones, scorpion species are not confined to specific zones (Kovařík et al. 2016). Their distributions overlap, and only a few species are restricted to specific habitats, like Hottentotta tamulus from Jaffna peninsula, Sri Lanka (Ranawana et al. 2013). We recorded five species, of which four are endemic to Sri Lanka. In the present study, Heterometrus swammerdami was the most abundant species, whereas Isometrus thwaitesi was the rarest. Reddyanus loebli and R. besucheti are the only two species sharing both habitat types. Importantly, Charmus laneus was recorded for the first time in Polonnaruwa.
Heterometrus swammerdami was the only burrowing species in this study. They prefer to burrow in termite mounds, though they are not constructing burrows. They displayed sit and wait behaviour expecting possible prey with extended pedipalp and open chela. Most of the time, one adult can be seen in the opening of the termite mound burrow, and sometimes several juveniles can be observed with their mother. Due to their burrowing behaviour it is difficult to observe them closely to determine age and sex. Higher opportunities to access resources might account for their higher abundance. Isometrus thwaitesi is known as an arboreal species. Kovařík et al. (2016) found I. thwaitesi running on branches and trunks of trees, and also sitting on leaves 1-4 m in height. In this study, however, all three individuals were observed on the ground near a wood debris pile among leaf litter, and they were only observed in forest habitat. The presence of a higher stratum in the forest habitat compared to open habitat could be influencing scorpion abundance by providing better foraging areas where moonlight cannot reach easily (Nime et al. 2013).
Reddyanus loebli is a tree-dwelling species. Most dry zone trees have fissured barks as an adaptation for harsh weather conditions, and this gives a suitable microhabitat. They were mostly (93.2%) observed in Manilkara hexandra, Drypetes sepiaria, and Ficus sp. trees among and under the scales, within the cracks in the bark. Most of the observed individuals displayed sit and wait behaviour under the scales of the tree bark, with extended pedipalp and open chela, remaining 6.8% individuals observed in brick walls of ruins. All juvenile individuals were observed in forest habitat. Vegetation cover in the forest provides a safe habitat from predators for these tree-dwelling scorpions. Reddyanus besucheti is a terrestrial species that is also found in both habitats. In the forest habitat, 55.6% of individuals were observed on the leaf litter, whereas 44.4% were observed in open habitat on sand.
Charmus laneus was the second most abundant scorpion species observed only in the open habitat. Lourenço (2002) recorded this species from Mannar District and in Wilpattu National Park (Northwestern part of Sri Lanka) and Kovařík et al. (2016) recorded this species from Puttalam District and Eluwankulama (western part of Sri Lanka). Therefore, this is the first record of C. laneus from the Polonnaruwa District (eastern part of the island), which is about 200km away from Mannar District. Their distribution was confined to the surrounding of exposed bedrock in an open area. Unlike H. swammerdami, they were very active and observed running among small grasses near to exposed bedrock on open land. None of the individuals was observed in the open grassy plains or among leaf litter.
The total Shannon diversity index was calculated as 1.0880 for both open and forest habitat. Since the normal range of the Shannon index is 1.5-3.5 (Magurran 1988), this value for the entire site indicates shallow species diversity compared to other taxa. This low alpha diversity is common among predators like scorpions because they are well known for their restricted movement, cannibalism, predation by nocturnal predators, habitat specificity, food size specificity, extreme climate adaptability, and adaptive radiation (Newlands 1972;Polis 1990;Pande et al. 2004). Together with a longer life span than many invertebrates, these factors may act as constraining factors as far as species diversity is a concern (Pande et al. 2012). Since, the 95% confidence intervals of Shannon index values are not overlapped, the Shannon index value for open habitat is significantly different than the forest habitat ( Figure 3a). This reflects open habitat has higher scorpion diversity compared to the forest habitat, because open habitat contains scattered boulders. Crevices under boulders are a preferred habitat for scorpions to spend the day time.
The number of species reflects the species richness. Species richness is strongly dependent on sampling size and effort (Help et al. 1998). The species abundance is often a more sensitive measure of a diversity parameter than species richness alone (Kempton 1979). To overcome this problem, the Margalef index was used. Since, the 95% confidence intervals for the Margalef J TT  similar species are equally abundant (Lloyd & Ghelardi 1964;Magurran 2004). Evenness value range from 0.0-1.0. When the species are equally abundant, evenness value is greater. When the few species are dominant in the community, evenness is less (Magurran 2004). Since J TT Image 3. Scorpion species found in Polonnaruwa Archaeological Reserve, Sri Lanka: a-Charmus laneus female | b-Isometrus thwaitesi female | c-Reddyanus besucheti female | d-Reddyanus loebli female | e-Heterometrus swammerdami female. © Kumudu B Wijesooriya same habitat or co-occur in the same shelter (Warburg 2000).
Arboreal scorpion R. loebli prefers to occupy around heights of 121-180 cm range. This might be mainly due to foraging opportunities and predator pressure. They are considered efficient predators of Isoptera, Hymenoptera, Diptera, Hemiptera, while civets, mongoose, land monitors, and lizards are the predators of them (personal observation). Thus, R. loebli might prefer to forage in this favourable height range without being consumed by another predator. On the other hand, R. loebli prefers to inhabit around DBH of 361-480 cm range, which is above the average DBH level. The diameter of a tree considered as contemplate of a niche area for an arboreal scorpion. Thus, they favour occupying a much larger niche for obtaining more resources like prey, sites to rest and hide from predators.
In conclusion, the five species reported in the Archaeological site of Polonnaruwa suggest high scorpion the 95% confidence intervals for the Shannon evenness values of two habitats are not overlapped, the Shannon evenness values are significantly different in the forest and open habitats (Figure 3b). The higher Shannon evenness value of forest habitat explains that scorpions found in forest habitat were more equally abundant than the open habitat due to the high dominance of H. swammerdami in open habitats (Figure 2a). Similarly, forest habitat has evenness value below 0.5, which is due to the high dominance of R. loebli in forest habitat ( Figure  2b). Beta diversity of habitats compares the species similarity between the two habitats (Magurran 2004). To compare the similarity between two habitats, which was calculated as 0.5882 in Sørensen index in a way reflecting a more than 50% shared species between two habitats. Similar results were observed in previous studies as intraspecific and inter-specific coexistence in several species of scorpions (Kaltsas et al. 2009;Shehab et al. 2011;Lira et al. 2013). Thus, species might either co-occur in the J TT richness in this area. This highlights the importance of conservation of historic ruins and forest patches of the archaeological site to maintain scorpion fauna. Thus, the results presented here contribute to the knowledge of the diversity of scorpions in these historically significant sites that can serve as a basis for future research on the impact of habitat modification and fragmentation on the population, distribution, and ecology of scorpions.