Use of anthropogenic structures by bats in rubber plantations: A preliminary survey from Sri Lanka

: Sri Lanka supports a high biodiversity and nearly one third of its mammalian fauna comprise of bats. Sri Lanka’s bats are understudied with few data available about their roosting habits, particularly in agricultural ecosystems. Here we report the findings of a preliminary survey, conducted to explore how bats use available anthropogenic structures in rubber plantations. We inspected abandoned buildings within six rubber plantations in Deraniyagala for the presence of bats. Eleven of the 14 abandoned buildings inspected were occupied by bats, viz , lesser false vampire bat ( Megaderma spasma ) and rufous horseshoe bat ( Rhinolophus rouxii ). Our results indicate that bats use anthropogenic structures for day roosting and as maternity roosts, as the presence of young-of-the-year (YOY) were observed in one of the roosts. We recommend implementing economically viable conservation measures such as creating artificial tree hollows and erecting artificial roosts like bat boxes in rubber plantations to provide permanent habitats for bats, as abandoned buildings may be demolished or repurposed in the near future.

Although the different ecosystems in Sri Lanka provide habitat for a wide range of bats (Yapa & Ratnaweera, 2013), these are declining rapidly due to habitat destruction and degradation.Approximately 30% of the land area of Sri Lanka is agricultural land dominated by seasonal (e.g., paddy) or perennial monocultures (e.g., rubber, coconut and tea) (Perera, 1984;Seo et al., 2005).Although these monocultures represent a significant proportion of the island's land area, there is a dearth of studies conducted on the faunal assemblages within them (Kottawa-Arachchi & Gamage, 2015).
Many countries from the Asian (Phommexay et al., 2011;Tanalgo et al., 2021;Wordley et al., 2014) and South-American (Heer et al., 2015) regions have produced studies on how agricultural lands specifically impact bat species.However, in Sri Lanka very few such studies have been conducted.The richness and distribution of bat species in Sri Lankan tea plantations were reported by Kusuminda et al., (2018) but no studies have yet been conducted within the island's rubber plantations, which constitute an area of 138,000 ha within the country (RRISL, 2020) and is 4.9% of the country's total agricultural land.
Rubber plantations offer potential anthropogenic roosts for bats in the form of abandoned worker houses, tool sheds and dilapidated factories.Roosting in anthropogenic structures offer a myriad of benefits to bats.these include lower energetic costs and lower predation risk (Lausen & Barclay, 2006), enhanced access to viable habitat (Mazurska & Ruczyński, 2008) and even range expansions (Voigt et al., 2015).However roosting in buildings could also expose bats to pernicious chemicals that reduce fitness (Bayat et al., 2014).Therefore, it is important to understand the extent of anthropogenic roost use by bats in rubber plantations in Sri Lanka.Our goal was to obtain a preliminary understanding of the use of anthropogenic structures by bats within rubber plantations of the Deraniyagala region (Sabaragamuwa province, Sri Lanka) and we expected to observe common species such as Schneider's leaf-nosed bat (Hipposideros speoris), false vampire bat (Megaderma spasma) and rufous horseshoe bat (Rhinolophus rouxii) within these structures.

Study area and time period
The study was conducted in nine rubber plantations (Sapumalkanda: 6.9310 lat, 80.3281 long; Ruecastle:  1).This area is situated within the wet zone of the country and has a mean elevation of 200 m with rubber plantations dominating the landscape.This study was conducted during a period of six months (August 2021 to January 2022).

Identifying and inspection of potential roosting sites
Potential roosting sites were selected by liaising with the field staff and administration of the plantations.We inspected abandoned worker homes, storerooms and disused and dilapidated buildings that were accessible for the presence of bats during daylight hours (0800h-1700h).
We were not made aware of any buildings in use that had bats roosting in them.Upon entering a building, a handheld torch was used to detect the bats.Counts were made by two independent observers and the average of the counts were used sensu Kusuminda et al., (2018).No bats were captured during the course of this study and we identified bats based on detailed photographs obtained using a digital camera checked against published field guides (Yapa, 2017) and a field identification key (Srinivasulu et al., 2010).We determined if individuals were adults or young-of-the-year (YOY) by morphological characters such as significantly smaller body size and differences in pelage (Phillips, 1980).
The location of each inspected building was determined using a handheld GPS unit (Garmin GPSMAP 64s).

RESULTS
Fourteen buildings were inspected for the presence of bats.
Out of these, 11 were occupied by at least one individual (Figure 2).M. spasma (n = 75 individuals, median = 4, SD = 5) (Figure 3) and R. rouxii (n = 2 individuals) (Figure 4) were the only species observed.Only one building was occupied by both species.The largest number of bats observed at a given location was 16 individuals.This was observed in two locations and in each case the colony comprised of M. spasma.At one of the sites three YOY were recorded.

DISCUSSION
Our preliminary survey indicates that very few species use anthropogenic structures within rubber plantations, despite their availability.However, our results indicate that 80% of the studied sites supported bats.Both species observed during our study (M.spasma and R. rouxii), have previously been reported to use anthropogenic structures for roosting both within Sri Lanka (Kusuminda et al., 2018;Yapa & Ratnaweera, 2013;Yapa, 2017) and elsewhere (Kingston et al., 2006;Molur et al., 2002).The number of individuals in the roosts we observed are comparable with these published records as M. spasma roost in groups ranging from 3-30 (Yapa, 2017) while R. rouxii may roost solitarily or form groups of hundreds of individuals (Phillips, 1980;Yapa & Ratnaweera, 2013).
The paucity of other species was unexpected as S. saccolaimus (Edirisinghe et al., 2013), M. lyra, Taphozous longimanus, T. melanopogon and H. speoris (Yapa & Ratnaweera, 2013;Yapa, 2017), have been observed roosting in buildings in the wet zone of the country.The relative rarity of S. saccolaimus, M. lyra and the two Taphozous species likely contributed to their absence.
Although we carefully examined each building for bats, the presence of diminutive species such as the Indian pipstrelle (Pipistrellus coromandra), known to inhabit buildings (Molur et al., 2002;Raghuram et al., 2014) could have been undetected, leading to a false negative result.Ruvinda et al.We suggest future studies using acoustic bat detectors and active sampling (using mist nets and harp traps) conducted throughout the year to establish the dynamics of roost use of bats in these plantations.
One species, H. speoris, was noticeably absent during our study.This species is considered the most abundant insectivorous bat in Sri Lanka (Yapa, 2017) and its absence is likely not due to rarity.This could possibly be due to the migration of this species in search of suitable maternity roosts during the breeding season (Yapa & Ratnasooriya, 2006).The diversity of tropical bat species, is negatively influenced by habitat destruction (Kingston, 2013), and this is particularly evident in rubber plantations (Danielsen & Heegaard, 1995;Phommexay et al., 2011).Lower arthropod prey abundance in rubber plantations (Zheng et al., 2015) is considered one of the drivers behind this lack of diversity (Harvey & González Villalobos, 2007;Phommexay et al., 2011).It may be that the insect population and diversity within the plantations we visited were insufficient to sustain a population of H. speoris despite their abundance and use of buildings.Conversely, H. speoris might be avoiding the roosts we examined for reasons hitherto undetermined but still feed within these plantations.This possibility should be explored through future research incorporating acoustic monitoring and active sampling.
Buildings offer several benefits to bats (Voigt et al., 2015) including reduced predation risk (Lausen & Barclay, 2006), and microclimatic stability (Whitaker & Gummer, 1992).Compared to natural roosts, abandoned buildings in the rubber plantation would provide less opportunity for predators such as snakes (Esbérard & Vrcibradic, 2007) and viverrids (Kusuminda et al., 2013) to prey on bats.Bats may also be using the buildings due to their favourable microclimate (Vivier & van der Merwe, 2007).
A warm, stable microclimate is particularly important for reproductive females and YOY (Chruszcz & Barclay, 2002;Sedgeley, 2001).Our observation of three YOY M. spasma suggests they may be using these building as maternity roosts.Conversely, the use of buildings as roosting sites could be a result of unavailability of natural sites within rubber monocultures and on a landscape level.Rubber trees are usually devoid of cavities that M. spasma are known to roost in (Balete, 2010;Yapa & Ratnaweera, 2013).During our study we only observed one natural roost; a cave inhabited by 15 R. rouxii and four M. spasma.
Roosting in buildings is not without cost.For example, bats in buildings have a greater chance of being exposed to chemical pollutants (Bayat et al., 2014), which can cause physiological damage (Clark & Shore, 2001).Bats inhabiting agricultural lands have been shown to accumulate pesticides and insecticides in their systems (Stahlschmidt & Brühl, 2012), which further aggravate physiological damage.Some of the buildings we inspected were used as storage depots for chemicals including pesticides and herbicides, and the bats roosting in them could be exposed to them.In addition, these roosts in plantations could be ephemeral and spontaneously demolished.If natural roost sites within a rubber plantation are already scarce then the destruction of these buildings may lead to further population declines.We recommend that more permanent artificial roosts (i.e., bat-boxes and artificial tree hollows) (Brittingham & Williams, 2000;Whitaker et al., 2006) be erected in rubber plantations as they are an effective conservation tool in many parts of the world (Flaquer et al., 2006;Rueegger, 2016) and would ensure the continued survival of bats within these rubber plantations.