Vertebrate road kills on State Highway 26 in Khandwa Forest Division, central India

: The road network is spreading rapidly in human dominated as well as forested areas in India. Roads passing through forested areas are known to negatively impact wildlife populations by contributing to wildlife-road kills besides causing habitat fragmentation, soil erosion, and disruption of gene flow. This paper documents encounters of wildlife road kills on the State Highway 26 (SH 26) passing through the forested area of Khandwa Territorial Forests in central India. Between October 2015 and June 2017 and from November 2021 to June 2022 we collected opportunistic data on wildlife mortalities on SH 26. We recorded 61 wildlife mortalities on a stretch of 34 km of SH 26. Of the road kills encountered, 71% were birds, 21% were mammals, and 8% were herpetofauna. Highest among bird kills were members of nightjars (Caprimulgidae) (26%) and owls (Strigidae) (23%). Among owls, we recorded two road kills of the endangered Forest Owlet Athene blewitti and individuals of five other owl species. Among other taxa, road-hits of Jungle Cat, Indian Palm Civet, Golden Jackal, Central Indian Langur, and a few snakes were recorded. Most bird kills on the road were recorded in February while the highest number of mammal and herpetofauna kills were recorded in the month of November. We report that the land use around the road and species trait contribute to vulnerability of a species to road-hit. Based on the frequency of road kills, we have identified seven hotspots of road kills on SH 26. Since road construction is one of the major mandates of the Government in India, it is suggested to carry out systematic environment impact assessment on existing and proposed roads to understand the patterns of spatio-temporal incidents of road kills. Planting of tall trees and installation of speed bumps and signboards to alert drivers should be planned at the initial stage of road construction. If appropriate mitigation measures to prevent wildlife mortalities are incorporated while the road is being constructed, it may be possible to bridge the gap between development and conservation.

While a small animal may get overridden, a large one will collide with the vehicle causing an accident and/or death. For all such instances, we have used the term wildlife-vehicle collision (WVC). Incidents of wildlife mortalities due to collision with speeding vehicles (WVC) in forested landscapes has far-reaching implications compared to poaching, hunting or natural death as WVC can also affect healthy individuals, leading to loss of genetic flow, subsequently resulting in population decline or isolation (Sutherland et al. 2010;Jackson & Fahrig 2011). Roadkill rates are attributed to the type of road, speed of the vehicles, and traffic volume (Forman & Alexander 1998;Underhill & Angold 1999;Clevenger et al. 2003;Saxena et al. 2020). Land use by the roadside has a major impact on accelerated rates of road kills. Presence of agricultural fields, forest patches, ditches, wetlands, canal crossing, shrubby vegetation and uneven height of road have been identified as hotspots of road kills (Main & Allen 2002;Mackinnon et al. 2005;Özcan & Ozkazanc 2017). Species traits such as group living animals, large home ranges, high dispersal rate,s and poikilothermic species are most likely targets of road kills (Saxena et al. 2020). Diet and morphometry of a species contributes to their vulnerability to road kills. Species feeding on rodents, insects, and carcass are vulnerable to road causalities since their life traits make them an easy victims of road kills (Adams & Geis 1983;Main & Allen 2002;Silva et al. 2019;Underhill & Angold 1999;Medrano-Vizcaíno et al. 2022).
The studies on road kills from India and outside India have commonly reported mortalities of mammals (Reed et al. 1979;Lavsund & Sandegren 1991;Behera & Borah 2010;Gubbi et al. 2012;Saxena et al. 2020;Lyamuya et al. 2022), birds (Channing 1958Dunthorn & Errington 1964;Dhindsa et al. 1988;Sundar 2004;daRosa & Bager 2012;Bishop & Brogan 2013;Siva & Neelnarayanan 2020;Sacramento at al. 2022), and herpetofauna (Vos & Chardron 1998;Aresco 2005;Langen et al. 2007;Glista et al. 2008;Baskaran & Boominathan 2010;Bhupathy et al. 2011;Quintero-Ángel et al. 2012;Samson et al. 2016;Jegannathan et al. 2018;Pallares & Joya 2018;Hastings et al. 2019). In India, the majority of studies on WVC have been carried out on national highways (NH) passing through protected areas of Western Ghats (Kumara et al. 2002;Seshadri et al. 2009;Baskaran & Boominathan 2010;Santhoshkumar et al. 2017), Madhya Pradesh (Pragatheesh 2011;Pragatheesh & Rajvanshi 2013;Saxena et al. 2020 (Pragatheesh 2011;Pragatheesh & Rajvanshi 2013;Saxena et al. 2020) but there is no published information on road kills outside the protected areas in the state. In this paper, we describe wildlife mortalities encountered on the State Highway 26 (SH 26) passing through the forested area of Khandwa District. The objectives of our study were (i) to assess which taxa are represented in the road kills and (ii) to identify the locations that record most wildlife mortalities. In the light of a recent notification on SH 26 proposed to be widened from a 2-lane state highway to a 4-lane national highway, the findings and recommendations of this paper would be of considerable importance since we describe the hotspots of the road kills and suggest precautionary measures and mechanisms to reduce wildlife kills on SH 26.

Study Area
We recorded the road kills on the SH 26 that passes through Khandwa Forest Division in Khandwa District (21.8259 0 N, 76.3678 0 E) of Madhya Pradesh (Figure 1). Both sides of the road were bordered by dry deciduous teak Tectona grandis forests and crop fields. Teak is the dominant species in the forests with J TT associated species such as Anogeissus latifolia, Lannea coromendelica, Terminalia alata, Butea monosperma, Diospyros melanoxylon, and Garuga pinnata. Bamboo was found mainly in hilly region of the road. Very few houses were close to the highway as villages were settled further away from the road. The topography in the area is hilly with gently sloping terrain interspersed with plain tracts of land. The elevation ranges 300-700 m. Among large mammals, the presence of Bengal Tiger Panthera tigris, Leopard Panthera pardus fusca, Jungle Cat Felis chaus, Chinkara Gazella benettii, Sambar Rusa unicolor, Nilgai Boselaphus tragocamelus, Four-horned Antelope Tetracerus quadricornis, and Wild Boar Sus scrofa have been recorded from the area (Shukla 2013

Methods
Since 2012, we have been carrying out ecological research on the Forest Owlet and other owls in Khandwa District of Madhya Pradesh. Every day, between 0600 h & 1100 h and 1600 h & 1900 h, we used to drive at the speed of 20 to 40 km/h by a jeep or a motorcycle on the road to monitor locations of owls for a stretch of 34 km to June 2022, we maintained a record of the road kills observed while driving on the road. The data were collected opportunistically, i.e., as, and when we came across a road kill. When we spotted a dead animal on the road, we recorded the following information: date, time of observation, species killed, and GPS location of the kill. To estimate approximate time of the kill, we categorised the road kills in two broad categories, as fresh (killed within last 24 hours) and old (killed over last 24 hours), based on the status of the carcass, skin texture, and blood condition (after Baskaran & Boominathan 2010). To assess the topography at the accident spot, we recorded the gradient of the road as plain, gently sloping, and hilly. For understanding the land use near the accident spot, we classified the habitat as teak forest, teak mix forest, agriculture, human habitation, and a combination of the given categories. After obtaining the data indicated above, we removed the carcass from the road to prevent double counting (Gomes et al. 2009).

Study segments
In order to have better understanding and interpretation of the WVC patterns we divided the road into five segments according to the location of nearby villages and recorded the landscape features on either side of the segment (Table 1). Study areas 2 and 4 differ in their extent and pattern of cover over agriculture land and teak forests. We used non-parametric Mann-Whitney U test to compare the frequency of road kills among different taxa and the number of mortalities between different road segments. All species encountered in road kills were classified in three broad categories of feeding guilds (Table 1). We used the software PAST 4.0 for the analysis.
Among the birds, family Caprimulgidae (Nightjars) represented the highest (26%) number of kills followed by Strigidae (Owls) (23%), and Cuculidae (Cuckoos) (14%). The most frequently encountered bird kill was that of the nightjars followed by Greater Coucal Centropus sinensis, Indian Roller Coracias benghalensis, and the Spotted Owlet Athene brama. One species of dove could not be identified because of its decomposed body. Among the mammals, members of the family Sciuridae (Squirrels) were most killed species (31%) followed by Viverridae (Civets) (15%), and Muridae (rats & mice; 15%) ( Figure  4; Image 2-9). Of the herpetofauna road kills, we found three species of snakes belonging to two families and one species of toad on the road during the survey. We found an almost equal number of mortalities of diurnal and nocturnal species (Table 2).

J TT
Type of habitat Total number of road kills were significantly higher on segment 2 compared to segment 4 (Mann-Whitney U = 484, P = 0.049) and segment 5 (U = 463, P = 0.039). Segment 1 recorded more road kills than segment 4 (U = 453, P = 0.03, Figure 3). Bird, mammal and herpetofauna road kills were found on segment 1, 2, 3, & 4 while on segment 5 we found road kills of two Three-striped Palm Squirrels Funambulus palmarum. However, we did not find significant differences in number of road kills of

Monthly patterns of road kills
Most bird road kills were recorded in February, while the highest number of mammal and herpetofauna kills were recorded in the month of November ( Figure 5). The number of bird road kills per month was significantly higher than that of mammal kills (U = 16.5, P = 0.033) and herpetofauna kills (U = 8, P < 0.00) per month.
A comparison of percentage of foraging guilds encountered as road kills revealed that the members of insectivore and carnivore-insectivore guilds were getting killed throughout the year while seed/fruit/grain feeders were killed mainly between January to May ( Figure 6). Carnivore-insectivore road kills were seen mainly in July and December. We do not have road kill data of August and September as we did not monitor the road during those two months.

Herpetofauna Road Kills
In India, most studies on road kills have commonly reported high mortality of amphibians, reptiles, and mammals (Kumara 2000;Das et al. 2007;Baskaran & Boominathan 2010;Gubbi et al. 2012;Jeganathan et al. 2018;Kumawat & Purohit 2020;Saxena et al. 2020;Sur et al. 2022). Among herpetofauna, snakes are commonly reported taxa of WVC. Ectothermic animals like snakes utilise roadways to regulate their body temperature during the winter season by resting on them, which makes them easy victims of road fatalities (Rosen & Lowe 1994). Snakes visit agricultural fields hunting for rodents (Pragatheesh & Rajvanshi 2013) and are at the greatest risk of death because their movement is slower on a smooth road than on other surfaces (Row et al. 2007). The higher road mortality of amphibians and reptiles could be attributed to their slow mobility, not reacting quickly to vehicles and the fact that drivers are less likely to notice these animals because of personal disregard for the species (Pragatheesh & Rajvanshi 2013). In our study, we found only four road kills of herpetofauna, including three species of snakes and one species of Common Indian Toad. We report road kills of Common Krait, Rat Snake, and Bronze-back Tree Snake, which appear to be victims of road kills commonly (Das et al. 2007;Baskaran & Boominathan 2010;Pragatheesh & Rajvanshi 2013;Jeganathan et al. 2018). Bronze-back Tree Snake is an arboreal species and its presence on the road is indicative of loss of canopy connectivity induced by road construction (Pragatheesh & Rajvanshi 2013). The above studies have reported the maximum mortalities of amphibians and reptiles during the monsoon months. The lower number of herpetofauna kills in our study may be explained by the fact that we

Mammal road kills
Among mammals, we found carcasses of Jungle Cat, Jackal, Domestic Cat, Rufous-tailed Hare, Palm Civet, and Palm Squirrels which have also been reported as road kills by other studies (Kumara 2000;Sundar 2004;Behera & Borah 2010;Baskaran & Boominathan 2010;Borah 2010;Jeganathan et al. 2018;Saxena et al. 2020;Sur et al. 2022). Nocturnal mammals (carnivores) often forage near the open areas by the road side hunting for small mammals and insects, and get temporarily blinded by powerful headlights thus making them victims of speeding vehicles (Orlowski & Nowak 2006). However, there was also a kill of a Central Indian Langur, which is a diurnal species. Langurs and Macaques often sit on the road to pick fallen fruits and grab food that is thrown out to them from passing vehicles. Langurs are group living animals and many of them run across the roads, thus increasing their chances of collisions with vehicles (Baskaran & Boominathan 2010).

Bird road kills
In India, a handful of studies have discussed the frequency of bird road kills (Dhindsa et al. 1988;Sundar 2004;Siva & Neelanarayanan 2020). Outside India, several studies have highlighted the frequency of bird road kills (Hodson 1962;Erritzoe et al. 2003;Gomes et al. 2009;Cook & Blumstein 2013;Husby 2016). Every year several millions of birds are estimated to die due to road hits in USA, UK, and Europe (Erickson et al. 2005;Bishop & Brogan 2013;Bíl et al. 2020), indicating the vulnerability of birds to speeding vehicles.
Our study also reports that avifauna were the major (71%) victims of the WVC. We recorded 33.33% mortalities of nocturnal birds, which included 11 individuals of nightjars and 11 individuals of owls belonging to seven owl species (Table 2). Nocturnal birds of prey move along the road side to hunt for insects collected near street lamps. Pools of water ditches, and wells by the road side make ideal feeding grounds for predatory birds feeding on amphibians and reptiles, and therefore get into collision with vehicles (Hernandez 1988). Studies reporting road kills of nightjars have surmised that nightjars often sit on the roads to feed on the insects near light posts. Also, during the breeding season nightjars sit by the road side to incubate their eggs because roads become warmer at night. The sudden flash of headlights temporarily blinds the nightjar, and they are unable to fly off (Erritoze et al. 2003).
Among owls, we recorded two road kills of the Forest Owlet. In 2015, one breeding male Forest Owlet was found dead on the road. After the incident, we could not locate the juvenile and the female in the area. It is likely that the female may have abandoned the area because the provisioning male had died. In 2022, one more breeding male Forest Owlet was found dead on the road. Although the Forest Owlet is a diurnal owl, they hunt during low-light hours during dusk and dawn (Prachi Mehta pers. obs. 2013). Losing an endangered species to road kill is a matter of grave concern as it has a direct implication on its in-situ conservation. We also report road kills of large owl species such as the Barn Owl, Mottled Wood Owl, Indian Eagle Owl, and smaller owl species such as the Jungle Owlet, Indian Scops Owl, and the Spotted Owlet. Several studies have documented Strigiformes to be common victim of WVC (Hernandez 1988;Gomes et al. 2009;Baskaran & Boominathan 2011;Le Gouar et al. 2011;Guinard et al. 2012;Siva & Neelanarayanan 2020;Sur et al. 2022). Authors have attributed greater road kills of Strigiformes to the hunting techniques of owls. Owls, unlike diurnal birds of prey, do not use hot wind currents to soar in the air but use perch and pounce technique (Mikkola 1983). Many owl species use trees, posts, fences, and cables as perches to scan the ground for prey (Mehta et al. 2018). Barn Owl and Eagle Owl often sit on the light pole near agricultural fields to prey on rodents. Smaller owls such as the Spotted Owlet often perch on low bushes, near a light source or on the edge of the road to catch insects (Mehta et al. 2018) and may be getting overrun by a speeding vehicle. Also, illuminated roads attract invertebrates, which in turn, attract the owls (Hernandez 1988). Abundance of prey attract owls to roads and flying from one bush to another can get them killed (Nero & Copland 1981).
A combination of hunting technique and availability of open areas make owls an easy victim of road kills. Large numbers of mortalities of Strigiformes due to WVC is an issue of serious conservation concern as it can affect their population structure (Forman & Alexander 1998;Le Gouar et al. 2011;this study).
Among the diurnal bird kills, we observed the carcass of doves, pipits, and coucal, who often feed by the roadside on the grains (Dhindsa et al. 1988). Most birdvehicle collisions occur because of the bird's quick flight ability and their tendency to fly across the road (Rilley et al. 2014). Also, roads that have agricultural land by the side is preferred by birds for feeding on spilled grains from the agricultural land or feeding on rodents from the crop field (Dhindsa et al. 1988;Erritoze et al. 2003).

Monthly patterns of road kills
From January to May, crops are harvested and bagged. During this process, there is a spill over of grains on the road and an influx of insects and rodents in the crop fields, which in turn attracts the movement of insectivores, carnivore-insectivores and seed/fruit/grain feeders near the road. Further monitoring and data will J TT be required to understand the influence of diet on the taxa encountered in road kills.

Habitat surrounding the road
On SH 26, road segments 1, 2, & 3 passed through plain terrain and the road side habitat was mainly teak forests interspersed with agriculture. Segment 4 passed through hilly terrain and had teak mixed forests all along the road. Segment 5 passed through plain terrain with agriculture fields and open forests by the road side. We found maximum wildlife mortalities of doves, munias, rollers, babblers, nightjars, coucals, owls and mammals such as Jungle Cats, civets, Jackal on segments 1, 2 and 3. Most of the above species feature in other studies as road kills (Dhindsa et al. 1988;Sundar 2004). Agriculture fields have plentiful spilled over grains for granivores, insects for insectivores, and rodents for birds of prey. Additionally, interspersion of agriculture and forests forms 'edge' habitat which offers foraging opportunities for a variety of bird (Dhindsa et al. 1988) and mammal species (Baskaran & Boominathan 2010). The speed of the vehicle also contributes to the frequency of road kills (Erritzoe et al. 2003;Saxena et al. 2020). Segments 1, 2, & 3 had plain terrain therefore vehicles move faster on this stretch of the road. Segment 4 passed through hilly portion, so the vehicles must slow down while driving through this segment. This could be the reason for relatively lower numbers of road kills on segment 4. Studies have reported lower frequencies of animal collision on roads passing through farm lands but much higher frequencies of causalities on roads passing through a mix of crop fields and forests (see review in Erritzoe et al. 2003). This may be a possible explanation of lowest number of road kills on Segment 5, which had mainly crop fields and patches of open forests.

CONCLUSION AND RECOMMENDATIONS
The road kills encountered in the present study shows that diverse wildlife exist outside the protected areas, and they move between different habitats such as agriculture fields, forests, and forest edge for foraging, and other ecological requirements. Our study reports that the vehicular traffic on SH 26 is taking a toll on the wildlife in the area. Numerous kills of Nightjars and different species of owls, including two individuals of the endangered Forest Owlet, is alarming as it indicates that these species use the road frequently and are at high risk of getting into collisions with vehicles. Through this paper we have provided information on type of species killed, the locations of road kills and monthly patterns of WVC. Our findings could be used as a baseline information to prevent further wildlife mortalities on SH 26.
Based on the frequency of road kills, we have identified seven hot spots for the accidents on SH 26 (Figure 7). Installation of speed bumps and rumblers are effective in controlling speeding vehicles. We suggest installation of speed rumblers at each hotspot.
For preventing mortalities of nocturnal animals, we suggest removing shrubs from the roadside and planting tall native trees on either side of the hotspots and all along the roadside on SH 26. On hotspots 1, 2, & 3, the road level is uneven, which allows accumulation of water in ditches. This can attract owls and other mammalian predators for hunting of small prey. The height of the road should be made even to prevent water accumulation.
Generalist species like Domestic Cat, Jungle Cat, Jackals, Langurs, babblers, robins, rollers, and mynas have got exposure to a variety of habitats and therefore are not wary of crossing the roads. Such species often come on to the road to scavenge and become a victim of road hits (Medrano-Vizcaíno et al. 2022). The habitat around hotspots 1, 2, 3, 6, & 7, is mainly agricultural fields interspersed with patches of forests. The owners of the crop fields should be requested to keep the road side free of grains, and fruits that may be attracting animals for foraging. The highways should be kept free of carcasses too as it attracts bird and mammalian scavengers.
Animal crossing signages and reflectors are helpful in alerting the driver to watch out for wildlife crossing the roads. We have recently put up a few such signboards on SH 26 urging the drivers to maintain slow driving speed ( Figure 8). The signboards should be shifted to new locations every few months, because if the signboards are at the same place, it will not invite the attention of regular drivers as they get used to seeing it (Sullivan et al. 2004).
The proposed widening of SH 26 for making into a four-lane road will lead to an increase in traffic intensity, vehicle speed and is expected to increase the wildlife death toll. This calls for attention of such studies to be carried out on a wider scale across the country to understand the impact of roads passing through forested areas.
Considering the rapid rate at which roads are being constructed in the country, it is imperative and urgent to carry out systematic and rigorous impact assessment on existing and proposed roads. Further, it is important to institutionalise a mechanism through which effective www.threatenedtaxa.org The Journal of Threatened Taxa (JoTT) is dedicated to building evidence for conservation globally by publishing peer-reviewed articles online every month at a reasonably rapid rate at www.threatenedtaxa.org. All articles published in JoTT are registered under Creative Commons Attribution 4.0 International License unless otherwise mentioned. JoTT allows allows unrestricted use, reproduction, and distribution of articles in any medium by providing adequate credit to the author(s) and the source of publication.