Relative Abundance of Mammalian Fauna of Chitral Gol National Park, Pakistan

1Wildlife and Ecosystem Research Lab, Department of Zoology, University of Chitral, KPK, Pakistan. 2Carnivore Conservation Lab, Department of Zoology, Quaid-I-Azam University, Islamabad 3Snow Leopard Foundation, Islamabad, Pakistan 4Directorate of National Parks, Wildlife Department Khyber Pakhtunkhwa, Pakistan 5Department of Biological and Environmental Sciences, Qatar University, Doha Qatar Article Information Received 25 July 2021 Revised 03 November 2021 Accepted 14 November, 2021 Available online 01 March 2022 (early access)


INTRODUCTION
M ammals are ecologically important and they directly or indirectly affect communities and ecosystems (Ripple and Beschta, 2004;Ray et al., 2005;Roemer et al., 2009). Both carnivorous and herbivorous mammals play critical role in maintaining the integrity of the ecosystem (Marcot and Heyden, 2001;Bakker et al., 2016). Carnivorous mammals are ecologically significant, as only a limited number of species cause strong predation, that both directly and indirectly affects ecosystem structure and function (Ray et al., 2005;Roemer et al., 2009). Many mammals are ecosystem engineers that significantly create or destroy habitats that changes the ecosystem structure and function (Lacher et al., 2019). These are the species that can shape both the biotic and abiotic component of the ecosystem in which they live (Jones et al., 1994, Jones andSafi, 2011;Sinclair et al., 2015). Relative to their abundance, mammals affect plant structure and function to the greater extent as compared to any other animal group (Hulbert, 1997;Paine, 2000;Sinclair et al., 2015). Mammals, like markhor Capra falconeri also generate revenue for the local community through trophy hunting programs Nawaz et al., 2016). Pakistan has varied range of ecosystems and unique ecological regions that support rich diversity of animals including mammals, birds, reptiles, amphibians and fishes (Altaf et al., 2014). Variations in climatic conditions and topography are the significant predictors of mammalian species diversity where different habitats support different species of mammals (Vaughan et al., 2000;Atnafu and Yihune, 2018). Faunal diversity also corresponds to the altitudinal variations which exist throughout the landscape of Pakistan (Altaf et al., 2014), particularly the mountain O n l i n e

F i r s t A r t i c l e
areas of the country show abrupt change in elevation and have many distinct habitat types which harbor large number of wildlife species (IUCN, 1996). Estimating species abundance is necessary for assessing extinction risk of endangered species, monitoring invasive species and particularly in managing population of threatened species (Yin and He, 2014). Knowledge of species abundance is necessary for decision making in biological management and conservation and for understanding the dynamics of population (Yin and He, 2014). Estimating abundance helps in setting hunting quotas, gauging prey availability for carnivores and managing wildlife areas for tourism (Palmer et al., 2018). Conservation efforts are assessed by using abundance estimates, it also provides insight into how a community functions (Danell et al., 2006;Verberk, 2011;Cox et al., 2017).
Investigation on abundance of mammals provides information for appropriate conservation actions (Galetti et al., 2009;Rabira et al., 2015;Atnafu and Yihune, 2018). Relative abundance is usually expressed as an index value calculated from the frequency of signs or observations per standardized unit of sampling effort (Wilson and Delahay, 2001). Relative abundance is used because it requires fewer resources and less sampling efforts as compared to population estimation (Lettink, 2012). The data obtained from camera traps can be used to estimate relative abundance of species (Obrien et al., 2003;Lahang, 2005;Obrien., 2011). Measure of relative abundance is the key element of biodiversity required for the estimation of species diversity (Hubbell, 2001;Yoccoz et al., 2001). The data of species relative abundance can be used by ecologists to infer relationships among different species in certain communities (Odat et al., 2009). It is also used to estimate ecological disturbance and perform population viability analysis (Odat et al., 2009).
Chitral Gol National Park (CGNP) harbors a large number of mammals; however, information on population size, estimated densities or any other ecological parameter is not available for those species except annual census counts for markhor. Since, absolute abundance, density and population size estimation are not possible due to financial and time constraints. The current study was conducted with the aim to determine the relative abundance of mammals in CGNP because it can be used as an alternative way to explore diversity of species in the National Park. It is of great importance for the species occurring there and utilizing the NP resources. Estimating relative abundance of mammals in CGNP could provide useful information needed for management of the National Park.

Study area
The study was conducted in Chitral Gol National Park, situated in the District Chitral (Lower), Pakistan. It is located on 35°51ʹ and 35°57ʹ N latitude and 71°43ʹ and 71°47ʹ E longitude while elevation ranges from 1500 to 4950 meters, having an estimated area of 78.6 km 2 . It is located in the Hindu Kush Mountain range lying in the west of Chitral town. The fauna of district Chitral shows affinity towards fauna of Palearctic region, and a little affinity with Oriental region in the south (Din and Nawaz, 2010). CGNP was established in 1984 meeting the criterion of the IUCN management category 2 of protected areas to protect populations and habitat of Snow leopard (Panthera unica) and Kashmir markhor (Capra falconeri cashmiriensis) (Arshad et al., 2012). The area of the NP is relatively steep with sharp ridges and vertical slopes with incline slope ranging from 45° to 120° (Hess, 2002;Arshad et al., 2012). CGNP has mean annual temperature of 16.8°C (Ali, 2008;Shabbir et al., 2013). The climate of Chitral is dry temperate and the area of the national park is out of the range of monsoon and it receives 462 mm of mean annual precipitation, mainly in spring and winter (Hess, 2002;Khan et al., 2010Khan et al., , 2011. During winter the mountains are covered with snow, and summer is long and dry ( Fig. 1).

Survey design
The study was carried out from June 17 to August 3, 2019. During the study, 30 camera traps were deployed within CGNP. The minimum Aerial distance between two (2) camera traps was kept one km. In every suitable location, a single motion triggered camera (Reconyx™)

O n l i n e F i r s t A r t i c l e
with infrared flash was installed. All the cameras were set to a steel pole at a height of 50 cm to capture both large and small sized wild animals. The steel pole was driven into the ground. The cameras were placed facing away from the sun, dense vegetation or any large object that could cause false trigger event. Camera trap units were programmed to take 3 images of size 3.1 MP at the interval of one second. Each image had time and date of capture. Coordinates and elevation of each camera were documented through geographic positioning system (GPS MAP 62s Garmin). The memory cards were of 64 GB and the batteries composed of lithium ions. The approach for sampling was passive as we did not use attractants (i.e. scent) to attract mammals towards the camera trap location. After 47 days of survey period, the cameras were taken off. A zoologist familiar with the region's mammalian fauna independently identified the species photographed by cameras. Photographs that could not be identified were not included in the analysis.

Data analysis Independent detection event
Detection for a species is considered an independent event, when the time interval between consecutive photographs of the same species is more than one-hour interval.

Trap night
It is a 24 h period when camera trap is set to be active. It is calculated by multiplying the total number of camera traps used with total number of nights of active trapping.

Operational day (for camera)
It is the number of days for which the camera traps remain in active condition.

Photos capture Rate
It is the total number of photos captured for one species divided by the total number of photos obtained during the survey per 100 trap nights.

Relative abundance estimation
Relative abundance was calculated using the following formula which is frequently used to calculate relative abundance index from the data generated by using camera traps (Din et al., 2013;Liu et al., 2013).
Where, Ai refers to the overall number of independent detections of species I during the survey period. N represents the overall number of independent events of all species during the survey period.

Camera trapping survey
Thirty cameras were installed for a period of 47 days. Most of the cameras remained operational throughout the study period while three cameras stopped functioning after some days of installation. The survey resulted in 1052 functional trap nights and 478 non-functional trap nights.  (Fig. 2). Prominent avian species captured through camera traps include monal pheasant Lophophorus impejanus, chukor partridge Alectoris chukar and snow cock Tertaogallus himalayensis while some other species were also captured which couldn't be clearly identified (Table  I).

Relative abundance of mammalian fauna
Markhor had the highest relative abundance of 49.631 with 19 spatial captures while golden jackal and golden marmot had the lowest spatial capture accounting for respective relative abundance index of 0.491 and 3.439 respectively. Relative abundance index of rest of the species was as follows: Cape hare 23.832, red fox 6.879, Himalayan lynx 4.668, feral dogs 4.422, stone marten 2.211, Kashmir flying squirrel 1.474, while 0.737 both for common leopard and grey wolf (Table II).

Taxonomic groups of recorded mammals
The results show that different animals belonging to different functional guilds and different taxonomic groups are dwelling in CGNP. Order carnivora was represented by eight species viz., leopard cat, Himalayan lynx, common leopard, grey wolf, golden jackal, red fox and stone marten. Order Artiodactyla is represented by one species, markhor which is near threatened in the IUCN red list and endangered in the IUCN red list of Pakistan. Lagomorpha is represented with only one species that is cape hare. Order Rodentia was represented by three species viz., Himalayan wood mouse, golden marmot and Kashmir flying squirrel (Table III).  Least concern Data deficient Himalayan lynx (Lynx lynx isabellinus) Near threatened Least concern Common leopard (Panthera pardus) Vulnerable Critically endangered Canidae Grey wolf (Canis lupus) Least concern Endangered Golden jackal (Canis aureus) Least concern Near threatened Red fox (Vulpes vulpes) Least concern Near threatened Feral dog Stone marten (Martes foina) Least concern Data deficient Artiodactyla Bovidae Markhor (Capra falconeri) Near threatened Endangered

DISCUSSION
CGNP has been a protected area since 1984 and gained much fame for successful conservation and increasing population of the markhor which has now been raised to around 3000 inside the park (Kakakhel, 2020). Markhor also turned out to be the most abundant species of the park with relative abundance index of 49.631 in the current study.
The relative abundance of cape hare was 23.832 in the study area. Cape hare was the only species photo captured (detected) from the Leporidae family. This species is of least concern status in the IUCN red list of species while in the national red list of Pakistan it is vulnerable (Sheikh and Moular, 2003). In Pakistan the genus Lepus is represented by 3 species Lepus capensis, Lepus nigricollis, Lepus arabicus (Ghazi et al., 2005). Its vertical distribution ranges from 600 meters to 5200 meters in winter and summer season respectively (Roberts, 1997). Most of the animals were captured at elevation varying from 1750 to 3500 meters.
Grey wolf, red fox and golden jackal are abundant species in the study area (Din et al., 2013). In the current study red fox and golden jackal had respective relative abundance index of 6.879 and 3.439. Abundance of red fox and golden jackal may be due to the reason that the National Park is comparatively safe and they are facing less human induced threats (Din et al., 2013). Their generalist food habits and scavenging nature allows them to thrive in the NP where multiple large predators exist along with many prey species (Flower, 1932;Roberts, 1997;Awan et al. 2004;Akrim et al., 2019). Roberts (1997) declared golden jackal to be uncommon in the mountains of Chitral but other latest studies indicate its presence in different parts of Chitral including the study area (Din et al., 2013). The species is highly adaptable, consumes variety of foods and rapidly expands to new habitats (IUCN, 2004;Laszlo et al., 2007;Shabbir et al., 2013). It is important to note that photo capture rate of wolf was lower in the current study as compared to the study conducted by Din et al. (2013) despite the extended duration of study and more cameras. The reason may be season of the study as wolves are known to change habitat during summer and winter seasons (Milacovik et al., 2011;Uboni et al., 2015). Although it is ranked with least concern status in the IUCN Red List, its global population has declined by 33% while it is listed as endangered in Pakistan (Sheikh and Moular, 2003). Feral dogs were also photo captured in large numbers with relative abundance index of 4.422. The NP might be proving a better place for them as compared to the nearby town. Staff of the park and local people report that packs of feral dog hunt on markhor fawns in eastern parts of the NP. According to the official sources, during 2014-2019 (n=183) feral dogs were killed by the park management in order to get rid of the menace, but the problem still stands (Khattak et al., 2021). A significant rise in the dog population has led to undesirable relations with both public and wildlife worldwide (Kumar and Paliwal, 2015;Home et al., 2017).
Family felidae was represented by Himalayan lynx, leopard cat and common leopard during the current study, each having relative abundance of 4.668, 0.491 and 0.737 respectively. Himalayan lynx has been reported from various parts of Pakistan but authentic and comprehensive reports on its population and interactions with humans are available only from Chitral; where the study area is an important stronghold of the species (Sheikh and Moular, 2003;Din and Nawaz, 2011;Din et al., 2015). The current findings and frequent sightings depict a stable and increasing population of the species in the NP and surrounding valleys. Leopard cat is more cryptic species of the Felidae family as compared to lynx and associated with forest habitats (Roberts, 1997). The species has been reported from Chitral, Azad Jammu and Kashmir and Gilgit Baltistan (Ahmad et al., 2016) The species is data deficient according to the national red list of mammals of Pakistan (Sheikh and Moular, 2003) due to lack of information on its distribution, habitat, population and environment (Sheikh and Moular, 2003;Shehzad et al., 2012). For the last five years, common leopard is frequently sighted and reported from the NP, it was also recorded in the current study whereas historically CGNP was famous for snow leopard. and wide ranging of the big cats and listed as vulnerable due to habitat loss, fragmentation and persecution (Nowell and Jackson, 1996;Bibi et al., 2013;IUCN, 2021). Presence of four subspecies of common leopard has been reported and claimed from different parts of Pakistan but molecular characterization of samples from the country doesn't support this notion (Ijaz et al., 2017). During the past several years, common leopard has been sighted or photographed in the snow leopard range and even in some cases a single camera recorded both leopard species (SLT, 2017). But in Chitral Gol National Park snow leopard has not been sighted or photographed since 2012, which needs plausible explanation. Order Rodentia was represented by Kashmir flying squirrel, golden marmot and Himalayan wood mouse which are frequent inhabitants of alpine and subalpine zones between 1000 to 4000-meter elevation (Roberts, 1997). Other studies have documented around nine species of rodents from the study area (Nasir, 2004). Golden marmot and Himalayan wood mouse are important food source for many large carnivores thus very important for ecological integrity of the park.
It has been established that body mass of the species has some effect on likelihood of being trapped through camera (Bengsen et al., 2011;Anile and Davillard, 2015). Some larger species may move more slowly and they have higher population density which results in a higher relative abundance for species like markhor in the park (Rowcliffe et al., 2008;Bengsen et al., 2011). We recognize that element of bias couldn't be eliminated in calculating RAI by using capture events (Anile and Davillard, 2015). Some species are more frequently seen in the NP premises such as markhor, feral dogs and Himalayan lynx although the relative abundance calculated from this study may not reflect the actual abundance of many species. An optimized survey based on knowledge of species' ecology and familiarity with the area coupled with multi-seasonal trapping efforts would increase the number of species recorded and more reliable estimation of their abundance.

CONCLUSION
The current work is the first comprehensive camera trapping effort on mammalian fauna in CGNP, which confirmed the presence of thirteen different species of mammals in the park. Herbivores like markhor, cape hare and carnivores like Himalayan lynx, golden jackal, red fox, grey wolf and common leopard are some major species dwelling in the NP. Although the previous reports are based on anecdotal surveys, the non-detection may be due to single season survey. Camera trapping study involving different seasons and targeting all different types of microhabitats of the park is suggested.

O n l i n e F i r s t A r t i c l e
Relative Abundance of Mammals of CGNP O n l i n e

F i r s t A r t i c l e O n l i n e F i r s t A r t i c l e
Relative Abundance of Mammals of CGNP