Small cat surveys : 10 years of data from Central Kalimantan , Indonesian Borneo

We present an update on the photographic detections from camera traps and the activity patterns of Borneo’s four small cats, namely, Sunda Leopard Cat Prionailurus javanensis, Flat-headed Cat P. planiceps, Marbled Cat Pardofelis marmorata, and Bay Cat Catopuma badia, at two sites in Central Kalimantan, Indonesia. Camera trap survey data of 10 years (2008–2018) from the first site in Sebangau provide details about the temporal partitioning of these small cats from each other but overlap with Sunda Clouded Leopard Neofelis diardi. The activity of Flat-headed Cat was higher after midnight and that of Leopard Cat at night with no clear preference before or after midnight. The Marbled Cat is predominantly diurnal, but the remaining three cats have flexible activity periods. While limited data are available from Rungan, the second site, we confirmed the presence of all four small cat species found on Borneo, though we have insufficient data to comment on the Bay Cat. The cat sightings, however, are intermittent and may reflect the unprotected status of this forest. Leopard Cats appear relatively unaffected by habitat disturbance based on encounter rates on camera traps. Conservationists, both NGOs and the government, must pay particular attention to specialists like Flat-headed Cats and Bay Cats when assessing habitat suitability for long-term cat conservation.


INTRODUCTION
In the absence of Tiger Panthera tigris, Borneo's cats represent a fascinating guild, and all are in need of conservation attention.Five species occur on Borneo: the Sunda Clouded Leopard Neofelis diardi (Endangered (EN) on the IUCN Red List; Hearn et al. 2016c), the Bay Cat Catopuma badia (EN and endemic to Borneo;Hearn et al. 2016a), the Flat-headed Cat Prionailurus planiceps (EN; Wilting et al. 2016a), the Marbled Cat Pardofelis marmorata (Near Threatened; Ross et al. 2016a), and the Sunda Leopard Cat Prionailurus javanensis (Least Concern; Ross et al. 2016).They are all protected under Indonesian law (P.20/MENLHK/SETJEN/KUM.1/6/2018).As they are rare and elusive, it is difficult to study them in the wild.Thus, there is limited knowledge about their ecology despite increased scientific interest.
Since 2008, Borneo Nature Foundation (BNF) and University of Palangka Raya, Centre for the International Management of Tropical Peatlands (UPR-CIMTROP), have been conducting long-term monitoring of the Sunda Clouded Leopard Neofelis diardi in Indonesia and, specifically, N. d. borneensis in Central Kalimantan.During this time, we also opportunistically obtained images of three of the four small cats in the Sebangau catchment area.Given the disturbed mosaic nature of Sebangau, we also hope to determine which areas are unsuitable for small cats, either due to habitat changes and/or due to human disturbance.Live Leopard Cats are more common in Kalimantan markets for sale as pets than Sunda Clouded Leopards, and their skins are more often found in homes (Rabinowitz et al. 1987).Far less is known about small cat movements, habitat preferences, seasonal movements, breeding patterns, and effects of anthropogenic disturbance across their range.These data are particularly lacking from tropical peat-swamp forests.We present here updated information on temporal activity pattern of four small cats from the Central Kalimantan region of Indonesia.

Study areas
The Sebangau catchment, Central Kalimantan, Indonesia (Fig. 1), is a peat-swamp forest (mixed-swamp forest sub-type) covering an area of ~5,600km 2 .This study took place in the 50km 2 research forest located in the northeast.The area was logged under a concession system between 1991 and 1997 followed by illegal logging between 1997 and 2004.The site is at an altitude of about 10m.The area was significantly affected by the forest fires that impacted Indonesia in 2015.

MATERIAL AND METHODS
Since 2008, a total of 210 camera locations were surveyed across both forest areas with an average of 30 units in each forest area at any one time.Between 2008 and 2012, cameras were set up in pairs in Sebangau and subsequently as single units.In the Rungan site, cameras were all set up as single units.Cameras were set in a stratified random survey design.Cameras were placed 500m to 1,000m apart and were in each location for a minimum of six months; some cameras were in the same location since May 2008.Locations were selected to cover a range of habitats and disturbances within the forests, avoiding streams and slopes wherever present.Camera traps were placed along established human-made trails (more than four years old) and, where possible, watering areas, to maximise the success rate of photographic captures.A combination of camera models were used, including Cuddeback Expert®, Cuddeback Capture IR® (Cuddeback Digital, Non-Typical Inc, WI, USA) Maginon, Crenova, and Bushnell.Cameras were checked every 40 days when batteries were changed and SD cards exchanged.Data were managed in a custom Microsoft Access database.Active behaviour times were calculated using the kernel density method ('href' bandwidth for kernel smoothing; Ridout & Linkie 2009;Meredith & Ridout 2016) to account for average dawn and dusk times in the sites, which are situated almost on the equator (for more information on the Sebangau study site, see Cheyne & Macdonald 2011;Cheyne et al. 2016b).Detection rate was estimated as number of detections/100 trap nights.Weather data were collected daily at each research site and fire data was obtained from the Indonesian Agency of Meteorology, Climatology, and Geophysics.A 30-min interval between photos of the same species was used to determine if photos of were an independent event at the same location and date.

RESULTS
The number of camera trap (CT) stations at each site varied annually due to broken units (Table 1).
Detailed descriptions of CT locations are in Appendices 1 & 2, including descriptions of the microhabitats, number of trap nights, number of detections, and careful descriptions of the setup around the CT stations.CTs were placed in different habitats across the two sites (Table 2).
All small cats in Sebangau were photographed in all main habitat types in the interior forest, <20m from the forest edge and in disturbed areas.All four small cats in Rungan were recorded in the interior forest; only the Flatheaded Cat was recorded near the lake.

Sebangau
Since the cameras were first placed in May 2008, we captured 157 independent images of Sunda Clouded Leopards (Image 1), but only 109 of Sunda Leopard Cats, 54 of Marbled Cats (Image 2), and 33 of Flat-headed Cats (Image 3).Compared to the average detection rates of small cats since the inception of the camera trap study in 2008, there was a decline in the detections of Marbled Cats and Flat-headed Cats (Fig. 2).From 2014 to July 2018, there was an average of 3.9 independent Marbled Cat images/month (min=0, max=16).No Flat-headed Cats were recorded by camera traps between January 2014 and February 2018 (Fig. 2), which coincided with a significant fire event from September to November 2015.
Of the three small cat species, the Sunda Leopard Cat is predominantly nocturnal with no clear preference for time of night.Flat-headed Cats also showed nocturnal activity but with a slight preference for post-midnight hours.Marbled Cats are strongly diurnal.Interestingly, 65% of 115 nocturnal records (18.00-05.59h) of Sunda Clouded Leopards were between 01.00h and 05.59h (Fig. 3), thus overlapping with the preferred active time for Flatheaded Cats.

Rungan
Three of the small cats were confirmed in Rungan in the first few months of the study but it took 12 months to confirm the presence of the Flat-headed Cat (Table 4).

Bay Cat
With over eight years of long-term camera trap surveying in the peat-swamp forest of the Sebangau catchment, the Bay Cat was not detected, and ongoing work suggests that it is not found in peat-swamp forests.The Bay Cat was confirmed in mosaic heath/ peat-swamp forest habitat for the first time (Sastramidjaja et al. 2015;Cheyne et al. 2016aCheyne et al. , 2017)).Through the use of camera traps, we present new location information on the distribution of Bay Cat in Kalimantan.This new location is approximately 64km southeast outside the range depicted by Hearn et al. (2016a).Our record of Borneo Bay Cat from the new habitat (heath/ peat-swamp forest) warrants further surveys in different habitat types to fully understand Bay Cat distribution and ecologic needs.

DISCUSSION
The small cats are appearing evenly across the habitat types in both Sebangan and Rungan, with the exception of the Bay Cat that likely does not exist in deep ombrogenous peat-swamp forest (Sebangau).Additionally, we have evidence of breeding in Flat-headed Cat and Marbled Cat in Sebangau (images of kittens) (Images 4 & 5).
Flat-headed Cats have a more irregular capture rate and though they are active throughout the day, more captures are obtained at night and therefore they are predominantly nocturnal.Leopard Cats have a more regular capture rate    and appear to be active both during the day and night, though they appear to avoid the hottest time of the day (11.00-13.00h).Marbled Cats have a regular capture rate with the majority of images taken during the day (05.00-16.00h), suggesting they are diurnal.There is only one image of a Bay Cat taken at 11.17h.These data are similar to those of Hearn et al. (2018), though these authors did not obtain sufficient images of Flat-headed Cats to make a detailed analysis.Peat-swamp and associated lowland wetlands are postulated to be an important habitat for Flat-headed Cats (Cheyne et al. 2009;Wilting et al. 2010Wilting et al. , 2016b;;Cheyne & Macdonald 2011;Adul et al. 2015).Marbled Cats are not believed to frequent roads or plantations (Hearn et al. 2016c) and prefer intact forests, though data are lacking on this cat (Rustam et al. 2016).
Peat-swamp and associated lowland wetlands were suggested to be poor or marginal habitat for Sunda Leopard Cat (Mohamed et al. 2016), but our work suggests that Sunda Leopard Cat are far more common (Cheyne & Macdonald 2011;Adul et al. 2015;Cheyne et al. 2016b).
The infrequent capture of the small cats in both sites is likely an artefact of the placing of the cameras (±1km apart) to focus on the wide-ranging Sunda Clouded Leopard.By moving the cameras closer (±500m) we hope to determine the population density for the small cats, determine if the Bay Cat is indeed absent from this forest, and to continue our monitoring of the Sunda Clouded Leopard population.The long period of time required to obtain images of small cats, possibly due to the placement of the cameras targeting Sunda Clouded Leopard, highlights the importance of long-term data and monitoring to avoid false-negative presence data.Sunda Leopard Cat is the most commonly recorded species in the study site.Marbled Cat is hard to study and, as many are arboreal, having cameras mainly on the forest floor means we could be missing out on key aspects of their behaviour.Flat-headed Cat is a wetland specialist and prefers forests with water (Wilting et al. 2010).Its diet likely consists of fish, frogs, and small mammals, and it may fill a niche on Borneo filled by the Fishing Cat Prionailurus viverrinus elsewhere in Asia (Iwaniuk et al. 2001).Due to this dependence on wetlands, we think that the devastating fires of 2015 may have severely impacted the Flat-headed Cat population, pushing it away from fire-affected areas.Our preliminary results suggest that Flat-headed Cats are returning to these areas, which BNF is actively working to restore.These data could be an artefact of survey effort (a high number of broken cameras) or a genuine reflection of this species behaviour.
Large parts of the peat-swamp are naturally seasonally flooded for up to nine months per year.The flooded nature of Sebangau does not always prevent the animals using the ground; indeed we have evidence of male Orangutans Pongo pygmaeus wading through water (Ancrenaz et al. 2014).Keeping a selection of camera trap locations the same over several months or years allows for variations in detection to be accounted for; given the regular flooding of the forest, it is likely that the wildlife is accustomed to this.We do notice animals using our boardwalks more regularly in the wet season.Peatlands and associated forest fires are a crucial conservation concern in Kalimantan (Gaveau et al. 2016;Miettinen et al. 2017).This is especially true during dry years such as in 2015 when a strong El Niño event led to particularly dry conditions.From August to November of that year, MODIS satellites detected over 50,000 fire hotspots in Kalimantan, 53% of which were on peatland (Gaveau et al. 2016;Miettinen et al. 2017).Since 2006, 17.35% of forest in the core Sebangau research area burned down (9.63% in 2015 alone-5.3km 2 of 55km 2 ).Of particular threat to the Flat-headed Cat are peat drainage and drying out due to logging canals, the loss of permanent water, and increased hydrologic instability (Page et al. 2009;Vanthomme et al. 2013).

Conservation impact
This work represents the first-ever comprehensive and long-term survey of small cats in Central Kalimantan.There is a severe lack of data on these species in nonprotected or small forest areas that may also contain viable populations.It is crucial to remember that, while these surveys indicate the continued presence of these cats, habitat loss, wildlife trade, and likely presence of populations in non-protected areas means that more work is needed to understand the impacts of anthropogenic activities on these cats.As detailed in Appendix 1, this project provides extensive and detailed data about many wildlife species in Sebangau and Rungan forests in addition to the cats-an additional 7,959 images (2,765 videos) of 74 species.Of these, two are IUCN Red Listed as Critically Endangered, five as Endangered, 14 as Vulnerable, 12 as Near Threatened, and 41 as Least Concern.

Figure 2 .
Figure 2. Mean detection of small wild cat species from the inception of the camera trap surveys in Sebangau since 2008.

Figure 3 .
Figure 3. Kernel density estimates of activity patterns of species using alpha of 0.1 smoothing parameter: a -Flat-headed Cat; b -Marbled Cat; c -Sunda Leopard Cat.