Introduction

For animals, the shelter is an indispensable resource for their survival and propagation (McComb and Noble 1981; Ruggiero et al. 1998). Animals use shelters predominantly as sleeping or resting sites (Miles et al. 1981; Carter and Encarnação 1983; Beck-King et al. 1999; Pardini and Trajano 1999), temporary resting sites (Prestrud 1992; Larivière and Messier 1998), food caches (Post et al. 1993), and natal burrows (Johnsingh 1982). A suitable natal burrow site is a key requirement for many small to medium mammals which provides a safe dwelling for reproduction and hence these structures become significant in population growth function (Kinlaw 1999), by supporting the breeding success of many burrow-dependent species (Alt and Gruttadauria 1984; Ruggiero et al. 1998). Mammals inhabit diverse natal shelter types including tunnels under trees, under mounds at hilltops, under fallen trees, riverbank shelters, rock shelters, and dens or burrows dug on a slope or on the plain ground (Beisiegel 2006). Among them, the underground burrows are highly significant shelters, critical for the survival of burrowing vertebrates (Kinlaw 1999), especially in arid and semi-arid regions with highly fluctuating temperatures (Campbell and Clark 1981; Alkon and Saltz 1988; Reichman and Smith 1990).

In dry and semi-arid conditions, the underground burrows have always been considered as significant ecological shelters, providing appropriate microhabitat conditions for breeding, wintering, and day-time resting for many species (Roper 1992; Kowalczyk et al. 2004). Among mammals, canids are “secondary excavators” and take advantage of the work done by “primary excavators” (Moore 1949; Sullivan 1956; Morrell 1972; Jiménez-Guzmán and López-Soto 1992; Thomson 1992; Yamamoto 1994; Corbett 1995; Macdonald and Courtenay 1996; List 1997; Kowalczyk et al. 2004; Sillero-Zubiri et al. 2004). Instances, where burrowing animals are sharing the burrows, are either obligatory commensals (found in only particular burrow type) or non-obligatory commensals such as “occasional,” “frequent,” and “accidental” (Kinlaw 1999). Also, competition between burrow sharing conspecifics suggests either negative (predation, interference competition) or positive (facilitation) interactions (Kowalczyk et al. 2008). The canids can dig their own burrows, but adapt to energetically efficient behavior by being secondary excavators (Murdoch et al. 2009) and scrupulously use the underground burrows of other species as natal or rearing sites where the pups are born and reared until dispersal (Pruss 1999). For example, the swift foxes (Vulpes velox) often use badger burrows and particularly select natal den sites on the tops of hills with a gradual slope (Pruss 1999) and unoccupied burrows with multiple entrances (Hillman and Sharps 1978). Similarly, kit foxes select natal burrows with more entrances with small entrance diameters (Arjo et al. 2003). However, in arid landscapes, swift foxes and kit foxes select natal sites in positions with little vegetation that maximize visibility from their burrows (Egoscue 1962; Kilgore 1969). Such studies on canids thus reflect the importance of suitable natal sites selection which is often based on burrow location and their physical characteristics (Hillman and Sharps 1978).

The golden jackal (Canis aureus) is one of the most common canids in its ranges of Asia and Europe (Macdonald and Sillero-Zubiri 2004; Giannatos et al. 2005; Jhala and Moehlman 2008). Golden jackals are monogamous (Kleiman 1977; Moehlman 1987, 1989; Asa and Valdespino 1998; Admasu et al. 2004) with rare observations of mate change (Moehlman 2014). Golden jackals demonstrate elaborate pre-copulatory sequences (Golani and Mendelssohn 1971), followed by making or occupying burrows at the onset of the breeding season (Sharma 1998). During the breeding season, golden jackals use burrows for rearing the pups; therefore, selection of suitable burrowing habitat is essential for the successful persistence of their population (Golani and Keller 1975).

In India, golden jackals search for suitable natal sites commences from February to March (Golani and Keller 1975) often preceded by scent marking through urination and defecation around burrows (Jhala and Moehlman 2008) from November through January. Favorable shelter sites for golden jackal include natural or human-modified sites, for, e.g., golden jackal selected open deciduous forests often with areas of sandy soil to build the burrows in Saurashtra region, Gujarat (Soni et al. 1995). They also prefer grassland, medium Prosopis, dense Prosopis, village outskirts, saline wasteland, halophytic scrub, fallow fields, mud flat, road edges, and canal as potential habitat types in Bhal region, Gujarat (Aiyadurai and Jhala 2006; Patil and Jhala 2008). The golden jackal as typical “secondary modifiers” has been observed to modify and use the existing burrows of Indian fox (Vulpes bengalensis), Asiatic gray wolf (Canis lupus), and Indian crested porcupines (Hystrix indica) (Jhala and Moehlman 2008; Mukherjee et al. 2017a).

While there are biological data on golden jackals from previous studies in India, these studies have been conducted on small populations and over short durations. However, long-term assessment and quantification of their activity pattern when they live in a high density and in different habitat condition will provide detailed insights into the biology of the species. Thus, to investigate the detailed spatiotemporal activity pattern, we selected Keoladeo National Park (hereafter KNP) in India, which has recorded highest density of golden jackal (14.84 individuals/km2) ever reported from any protected area around the world (Singh et al. 2016). The golden jackals and other burrow conspecifics that include Indian crested porcupines, leaf-nosed bat (Hipposideros spp.), and Indian rock python (Python molurus) have been observed to compete for the ecologically significant burrows in KNP (Mukherjee et al. 2017a, b). It is, therefore, an imperative to understand the factors influencing the golden jackals to select appropriate natal burrows, which in turn influences their propagation. The present study was thus attempted to understand the pattern of habitat preference for burrow site selection by golden jackals in case of a large population of KNP over a span of 3 years, wherein the existing conditions of habitat availability, social organization, and level of competition varies. In view of this, we explored the biotic or abiotic elements influencing the golden jackals in selecting specific burrows during the breeding season in KNP. The study describes the site-specific habitat characteristics and physical characteristics of these natal sites, and the microhabitat conditions, the environmental conditions, and the presence of conspecifics playing role in site selection by a golden jackal. The breeding season of the golden jackal was also compared with the seasonality of other canids across different geographical regions to comprehend the possibility of any specific pattern in their breeding seasonality. The study also explores the burrow-specific activities particularly during the breeding season and the possible factors affecting the activities of the cathemeral canid in its native range of semi-arid conditions.

Materials and methods

Study area

KNP (27° 7.6′–27° 12.2′ N and 77° 29.5′–77° 33.2′ E, Fig. 1) is in the Bharatpur District of Rajasthan and falls under the semi-arid zone (Province 4A) of India (Rodgers et al. 2002). The total area of KNP is 29 km2, of which 20.5 km2 is terra firma and 8.5 km2 is a wetland area. The park has a mosaic of habitats which includes physiognomic types of forest, woodland, scrub woodland, savanna woodland, low grasslands with scattered trees and scrub, plantations, and wetland (Mathur et al. 2009). The vegetation type and their present status have been recently described in detail (Mukherjee et al. 2017c). KNP is rich in mammal diversity with 34 species (Singh et al. 2017). The mean elevation of the area is ∼ 174 m asl and the temperature varies from 0.5 to 50 °C with annual mean rainfall being 655 mm (Vijayan 1991).

Fig. 1
figure 1

The location of KNP in the state of Rajasthan, India (inset). Map showing all the burrows occupied by golden jackals during the 3 years of the study

Full size image

Identification of covariates

The golden jackal population in Gujarat preferred scrub habitat for burrowing with one to three openings (Jhala and Moehlman 2013) as also shown for Coyotes (Canis latrans) preferring bush-covered south-facing slopes and thicket cover (Gier 1968). In KNP, jackals are known to occupy existing burrows of Indian crested porcupine co-occupied by pythons (Mukherjee et al. 2017a). Therefore, we expected the vegetation parameters around the burrows, the external burrow structural features, and the presence of both porcupine and python would determine the selection of burrows by golden jackal in KNP. Hence, we identified five habitat covariates: Soil Type (ST), Percent Ground Cover (GC), Percent Canopy Cover (CC), Percent Wood Cover around Burrow (WC), Percent Herb Cover Over Burrow opening (WC); four burrow structure covariates: Slope of Burrow Mound (SL), Orientation of Burrow Opening (OR), Number of Openings (NOP), Area of Primary Opening (APM) along with abundance of porcupine (POR) and python (PYT), which were presumed to determine the burrow selection by jackals. The predicted response of the golden jackal to each of these covariates is shown in Table 1.

Table 1 Predicted species response to each covariate based on our a priori hypotheses for selection of burrows by the golden jackals in KNP
Full size table

Field methods

Natal site selection of burrows

The study was conducted from September 2013 to October 2016. A total of 47 burrow systems were recorded in KNP (Fig. 1; Mukherjee et al. 2017a). The geo-coordinates for all the detected burrows (n = 47) were recorded using handheld GPS and information about all identified variables (ST, GC, CC, WC, HC, SL, OR, NOP, APM, POR, PYT) were collected from these sites. Circular plots of the 10-m radius were laid keeping the burrows in the center to assess the percent ground cover, canopy cover, and wood cover around the burrow. Soil samples were collected from each burrow site and the soil type was assessed using feel analysis method (Thien 1979), and classified into six soil types, i.e., clay, clay loam, sandy clay, sandy clay loam, silty clay loam, and silty loam. External burrow characteristics were recorded including slope of burrow mound, the number of openings in a burrow system, the size of openings (in square meter), and orientation of openings (in cardinal degrees). The number of porcupines and pythons present prior to jackal’s permanent occupancy of the burrow was recorded using camera trap images and also by a visual encounter of the animal.

Major activity pattern around the burrows

Among the 47 recorded burrows, a subset of 20 random burrows was selected for intensive monitoring of the major activities around the burrows over 3 years. A total of 11 high sensitive passive infrared (PIR) motion sensor camera traps (Boskon Guard Scouting, BG-520 series) were deployed in 20 selected random burrows. The camera trap trigger gap was set to 10 s in case animals are continuously present in camera view. With each trigger, a set of three pictures were captured. The sampling effort for 20 burrows was systematically divided through the months. The camera traps were placed for five consecutive days in 5–10 burrows depending upon the number of openings and then were shifted to the next set of remaining 20 burrows for another 5 days. This deploying method was repeated for the first 24 months continuously. The efforts have varied for some months due to weather constraints and malfunctioning of some camera traps. In a month, the trapping effort varied from 15 (3 burrows for 5 days) to 300 camera trapping sessions (20 burrows for 5 days with repetition) with a total of 3254 days of camera trap effort.

Every fifth day, the photographs were downloaded and saved with the identity of burrow and month. The photographs were screened and managed for further analysis. The time of the initial trigger was considered, in cases, jackals repeatedly triggered cameras without leaving the field of view. The major activities around the burrows were recorded as burrow inspection, marking, burrow modification, and occupancy during the breeding season. The definition of each of the activities is provided in Table 2.

Table 2 Ethogram of the behaviors by golden jackal recorded around the burrows in KNP
Full size table

Specific activity pattern after occupancy

To understand the specific activity pattern, a pair of jackal was tracked until they chose to select a burrow during the breeding season in 2016 (April–May; dry season). Upon their permanent occupancy, camera traps were deployed and utmost care was taken to minimize the disturbance while changing the traps every third day. The activity of the pack (adults and pups) outside the burrow was continuously monitored until they leave the burrow.

Along with recording the activities of golden jackal, data on environmental parameters were simultaneously collected. The information on lunar phase was obtained from the official website of Indian Meteorological Department, Govt. of India. The ambient temperature and relative humidity data (at 10-min interval × 24-h × 365 days) were recorded using Onset® Hobo™ automatic data loggers, to determine whether any of these values influence the activities of a jackal. Four data loggers were deployed to spatially cover the entire KNP. The data from the nearest logger was extracted for each burrow system. The data were recorded six times in an hour at the interval of 10 min. The hourly mean of the logged data was then used for further analyzing the effect of temperature and humidity on each activity event recorded in a particular hour of the day.

Data analysis

Natal site selection of burrows

To ascertain the effects of various habitat variables on the likelihood of burrow selection by the golden jackal, binomial logistic regression was performed using SPSS v.16.0 (SPSS Inc. 2007) on the data from 47 burrows. A binary response variable (y) was defined for each observation, such that y = 1 if jackal selected and used the burrow and y = 0 if it did not. Binomial logistic regression estimates the probability of an event (in this case, jackal occurrence in a burrow) occurring (Keating and Cherry 2004). If the estimated probability of the event occurring was greater than or equal to 0.5 (better than even chance), the event was classified as “occurring” and if the probability was less than 0.5, the event was classified as “not occurring.” To predict the response of our dependent variable (DV, here jackal occurrence) on other explanatory variables (ST, GC, CC, WC, HC, SL, OR, NOP, APM, POR, PYT), combination of several models were tested in the form of the equation: Y = A + Bx 1  + Cx 2  + …, where Y is the response variable, A is the β-coefficient of the constant, B, C are the β-coefficients of the predictor variables, and x 1 , x 2 are odds ratio. The product of odds ratio and predictor determines the value of the response variable which changes according to the odds ratios. The models were defined based on the β-coefficients and Predicted Group Response (PGR) for each of the models. Usually, the R2 (Cox and Snell R2 or Nagelkerke R2 values) derived in logistic regression are pseudo R2 values. Hence, an approach of bivariate correlation between DV and PGR was considered for obtaining the real R2 values. Different models were tested with different independent variables, followed by step by step removal of variables according to their significance, and finally, the model with highest R2 value was considered the best. R2 gives how much percentage of DV is predicted by the explanatory variables and hence the best model is the one with the highest value (Hosmer and Lemeshow 2000).

Major activity pattern around the burrows

Conceptually jackals recorded by camera traps are considered active only when they move out of the refuge and their activity records were defined as the times of day at which cameras were triggered (Rowcliffe et al. 2014). Frequency data of each activity from the subset of 20 burrows were pooled across 12 months from 20 randomly selected burrows and standardized as per sampled burrow per camera trap days in each month, as the efforts varied across the months. We calculated the mean number of detections per day and compared it across all the months using G-test of independence (McDonald 2009).

Specific activity pattern after occupancy

The specific circadian activity patterns were determined during the breeding season of 2016, wherein the activities of one pack (adults and pups) from one burrow were continuously monitored for 40 days. The data on 24-h timescale was analyzed by using Rayleigh Z statistics (Batschelet 1965; Zar 1999) in Oriana software ver. 4.01 (Kovach 2009). In circular statistic, the mean vector length “r” represents the measure of dispersion of points around the mean vector “μ” (Batschelet 1965) and its length (0 < r < 1) varies according to the concentration of data around the mean angle, where r = 1 indicates minimum dispersion and r = 0 means maximum dispersion (Batschelet 1965). To ascertain the effect of temperature and relative humidity on each of the activities, Pearson’s correlation test was performed in SPSS. To determine the relationship between the intensity of moonlight and jackal’s activity (if any), the activity data was screened only for the night (1900 to 0600 h) hours and then analyzed across eight lunar phases in Oriana software. The mean of each specific activity (in minutes) was analyzed across 24-h timescale and Kruskal-Wallis H test was performed to determine if there are statistically significant differences between the intensity of specific activities across 6 weeks of continuous burrow use by the jackal pack. All the data were tested for normality and appropriate parametric or nonparametric statistics was employed.

Results

Natal site selection of burrows by golden jackal

Of the known 47 burrow system (2013–2014), golden jackal occupied 11 of these burrow systems (Fig. 1). The binary logistic regression suggested Model Y = A + BxAPM + CxWC + DxHC statistically significant (Walds χ2 = 11.84, p < 0.001), with highest R2 value (Table 3). The model explained 78.0% (Table 3) of the variance in different covariate suitability in selecting burrows and correctly classified 95.7% of cases. Jackals were 1.12 times more likely (Table 4) to choose an area with more wood cover (β = 0.12; Walds χ2 = 4.18, df = 1, p = 0.04). Area of primary opening negatively affected (β = −0.03) and herb cover positively affected (β = 0.03) the burrow selection by a jackal. However, the responses of jackal selecting the burrow in corresponding to the latter variables in the model were statistically insignificant (Table 4).

Table 3 Summary of model selection procedure for golden jackal burrow choice in KNP based on correlation between predicted group response (obtained from binary logistic regression) and dependent variable
Full size table
Table 4 Best-fit model parameters, β-coefficient of each determinant variable with their statistical significance and odds of selecting burrows by golden jackal in KNP
Full size table

Major activity pattern around the burrows

A total of 3254 camera trap days’ effort produced 2074 activity records of golden jackal around the subset of 20 burrows. The activities namely “burrow modification” and “rearing pups” were restricted to the months of March–July and April–May, respectively (Fig. 2). The mean detections per day (Table 5) were significantly highest in the month of May for “rearing pups” (5.81 ± 16.31SD; G = 5741.00, df = 11, p < 0.001) and “burrow inspection” (1.45 ± 3.75SD; G = 1131.30, df = 11, p < 0.001). This was followed by “burrow modification” (0.13 ± 1.08SD; G = 161.70, df = 11, p < 0.001) and “marking” (0.10 ± 0.63SD; G = 51.47, df = 11, p < 0.001) in the month of April.

Fig. 2
figure 2

The major activities by golden jackals around the burrow system from 2013 to 2016 in KNP

Full size image
Table 5 The mean (\( \overline{\mathrm{X}} \)) detections per day for major activities throughout year by golden jackal around the burrow systems in KNP from 2013 to 2016. [N = 3254 camera trap (CT) days; F = frequency; SD = standard deviation]
Full size table

Specific activity pattern after occupancy

The occupied burrow by the jackal pack (a pair with pups) was continuously monitored (N = 40 days) and was found active throughout the day with specific activities peaking at a particular period. The time spent outside (Fig. 3) the burrow significantly peaked between 0600 to 0800 h (Z = 37.50, p < 0.001, Table 6). The emergence of jackal from the burrow was not significant for a particular time. However, the retreat into the burrow significantly peaked between 0800 and 1200 h (Z = 7.96, p < 0.001).

Fig. 3
figure 3

Daily records of emergence, retreat, and time spent outside the burrows after permanent occupancy by golden jackals in KNP (bar represents the individual records and the line represents the mean vector and its circular standard deviation)

Full size image
Table 6 Daily emergence, retreat pattern, and peak time spent outside burrows after permanent occupancy; circadian activity patterns outside the burrows, and nocturnal activity of golden jackal during different lunar phase in KNP
Full size table

Other activities (Fig. 4) like feeding, grooming, playing, resting, exploring the surrounding, and guarding significantly peaked between 0500 and 1000 h (Z = 4.04, p = 0.02), 0600 to 1100 h (Z = 5.13, p = 0.01), 0300 to 0900 h (Z = 3.51, p = 0.03), 0630 to 0830 h (Z = 27.14, p < 0.001), 0300 to 0700 h (Z = 6.43, p = 0.002), and 0030 to 0500 h (Z = 5.54, p = 0.004), respectively (Table 6). Frequency histogram of feeding activity also projected three peak time-slots, namely 0130 to 0230, 0830 to 1000, and 1530 to 1630 h (Fig. 4).

Fig. 4
figure 4

Circadian activity of golden jackal adults and the pups outside the burrows in KNP (bar represents the individual records and the line represents the mean vector and its circular standard deviation)

Full size image

The lunar phase was associated with the nocturnal activity of jackal, which was observed to be higher between “third quarter” to “new moon” phase (Fig. 5) with a significant peak (Z = 118.96, p < 0.001, Table 6) during the waning crescent phase.

Fig. 5
figure 5

Nocturnal activity of golden jackals during corresponding lunar phase (one burrow; N = 40 days; total snapshots = 696; bar represents the individual records and the line represents the mean vector and its circular standard deviation)

Full size image

Major activities including “burrow inspection” and “burrow modification” was significantly associated with increasing temperature and decreasing humidity. “Marking” however was associated with neither. Low temperature and corresponding high humidity was significantly associated with (Table 7) specific activities like feeding (RT = −0.47, p = 0.02; RRH = 0.45, p = 0.03), grooming (RT = −0.52, p = 0.01; RRH = 0.46, p = 0.03), playing (RT = −0.48, p = 0.02; RRH = 0.46, p = 0.03), and guarding (R = −0.67, p < 0.001; RRH = 0.66, p < 0.001). Any change in temperature or humidity did not significantly affect other activities including, emergence, retreat, exploring surrounding, huddling, and resting.

Table 7 Pearson’s correlation between temperature (RT) and humidity (RH) influencing the activities of golden jackals around the burrow system in KNP
Full size table

Intensity of activity around the burrow

A total of 696 activity records were captured during the breeding season of a jackal that lasted for 40 days. The intensity of each activity was analyzed across 6 weeks (Fig. 6). The activities including feeding (χ2 = 30.76, p < 0.001), grooming χ2 = 33.85, p < 0.001), guarding (χ2 = 28.91, p < 0.001), and playing (χ2 = 34.65, df = 5, p < 0.001) significantly occurred between fourth and sixth weeks (Table 8). However, “exploring surroundings” and “resting” were not significantly restricted to any week. Jackals spent an average of 103.60 min (±154.74SD) per 24 h on playing, followed by guarding (36.13 min), grooming (20.18 min), feeding (6.95 min), resting (3.74 min), and exploring the surroundings (1.17 min).

Fig. 6
figure 6

Frequency of activities outside the burrow across 6 weeks during the breeding season of the golden jackals in the year 2016 in KNP (one burrow; N = 40 days; total snapshots = 696)

Full size image
Table 8 The mean time spent on various activities outside the burrow by the golden jackal breeding pair and the pups in 24 h during the breeding season of 2016. (N = 40 days; total snapshots = 696)
Full size table

Discussion

Suitable and safe natal burrow site is important for the survival of golden jackal populations wherein they give birth and rear their young ones. The results of the study indicate that the golden jackals in KNP frequently visited the burrows throughout the year. The general activities around the burrows begin with “inspection” of the burrow sites and scent “marking,” followed by specific activities such as “burrow modification” and “rearing of pups” restricted to March–July and April–May, respectively. The jackal in KNP occupied 23.4% of the 47 burrows and selected natal sites with more vegetation cover and small openings. During the pup rearing season, jackals were found active outside the burrow throughout the day, executing specific activities around the burrows. These include feeding, grooming, playing, resting, exploring the surroundings, and guarding. The burrow-specific activities during pup rearing season lasted for 6 weeks.

Though the golden jackals were frequently seen around the burrows, their nocturnal activity significantly peaked between “third quarter” and “new moon” phase in KNP. Similarly, other canids have also shown this peak in activity, for, e.g., crab-eating fox (Yanosky and Mercolli 1990; Faria-Corrêa et al. 2009) and maned wolf in Brazil, wherein the latter traveled significantly more on nights of the new moon than compared to the full moon (Sábato et al. 2006). It is likely that the increased activity in the last quarter and new moon nights could be associated with a circumvention behavior including possible detection by prey and protection from predators (Yanosky and Mercolli 1990; Sábato et al. 2006).

In KNP, interspecific burrow sharing and facilitative interaction have been observed between the primary burrower, i.e., Indian crested porcupines, and non-obligatory commensals, i.e., golden jackals, leaf-nosed bats, and Indian rock pythons (Mukherjee et al. 2017a, b). In the present study, the jackal as “secondary modifiers” (Kinlaw 1999; Sillero-Zubiri et al. 2004) has occupied only 20 to 25% of the existing porcupine burrows in KNP during April–May. The present data is similar to the previous findings on several other canids which are also secondary excavators. For example, raccoon dogs, used 8% of the Eurasian badger burrows in Žemaitija National Park, Lithuania (Ulevičius 1997); raccoon dogs and red foxes occupied 33 and 20% of the burrows of Eurasian badgers, respectively, during the breeding season in Białowieża Primeval Forest, Poland (Kowalczyk et al. 2008). Likewise, kit fox, dingo, crab-eating fox, short-eared dog, and bush dog were also known to use existing burrows of prairie dog, wombat, armadillo, and paca (Sillero-Zubiri et al. 2004; Faria-Corrêa et al. 2009). Since jackals inhabit the burrows only for a short period during their breeding season (Jhala and Moehlman 2004), they seem to have opportunistically selected the existing burrows to avoid energetic cost in digging new burrows (Vleck 1979; Frafjord 2003; Zelová et al. 2010). In addition to regular visits to burrow sites, the jackals selected abandoned Indian crested porcupine burrows as the natal sites in locations with more wood cover and entrances with small opening with more herb cover. The thick cover at burrow sites of golden jackal reduces the predation risk of vulnerable pups as also observed by Saad et al. (2015) in Jhelum District of Pakistan. Similarly, other canids such as female red wolf (Canis rufus) in eastern North Carolina, USA, seek shelter in shallow surface depressions located in dense vegetation (Kelly et al. 2004). Likewise, the natal burrows of kit foxes in Tooele County, Utah, had small entrances with high vegetation cover. The physical burrow characteristics and the surrounding habitat are critical for the successful rearing and propagation of the young ones (Bekoff and Wells 1982).

The mean detections of the pups and rearing of pups peaked during April–May corresponding to the rearing season of golden jackal in Israel (Golani and Keller 1975) and India (Jhala and Moehlman 2008). Most of the canids are seasonal breeders (Holt et al. 2014), albeit canids such as bush dog exhibiting no seasonality. The breeding timings of different canid species differed across the globe; however, the season of the breeding remained during the dry season. For example, the short-eared dog in Peru (Leite Pitman and Williams 2004), maned wolves (Chrysocyon brachyurus) in South America (Rodden et al. 1996, 2004), the Indian fox (Acharjyo and Misra 1976; Johnsingh 1978; Johnsingh and Jhala 2004), and the dholes (Cuon alpinus) (Davidar 1973; Johnsingh 1982; Venkataraman 1998) in India are all breeding during the dry spell. In India, the dry spell is followed by the monsoon and several of these abiotic factors affect the breeding onset and subsequent birth of young ones (Lehman et al. 1997; Prendergast 2005; Holt et al. 2014), thus optimizing the survival of the young ones, due to factors such as appropriate ambient temperature, abundant nutrition and water availability, and probable variations in the predation activities of other species (Moehlman 1979, 1983, 1987).

The specific activities of jackal around the burrows are throughout 24 h during breeding and rearing season, which is in concordance with jackal’s cathemeral behavior (Sharma et al. 2013). Specific activities like feeding, grooming, playing, resting, and exploring the surroundings significantly peaked during early hours of the day in concurrence with lower temperatures. Increased temperature with the day’s progress significantly reduced the activities of the jackal around the burrows and they were observed retreating around 1200 h. Guarding of the burrows by adults significantly increased during night hours (0030 to 0500 h), when the pups are more susceptible to predation. Some predators like striped hyena can be a threat to the unattended pups. Similarly, pup-guarding by adults was observed in other canids, e.g., African wild dog (Lycaon pictus) (Courchamp et al. 2002), hoary fox (Lycalopex vetulus) (Dalponte and Courtenay 2004), and pampas fox (Lycalopex gymnocercus) (Lucherini et al. 2004). The pups were suckling every 8 h, around 0200, 0900, and 1500 h, and when the pups were about 3 weeks old, in addition to milk, regurgitated food also became part of the diet. This was similar to the golden jackals in Serengeti (Estes 1991) where the pups suckle at least five times a day and twice at night for the first month followed by consuming disgorged food by the end of the month.

It was observed in the present study that both the sexes participated in parental care activities, taking turns to guard the burrow, grooming, and playing with the pups. The male exhibited female care, including food supplementation prior to birth and throughout nursing, which is similar to male parental care in few other canids (Moehlman 1978; Moehlman 1989; Sillero-Zubiri et al. 2004; Dalponte and Courtenay 2004; Lucherini et al. 2004; Bothma 2013). In few quantitative field studies of jackals (Moehlman 1978; Malcolm 1985), the female attendance around the burrows dropped to ~ 35% of the time from the third week onwards as they were engrossed in extensive foraging after the first 2 weeks of confinement, with subsequent increase in paternal attendance to almost twice the time from the second to the fourth week.

The breeding success of all species depends on several factors, of which availability of a suitable natal site and a microhabitat as a refuge for rearing the young is a vital factor. Except in Antarctica, canids occur across the globe (Sillero-Zubiri et al. 2004), and hence, the availability and use of an appropriate burrowing site for rearing their young ones are highly significant for the propagation of their population. Despite prominence in their range and some of them being highly generalist, they still prefer to select safe shelter for breeding to avoid running into the risk of predation (Tannerfeldt et al. 2003). In view of this, the study is particularly relevant as it delivers a collective and detailed account of the use of the sub-surface earthen refuge by the golden jackal in semi-arid conditions of KNP. The present study affirms that the jackals are highly selective of their natal sites with a preference for appropriate cover. The burrow usage is largely restricted during the breeding season, corresponding to the dry season prevailing in the region. The rearing of the pups around the burrows lasted for 6 weeks, wherein the adults and the pups exhibited true cathemeral behavior and remained active throughout the day. The low temperature conditions and changes in lunar phase influenced the variations in burrow-specific activity patterns including the timing of emergence and retreat from the burrows. All the comprehensive information thus provides noteworthy inferences for further understanding of the biology of this cathemeral canid reportedly experiencing a decline in its population across its range including India (Giannatos et al. 2005; Jhala and Moehlman 2008).