Impact of Soil Types on the Abundance of Murid Population in North Punjab, Pakistan

North Punjab (Pakistan) includes hilly undulates of Pothwar Plateau and northern Parts of the Indus Plain is land of two major rivers Indus and river Jhelum, several seasonal streams, ravine belts and cultivated areas. The study area has various mountain ranges, such as, Kala Chitta Range, Margalla Hills Range, Salt Range and Khairi Murat (www.dawn.com/weekly/herald/herald33.htm). The climate is sub-humid to semi-arid, annual rainfall varies from 38 cm in southwest to 150 cm in the northeast (Anonymous, 1976).

The small mammals were trapped from 8 randomly selected localities from four districts, viz., Attock (Chura Sharif, Makhad Sharif, Mankur), Chakwal (Bharpur, Noor pur Sethi, Khokherzer), Rawalpindi (Dullah), Jhelum (Sroba). In each location uncultivated wild land was selected, where four trap lines transect were set trapping rodents over some 100 hectare following (Cunningham and Moors, 1996). A trap-line transect consisting of 30 trap stations laid on a straight line at of distance of 17 m from each other. At each station three snap traps (2 Rat and one mouse) were set and trapping executed for three nights consecutively. Each trap was baited with bread with application of vegetable ghee (Cunningham and Moors, 1996). Trapping nights were calculated by multiplying number of traps with number of nights. The traps were set late in the evening and checked next morning, specimen removed and identified to species level using key (Robert, 1997) and recorded separately for sampling site and day.

Three soil samples (total 24 from eight villages) were taken from each locality from at a depth of 30 cm, which were packed in plastic bags for laboratory processing.

The texture class was determined as relative proportions of different soil fractions (dried and grinded sand, silt and clay) using the USDA textural triangle. The soil samples (10g) were taken in a 600 ml beaker to which 60 ml of dispersing solution (40 g of sodium hexametaphosphate, 10 g of sodium carbonate dissolved in distilled water and the volume made up one liter) was added and left for 24 h. The contents of beaker were transferred into soil-stirring cup, and the cup was filled with water to about three-quarters. The suspension was stirred at high speed for 3 min using the special stirrer, when stirring paddle was rinsed in the cup and was allowed to stand for 1 min. The suspension was transferred into a 1-L hydrometer jar and distilled water added to make a volume of 1 L (Bouyouces, 1962). The heights of clay, silt and sand layers was recorded and relative proportions calculated by dividing each height with total thickness of the solid layers and converted into percentiles for the proportional representation of each soil component.

The saturation paste prepared by dissolving soil sample in KCl solution (0.7456 g in 1 liter of distilled water). was passed through moist filter paper and extract obtained by applying pressure. The electrical conductivity determined by passing the extract through conductivity meter in mS cm-1 or dS m-1. Saturated paste of soil (10 g) was prepared by adding deionized water and allowed to equilibrate for 18-24 h, pH was measured with pH meter (PHS-3D-02 pH Meter) (Schofield and Taylor, 1955).

Principal components analysis (PCA) was used to find association between rodent population abundance and soil content (sand, silt and clay).

Table I shows a total of eight soil types recognized on the basis of relative proportions of three basic soil components, i.e., clay, silt and sand The percentage of sand varies from 20% to 85%, silt from 7% to 55%, and clay from 8% to 25%. The soil samples were normal to alkaline with pH ranging from 6.9 to 8.0 and EC varied from 0.3 to 0.92 EcdS m-1.

We captured 1571 samples which belonged to the following 6 species Nesokia indica, Mus musculus, Mus budooga, Tatera indica, Millardia meltada and Golunda ellioti trapped over 3840 trap nights (Table II). Trapping success was the highest in loam (66.6%), consecutively followed by clay loam (55.2%), sandy (42.7%), loamy sand (35.4%), clay loam (34.3%), sandy loam (33.3%), clay loam (34.3%), silt loam (31.4%) and the lowest trapping success in silt clay loam (28.1%) (Table II).

The house mouse (Mus musculus) prefers sand soil type (22%) followed by loamy sand (19%), sandy loam (17%) sandy clay loam 14%, loam 11%, clay loam 8%, silt loam 6% and least preferred soil was silty clay loam (3%) (Table III). This species shows negative relation with silt and clay and no relation with EC and pH, in correlation matrix of principal component analysis. The most preferable habitat

Table I.- Characteristics of the soil and coordinates of at different localities.

of Indian field mouse (Mus booduga) was sand (24%), followed by loamy sand (20%), sandy loam (16%), sandy clay loam (12%), loam (10%), clay loam (8%) and lowest preferred soil type was silt loam (4%) and silty clay loam (4%). This species also show similar behavior like Mus musculus and show negative relation with silt and clay and no relation with EC and pH, in correlation matrix of principal component analysis.

The Indian gerbil (Tatera indica) prefers loam soil (46%), followed by clay loam (26%), sand (18%), silt loam (7%) and least preferred soil was silty clay loam (3%). There was no capturing of T. indica in soil type, loamy sand, sandy loam and sandy clay loam (Table III).

The order of preferred soil by fossorial rodent species the short-tailed rat (Nesokia indica) was clay loam (20%), (Loam (18%), sandy clay loam (16%), sandy loam (14%), loamy sand (11%), silt loam (10%), silt clay loam (9%) and sand (2%) (Table III).

The complete absence of Indian Bush Rat or coffee rat (Golunda ellioti) in soil type sand, loamy sand and sandy loam was registered. The most preferred soil for this rat species was silty clay loam (30%) followed by silt loam (26%), clay loam (22%), loam (17%) and least preferred soil was sandy clay loam (4%) (Table III). The species show negative relation with sand, silt and clay in correlation matrix of principle component analysis (Supplementary Table I).

The eigenvalues was the highest in F1 (6.663), consecutively followed by F2 (2.094), F3 (1.144), F4 (0.729), F5 (0.285), F6 (0.055), and F7 (0.030) (Supplementary Table I). The eigenvalue corresponds to a factor in one dimension. The eigenvalues correspond to a high % of the variance; ensure that the factors are a good quality projection (Fig. 1).

The correlation circle on axes (F1 and F2) shows a projection of variables in the factors space. The variables Nesokia indica, Tatera indica density (loam are far from the center, and are significantly positively correlated (r close to 1). The variables (G. elliotti, M. meltada) density (silty clay loam) are also far from the center and are significantly positively correlated (r close to 1). The distribution of Mus musculus and Mus booduga and sand concentration are also far from the center, and are significantly positively correlated (r close to 1) (Fig. 2).

The study revealed six species of rodents (Table III) in six type of soil (sand, loamy sand, sandy loam, sandy clay loam, loam, clay loam, silt loam, silty clay loam). We identified three type of rodent communities: Mus musculus and Mus booduga (Community 1), Nesokia indica and Tatera indica (Community 2), Glounda elliotti and Millardia meltada (Community 3). The composition of rodent community varies in different soil types. The Community one is widely distributed and abundant in sand soil type, while the community two is widely distributed and abundant in loamy soil type. The distribution and

abundance of third rodent community is limited to silty clay loam soil type. The results support the conclusion that soil type is limiting factor in distribution and abundance rodents according to the findings of Massawe et a. (2008).

The Indian bush rat (Glounda elliotti) is widely distributed and abundant in silt clay loam soil type and the complete absence in soil types such as sand, loamy sand and sandy loam, indicates that soil texture (sand, silt, clay) directly affect the distribution and abundance of these rodent species and is support the results according to finding of (Reichman and Smith, 1990). The distribution of (G. elliotti is already reported in silty soil (http://www.iucnredlist.org2017). The samples of the species was captured from vicinity of rocky and hilly tracts and analysis of soil result (silty clay loam) is also confirmed by previous findings (Molur et al., 2005)

The species Millardia meltada is widely distributed in south Asia (Sri Lanka, India and Nepal) and is Least Concern (http://www.iucnredlist.org2017), with no major threats known affecting its population (Molur and Nameer, 2008). The results of our study indicate that the species shows similar habitat association like Glounda elliotti and is distributed in silty clay loam soil types at cultivated crop arid lands and our results are also conform the findings of Molur et al. (2005).

The species Tatera indica is widely distributed in south Asian countries (Molur and Nameer, 2008) and is reported in abundance at Pothwar (Punjab) (Faiz et al., 2015). The results of our study also indicate burrowing rodent Tatera indica prefers loamy soil (Fig. 2) of dry river slopes which is mostly loamy and confirm our findings.

The species Nesokia indica is also widely distributed in south Asian countries but there are no study on its population abundance known to date (Molur et al., 2005). The results of our study also indicate that it is an example of another one species which prefers loamy soil (Fig. 2).

The house mouse (Mus musculus) is widespread, abundant and commonly regarded as a pest species (Giovanni Amori, 2007). The results of present study indicate that the species show direct relation with soil type (sandy loam) and an inverse relationship with the percentage of silt and clay (Feldhamer, 1979). It was captured from all soil types.

The mouse Mus booduga is another widespread, abundant rodent native to South Asian countries (India, Sri Lanka, Pakistan, southern Nepal, Bangladesh) (Molur et al., 2005). The results of present study indicate that it is also shows direct relation with soil type (sandy) and an inverse relationship with the percentage of silt and clay (Feldhamer, 1979). It was as well captured from all soil types.

There is supplementary material associated with this article. Access the material online at: http://dx.doi.org/10.17582/journal.pjz/2018.50.1.sc8.

Authors have declared no conflict of interest.