1. Introduction

Macrotermes bellicosus (Smeathman) belongs to group of insects commonly called termites. Termites are polymorphic social insects that live in self-constructed mounds called Termitarium. The individual termites within the termitarium constitute a colony that comprises several castes that are morphologically and functionally distinct (Lee and Wood, 1971; Richards and Davies, 1977). They are soft-bodied, moderate to small size insects with biting mouthparts. Body of flying individuals is dark, but those that remain in the nest are whitish with only their heads being pigmented heavily (Krishna and Weesner, 1969; Richards and Davies, 1977). The wings in termites are temporary, long and slender and in two pairs that are similar to each other. They are deciduous and are quickly shed with a single flick when the swarming termites find a new nest site, pair up and dig in (Collins, 1981; Ekpo and Onigbinde, 2007). M. bellicosus is a fungus- growing termite that cultivates symbiotic fungi in their nest (Binate et al., 2008) and was considered a popular termite in Nigeria (Ekpo and Onigbinde, 2007). It was reported as pest of several crops such as Maize (Wood et al., 1980), Groundnut (Johnson and Gumel, 1981), Sugarcane (Boboye, 1986), Rice (Alam, 1992) and Cocoa (Ndubuaku and Asogwa, 2006). Termite mound is one of the common features of most agro- ecosystem in tropical Africa (Wood and Sands, 1978; Collins, 1983; Yamada et al., 2005). The mounds of Macrotermes were found to affect the tree flora of several ecosystems being a source of heterogeneity in the landscape (Traore et al., 2008).

The population of the mounds tend to prevent farmers from utilising land masses for cultivation where their densities were high (Ehiet al., 2007). Trapnell et al. (1976) estimated a density of 3-4 mounds ha^-1 in Zambia for Macrotermitinae, Pomeroy (1978) reported 1-4 mounds ha^-1 for Macrotermes in Uganda, Abe et al. (2009) reported a 3- 10 mounds ha^-1 for M. bellicosus. Studies by Collins (1981) and Eggleton et al. (1996) on pattern of distribution of mounds of M. bellicosus revealed a random and clump distribution pattern in Southern Guinea Savanna of Nigeria and Cameroun, respectively.

The present research is undertaken to study the density and distribution pattern of mounds of M. bellicosus being a popular and most prevalent mound- building termite species in Sokoto State, Nigeria.

2. Materials and Methods

The Study Area

The research was carried out in Sokoto State, located in the Northwest of Nigeria, between latitudes 11°30’N and 14°00’N and longitudes 4°00’E and 6°40’E. Sokoto State covers a total land area of 32,000 km² (Abdu et al., 1982; Mamman et al., 2000; Tureta et al., 2006). The State has a tropical continental climate and entirely falls within the semi-arid climatic environment. The annual rainfall is between 500 and 750 mm with a high peak in August and mean monthly temperatures varying between 13℃ in December/ January and 42℃ in April, while the average annual temperature is 34℃ (Kowal and Knabe, 1972; SERC, 2010).

Four (4) Local Government Areas in the State were selected following preliminary studies of areas with high number of termite mounds, namely; Shagari, Wamakko, Wurno and Yabo.

Selection of Sample Plots

In each Local Government Area, six (6) sample plots of 1 hectare (500 m x 20 m) were selected, from which three (3) plots each, from both fadama and upland were chosen by applying the standardised protocols of Jones and Eggleton (2000).

1) Survey of Mounds of M. bellicosusin the Sample Plots

A transect survey was done on foot in each of the selected sample plots to identify the mounds of M. bellicosus in each Local Government Area. Mounds belonging to M. bellicosus within the sample plots encountered were counted. Similarly, distance between mounds was measured using a measuring tape. The method used by Collins (1981) and Lepage (1984) for determining spatial distribution using the dispersion coefficient was employed in the study for determination of spatial distribution pattern of the mounds. The dispersion coefficient of the mounds was calculated using the Clark and Evans (1954) formula for assessing the dispersion coefficient;

Where, a is the dispersion coefficient, m is the real density and d is the average distance between mounds. The real density was determined using the formula;

Where, n is the number of mounds sampled and s is the area sampled in m² as applied by Meyer (1997). Values where a=0 are described as having clump distribution, a= 0.25 as random distribution and a= 1.158 as regular distribution (Collins, 1981).

2) Statistical Analysis

Data generated were subjected to one-way analysis of variance where significant differences between means were separated using Least Significant Difference (LSD) at p<0.05. Analysis was done using SAS 9.3 Statistical package (SAS, 2003^®).

3. Results and Discussion

Densities of mounds of M. bellicosus in the study area are shown in Table 1. A total of 242 mounds with a mean density of 10.08 were observed in the 24 plots. Wurno Local Government Area had the highest mean of 18 mounds ha^-1, while Shagari had the least 4.67 mounds ha^-1(Table 1). The observed mean density of 10.08 mounds ha^-1 appeared to be higher than earlier reports of Lee and Wood (1971), which indicated that large mounds of Macrotermitinae that supports large number of individuals are usually less than 10 mounds ha^-1. Similarly, Pomeroy (1978) reported a density of 1- 4 mounds ha^-1 for Macrotermes in Uganda, Trapnell et al. (1976) also, estimated a density of 3- 4 mounds ha^-1 in Zambia for Macrotermitinae and Abe et al. (2009) reported a density of 3-10 mounds ha^-1 for M. bellicosus in an upland site of central Nigeria and attributed the variation to anthropogenic disturbance. The observation in the present study may be because of the favourable environmental conditions and non-destruction of the mounds by people due to the spiritual believes that the mounds house evil spirits.

Figure 1 showed mean number of mounds of M. bellicosus in different land types. It revealed that Wurno upland had the highest mean, followed by Shagari upland while Wamakko upland was the least. Also, fadama lands have more mounds than the uplands in all the Local Government Areas except in Shagari Local Government Area where uplands had higher mounds than the fadama lands (Figure 1). The Figure also showed significant difference (p<0.05) between fadama lands and uplands in the number of mounds in all the Local Government Areas.

Table 1. Abundance of Mounds of M. bellicosus per hectare in the Study Area

The significant difference observed between the density of mounds in the fadama and upland areas, with the fadama having higher number of mounds in all the Local Government Areas except in Shagari Local Government Area, could be attributed to the moisture content in the fadama land type. But in Shagari, probably because of the pestiferous activities of the termites, there were various attempts to control the termites from the fadama as it is the major farming area in the community, this might have influenced the mound density. This is in agreement with the report by Pomeroy (1978) that the abundance of mounds of M. bellicosus was strongly correlated with moisture, but disagreed with Abe et al. (2009) which reported that the number and volume of termite mounds are lower in lowlands than the uplands and attributed interference with shallow water table in poorly drained soil as a possible factor responsible for the difference.

Figure 1. Mean Number of Mounds of M. bellicosus per hectare in Different Land Types in the Study Area. Bars followed by the same letter at the top were not significantly different (p> 0.05). LSD (SAS, 2003®)

Dispersion coefficient as presented in Table 2 ranges from 0.03 (Wamakko upland) - 0.71 (Wurno upland). This showed that the mounds were clumped in distribution in all the fadama lands and Wamakko upland, randomly distributed in Yabo upland and Shagari upland, while in Wurno upland the distribution indicated a degree of over-dispersion. The clumped distribution patterns observed among the fadama lands could be because of sufficient plant material in the land type, also the random distribution observed among the uplands implied that intraspecific competition was not a key factor regulating the distribution of mounds in those areas. This pattern implied that a few major factors are dominant in regulating the distribution of the mounds in the area, which could be one of the reasons for the unexplained observed variation in some areas in the study with apparently no difference in the environmental variables.

The observed over-dispersion in Wurno upland may be because of rocky nature of the area making it difficult for mound construction by the termite. The observed clump dispersion in Wamakko upland could have resulted because of the low density of the mounds in the most of the land type with few mounds aggregating in one area possibly because of its richness in plant materials due to improved soil quality. This was in line with Collins (1981) which reported a random distribution pattern of M. bellicosus in Southern Guinea Savanna of Nigeria and Eggleton et al. (1996) in Cameroun that termite abundance was highly clumped.

Table 2. Dispersion Coefficient of Mounds of M. bellicosus in the Local Government Areas

4. Conclusions

M. bellicosus have high mound density in Sokoto State. The insect being a pestiferous species could promote erosion; appropriate control strategies are required for its control. The clump distribution pattern of the termite suggests that they are aggregated in a more favourable environment, thus paving a way for their control.