African Journal of
Microbiology Research

  • Abbreviation: Afr. J. Microbiol. Res.
  • Language: English
  • ISSN: 1996-0808
  • DOI: 10.5897/AJMR
  • Start Year: 2007
  • Published Articles: 5215

Full Length Research Paper

Morphological and molecular characterisation of termite species in Taita Taveta County, Kenya

Paulyne Akoth
  • Paulyne Akoth
  • School of Biological Sciences, Jomo Kenyatta University of Agriculture and Technology, P. O. Box 62000-00200, Nairobi, Kenya.
  • Google Scholar
Huxley M. Makonde
  • Huxley M. Makonde
  • Department of Pure and Applied Sciences, Technical University of Mombasa, P. O. Box 90420-80100, Mombasa, Kenya.
  • Google Scholar
Nancy L. M. Budambula
  • Nancy L. M. Budambula
  • Department of Biological Sciences, University of Embu, P. O. Box, 60-0100, Embu, Kenya.
  • Google Scholar
Hamadi I. Boga
  • Hamadi I. Boga
  • School of Biological Sciences, Jomo Kenyatta University of Agriculture and Technology, P. O. Box 62000-00200, Nairobi, Kenya.
  • Google Scholar

  •  Received: 02 November 2021
  •  Accepted: 22 February 2022
  •  Published: 30 April 2022


The diversity of termites remains obscure despite having seized the equatorial ecological community with numerous effects both environmental and solvent. In the current study, morphology and DNA based techniques were used to appraise the heterogeneity of termites in Taita Taveta County. Soldier termites were arbitrarily fragmented from the five main plant lives in the study region. A dissecting microscope was used to discern the external features of the termites and was noted. The features were compared to the identification keys and then transformed into quasi-characters to generate a dendrogram. DNA was isolated from the 31 specimens from which mitochondrial cytochrome oxidase II (COX II) gene was sequenced and probed. We tested whether or not molecular characterisation underpins morphological characterization and then obtained similar sequences from the Genebank Repository of National Center for Biotechnology Information (NCBI) and used them together with our sequences for the phylogenetic tree construction. The results were supported by genetic specifications such as nucleotide composition and dichotomized genetic range. Termite of three different genera; Macrotermes, Amitermes and Odontotermes were identified based on their mound structure and soldier morphology. Phylogenetic analysis also placed the termites into three clusters, which were affiliated with the genera above. Genera Amitermes and Odontotermes were confined to distinct vegetation whereas the genus Macrotermes were all-pervasive. The study confirmed that both approaches identified the termite genera but a combination of both is necessary for confirmation and higher taxonomic classification.


Key words: COII, diversity, Macrotermitinae, termites, phylogeny.


Termites (Isoptera) are known both as ecological drivers and significant menace to the environment (Vidyashree et


al., 2018). They contribute to various ecosystem processes such as decomposition of organic matter into nutrients, and also modify soil properties decomposition of organic matter and nutrient cycling (Arif et al., 2019). Moreover, their termitaria provide special microhabitat for a number of organisms (Enagbonma and Babalola, 2019). However, they are also pests of agricultural crops and household items made of cellulosic material, a trait that tends to conceal their role in offering ecosystem services (Jouquet et al., 2018). Both positive and negative effects of termites arise from their feeding activities abetted by symbiotic microorganisms such as fungi and bacteria which facilitate cellulose digestion (Adaobi and Abiye, 2019).


Identification of termites still poses challenges due to inadequate taxonomic knowledge as well as their eusocial behavior coupled with distinct caste system (Murthy, 2020). The latter is because termites present limited or no species-specific morphological traits which allows identification only up to the genus level (Korb et al., 2019). In addition, there is a shortage of experts in most of the equatorial and subtropical countries (Mugerwa et al., 2014). As a result, there is scanty information on prevalence, identity, diversity and distribution of termites in Kenya and the African continent as a whole. Research leading to character level designation is thus necessary for accurate biodiversity assessment (Murthy et al., 2017).


Currently there are about 3106 sentient and remnant identified termite species (Effowe et al., 2021) allied to 330 genera, 12 families, 21 subfamilies and eleven families (Krishna et al., 2013). The order Isoptera is formally divided into inferior (underdeveloped) and superior (more advanced) termites (Zhu et al., 2012). The latter belong to the family Termitidae which have varying and sometimes conflicting characteristics. Members in the family Termitidae are capable of building termitaria albeit some just make galleries and tunnels underground. As a matter of fact, the family Termitidae contains the greatest number of studied genera and described species that is more than 2/3 in each case (Kanwal et al., 2011). The African continent leads in the number of species and this is due to its favourable climatic conditions (van Huis, 2017). There are over 650 species in the family Termitidae. The subfamilies under this family have grown to eight over the years (Engel, 2011). Majority of the described species belong to the eight families. East Africa alone has over 177 termite species and more could yet be identified (Ahmed et al., 2011). The lower termites which are majorly dependent on protozoa and bacteria (Huda et al., 2019) are; Rhinotermitidae, Kalotermitidae, Hodotermitidae, Termopsidae, Archotermopsidae, Stolotermitidae, Stylotermitidae, Serritermitidae and Mastotermitidae. The number of species in these families reduces respectively with the last two families having only one species each (Namanda, 2015; Nkunika, 1998).


External morphology of termites has for a long time been given importance in systematics and classification studies (Mandal et al., 2014). Taxonomic characters found in soldier caste present conspicuously reliable features for diversity investigations (Himmi et al., 2020). However, without keen conduction, morphological identification could present some ambiguity when solely applied. It should therefore be complemented with other methods (Patel and Jadhav, 2019). Many studies have opted to couple morphological identification with molecular characterization to avoid opaque identification (Schyra et al., 2019). The use of mitochondrial DNA (mtDNA) sequences has been successfully complemented with morphological characterisation for species identification. Mitochondrial COII gene in particular has been established to be reliable in species identification of members of different invertebrate groups as well as in establishing phylogenetic relationships among these groups (Garrick et al., 2015). To analyze mtDNA, little DNA samples is required and gives accurate results (Kapli et al., 2017); thus it remains a marker of choice for species identification. In addition, the COII gene has adequate representation of posted sequences in the GenBank database (Ghesini et al., 2020). This study aimed to determine termite colonies of the equatorial savannah of Taita Taveta County, Kenya based on the two approaches.



Study site and sample collection


Collection of termites was done in four main districts of Taita Taveta County. The county is situated northwesterly of Mombasa city in the coast and 360 km southeast of Nairobi. Its expanse stretches an area of about 17,000 km2 and is within a tropical savannah with various plant life systems. We highlighted Herbaceous Grassland (HG) [Mwatate district Latitude: -3° 29′ 59.99′′ S, Longitude: 38° 22′ 59.99′′ E]; Rain Forest (FS) [Wundanyi district Latitude: -3° 24′ 6.95′′ Longitude 38° 21′ 50.47′′]; Shrub Savanna (SV) [Voi district Latitude: -3° 23′ 45.78′′ S, Longitude: 38° 33′ 21.92′′ E]; Rain Fed Trees (RFT) [Taveta district Latitude: -3° 39′ 63.20′′ S, Longitude: 37° 67′ 36.20 E′′]; and Disturbed Areas (DL) all over of the county.


Aggregately, 54 active termite assemblages from the highlighted plant life systems were haphazardly chosen for sampling. Within each vegetation type we sampled mounds, which were at most 100 m apart. To obtain the exact points of reference of the collection points, Gamin software was installed in the phone and used to take the points (Table 1). Live termitaria were able to be distinguished by the absence of plant establishment as they cause developing soils around to harden and become impregnable (Yamashina, 2013). In order to get to the termite territory, the mounds had to be dug to a vertical extent of 50 cm downwards or depending on how far the nest was. 20 soldier termites were picked using pincers and preserved in sterile falcon tubes containing absolute ethanol. The collected samples were put in a cool box after which they were taken to the institution’s research laboratory. In the laboratory, they were kept in the refrigerator for subsequent morphological and molecular characterization.



Assessment of mound characteristics, morphological characterisation and analysis


Termite mound structure and features including height; size, presence and absence of ventilations, as well as shapes of the mounds were observed and noted. These features were collated with the identification features posted by Roonwal (1977). Morphological characterisation of soldier termites was conducted under a dissecting microscope. We measured the breath and linear distance of the soldier heads, entire linear distance, breath and linear distance of prothorax as well as the shape, and similarly for the mesothorax and metathorax. Apart from these characteristics, we also used; sensory features on the head by counting them and the parts thereof, head features including size and shape and maxillary characteristics. A combination of these characteristic features was employed in the generic level classification and identification of the termite samples. From every active mound sampled, only four termites were used to take measurements and features, and an average of the four was worked out. A high accurate vernier caliper (Mitutoyo 530-312) was used for taking measurement. We used the criterion published by Sornuwat et al. (2004) and Wijerathna and Dias (2012) respectively to ultimately identify the termites to the genus level. Further, Euclidean distance measure was employed and a hierarchical cluster analysis with average linkage method was performed. The impact brought about by scale was avoided by standardization of crude phonological information observed to zero mean and unit variance. The resulting covariance was employed in working out the interrelation amidst termite communities with cluster examination through hierarchical cluster analysis (HCA) with UPGMA (unweighted pair-group method using arithmetic averages). Paired range for all the sample (covariant equal advantage) were calculated. This was in a bid to establish those specimens with the smallest genetic range, which was done repeatedly for successive couples of specimens having trivial pair-wise range. These ranges are thus grouped repeatedly in a continuous manner. Proceeds from this continuous clustering were concocted to generate a rooted dendrogram with branching values using Euclidean diversity estimate in DARwin v.6.1 software. The dendrogram was generated using GenStat 16th Edition statistical software. Quality pictures of the exemplary termite galleries and termite samples were taken with an electric camera (OPPO A 93). A comparison of our findings with the records available in the National Museum of Kenya was done after which our representative specimens bearing the name and identity number of the researchers were conserved at the Entomology Department of the Kenyan archives.


DNA based delineation


Extraction of DNA


Slightly moderated Phenol:Chloroform method (Sambrook et al., 1989) of DNA extrication was used to extract complete DNA from the heads of soldier termites. The heads were sterilized, combined and crushed in 200 μl of TE (Tris EDTA pH 8) with mortar and pestle. For easy and faster disintegration of the tissues, we used 500 μl of lysis buffer (400 μl of TE and 100 μl of 5% SDS) and 10 μl of Proteinase K. The blend was inoculated at 65°C for 1 h. A mixture of 120 μl of phenol: chloroform: isoamyl alcohol (25: 24: 1) was put in the tubes followed by swirling for 30 s, and then centrifugation for 10 min at 10,000 rpm. The uppermost watery film was cautiously taken and put in a sanitized Eppendorf tube ensuring that the protein layer remained intact at the interphase. For precipitation of DNA, 500 μl of isopropanol was appended to the separated watery film and stored at -4°C for a whole night after which centrifugation at 12,000 rpm for 10 min was done. This led to the formation of a pellet which was cleaned severally by first getting rid of the supernatant followed by rinsing with 70% ethanol and sterile water. The resulting pellet was dried in air for half an hour or thereabout and eluted in 40 μl of TE buffer to stabilize it against disintegration. To confirm whether the gDNA was up to standard, we ran the extract in 0.8% agarose gel. The DNA was stored at -20°C for subsequent use (Sambrook et al., 1989).


Amplification and sequencing of COII gene


Using these set of primers; A?tLeu (5’?CAG ATA AGT GCA TTG GAT TT?3’) and B?tLys (5’?GTT TAA GAG ACC AGT ACT TG?3’), forward and reverse directions respectively (Miura et al., 1998), the extracted genomic DNA was used as a prototype to multiply the targeted COII gene portion. The mix therefore comprised of 2.5 μl of 10X PCR buffer, 2.0 μl MgCl2 (2.5 mM), 2.0 μl dNTPs (200 μM), 0.25 μl of Taq Polymerase (5 U/μl), 1 μl of each forward and reverse (5 Pico moles) and 0.5 μl of DNA. This was topped up with 15.75 μl of PCR water making a final volume of 25 μl. A reaction check with all the components except the prototype was also set for internal validity. Temperature cycles consisted of; primary denaturation at 95°C for five minutes, denaturation at 95°C for 30 s, annealing at 52°C for 45 s, and extension at 72°C for 1 min using automatic PCR unit. This was repeated 30 times, followed by 5 min of extension at 72°C. The resulting PCR by-products were authenticated by gel electrophoresis using 1.5% agarose in 1x TAE buffer stained with ethidium bromide and visualized under ultraviolet light. The by-products were refined using the QIAquick PCR purification Kit protocol (Qiagen, Germany). The refined by-products were preserved at - 20°C. This was then sent for sequencing through international courier services to Macrogen Netherlands, which is a commercial service provider ( The amplified DNA fragments were sequenced employing Sanger sequencing, the most common dideoxy method using COII universal primers, whereby there was use of redesigned dideoxy bases (ddNTP’s). The sequence data of the termites were recovered in the form of chromatograms which were extracted and saved for editing and further analyses.


Phylogenetic analyses


Out of the effectual 31 sequences amplified, only four were utilized in the eventual probe to maximize on the quality of the results. These sequences were manually edited in CLC main workbench version 7.2.6. It was done by first trimming the sequences to eliminate reads of poor quality, followed by assembling the forward and reverse sequences to align them where they overlap to get a contiguous sequence. Conflicts arising as a result of mismatch in overlapping regions of the forward and reverse sequences were assembled. Using Mallard software (Ashelford et al., 2006), we looked out for any chimerical structures and artefacts present. The edited sequences were submitted to the Genebank to obtain the following accession numbers; KT845956, KT845957, KT845958, KT845959, KT845960, KT845961, KT845962 and KT845963. Further, we did an exploration of identical sequence arrangement using BLASTN. Closest relatives of the sequences with 99-100% identity were retrieved from the GenBank for subsequent phylogenetic analysis. The sequences were aligned with the help of CLUSTAL Omega program ( The evolutionary history was hypothesized employing UPGMA formula (Sneath and Sokal, 1973) after which it was possible to highlight optimal tree with total branch length that was equal to 1.77017127. The rate percent of duplicate trees in which the related species were grouped simultaneously in the bootstrap test (1000 replicates), were presented above the nodes (Tamura and Kumar, 2002). The tree was drawn to scale, using similar units like the units of the branch length in evolutionary distances for phylogenetic tree inference. Using Maximum Composite Likelihood formula (Tamura et al., 2004), we determined the genetic range in the units of the number of base interchanges in each site. The differentiation in constitution partiality among sequences was given consideration in evolutionary differentiation (Tamura and Kumar, 2002). We incorporated 15 homolog sequences. Codon sites considered were 1st+2nd+3rd. For every set of sequences, all obscure sites were eliminated leaving 682 sites in the eventual data-set. MEGA 7 (Kumar et al., 2016) was used to do the evolutionary analysis. To construct the tree, we exported the aligned sequences to MEGA and converted them into Mega format. We then set the analysis parameters like bootstrapping values (in this case 1000 times) as well as appropriate substitutional model and left it to compute and finally saved the generated tree. All prototypical sequences were retained in a collective database (GenBank) for easy access by interested stakeholders.


Morphological identification of termites 


In this study it was observed that termite mounds were present in all the types of plant life surveyed. Collectively, 54 active termitaria were documented. Of the total number, 17 were located in the herbaceous grassland vegetation, seven in the rain-fed trees, another seven mounds in the shrub savannah vegetation and six mounds forested in the vegetation. The disturbed land recorded 17 active termite mounds. It was noted that in the disturbed areas such as the area occupied by the institutions, there were more active mounds compared to farmlands, which had mounds that were either destroyed or inactive. Despite encountering few mounds in the rain-fed trees and shrub savanna vegetation, occurrence of inactive mounds was uncommon. In the herbaceous grassland vegetation type (the most dominant vegetation in the County), the active mounds were dominant.


Mound architectural characteristics


A comprehensive analysis of the mound characteristics revealed the presence of three types of mounds. These were regular and irregular conoid mounds with several vents, elongated closed dome-shaped mounds and above ground mounds. Of these mounds, 32 were huge irregularly conoid with several vents. The elongated closed dome shaped mounds were 17, the slightly above ground mounds were three while the arboreal nests were two (Figure 1).



Termite identification


Termite samples identified from this study clustered into three groups based on the mound structures they construct and the observed termite characters noted, including the head capsule, antennae segments, mandibular features, pronotum, labrum, mesonotum and metanotum (Table 2).



Cluster analysis of morphological characters


There were five clusters distinguished as a result of examining the phenotypic characteristics of the termites when a dendrogram was generated (Cluster I, II, III, IV and V). Most samples (41 members) clustered together in Cluster I (Figure 2). The populations in this cluster were a mixture from the different vegetation types; DL (37%), RFT (15%), HG (24%), shrub savannah (20%) and forest (4%). Sample 12DL formed a cluster of its own in Cluster II. Cluster III comprised six members, namely 18RFT, 19FS, 26FS, 21HG, 32SV, and 38RFT. Furthermore, five samples (24SV, 25FS, 29FS, 51RFT, 28FS, and 27FS) formed Cluster IV whereas 27FS formed a singleton in Cluster V. Clustering of the termites based on morphological characters indicated that the soldier characters were not in any way dependent on vegetation type as some characters were shared by members of different vegetation types. Members sampled from the same plant life system also clustered separately, which is an indication of diversity among them. It also pointed to heterogeneity between members of the same vegetation and those of other vegetation types.



Molecular identification of termites


Nucleotide analysis, genetic distances and sequence divergence


The approximately 750 bp fragment of COII gene of the termite specimen amplified, suggests that there was no transfer of mitochondrial DNA to the nucleus which were sequenced. Stop codons were hardly present in majority of the samples. This is congruent with all amplified sequences depicting the nature of a functioning mtDNA cytochrome oxidase II arrangement of nucleotides. The parts omitted in some sequences were automatically fixed with the MEGA program applied before testing. The third codon set was weighted with A+T sequences at an average content of 63%. Figure 3 shows the nucleotide constitution of varied termite species.



We worked out the two-by-two genetic ranges between the sequences using Kimura two parameters (Kimura, 1980) presented in Table 2. Distance within species of the first set containing specimens 1DL, 3DL, 4DL, 21HG, 14SV. 34HG and 32SV haplotypes was 0.003 while that of group two comprising specimens 7RFT, 17RFT, 20RFT, 42RFT 41HG and 8SV was 0.004. Sequence of sample 28FS clustered together with Amitermes conformis. Interspecific distances among the four different clusters of the termites ranged from 0.04 between first and second clusters to 0.23 between third and fourth clusters (Table 3). Within the same plant life, there was low genetic divergence measure between species which ranged between 0.00 to 0.01 for the DL, 0.00 to 0.01 for RFT, 0.03 to 0.14 for SV samples, 0.22 for FS specimens and 0.00 to 0.04 for HG specimens. This same divergence was higher for species from disparate plant life systems. An example is the genetic range linking forest species and those from the RFT vegetation which fell between 0.19 and 0.21 (Table 3). COII Intercluster divergences were approximately 10-fold higher than intracluster divergences.



Phylogenetic analysis


UPGMA formula of evaluation used for inferring

the revolutionary account on MEGA 7.0.2 revealed three main clusters, representing three different genera (Macrotermes, Odontotermes and Amistermes). One cluster (supported by a bootstrap value of 84%) had three samples indicated as 1DL [KT845957], 17RFT [KT845959] and 21HG [KT845960] on the phylogenetic tree, plus a close association to representatives of genus Macrotermes (Figure 4). These members (including Macrotermes species, Macrotermes subhyalinus and Macrotermes michaelseni) had a close taxonomic affiliation with ≥98% sequence similarity amongst them. Another cluster was related to Odontotermes species. Amitermes representatives assembled separately in a group with a strong bootstrap rate of 80% together with sample 28FS [KT845963]. The two Amitermes spp. (Amitermes conformis and Amitermes obeuntis) had 91% sequence similarity with sample 28FS [KT845963].



Termite taxa dispensation in the different vegetation types of the study site


Only a single termite species (M. subhyalinus) subjugated the fragmented land areas or DL. In the same way, the RFT land type comprised singly of the M. michaelseni. The herbaceous grassland vegetation type was inhabited by both M. subhyalinus and M. michaelseni. Notably, the shrub savannah vegetation type was rich species composition with three termite species; M. subhyalinus, M. michaelseni and an Odontotermes sp. Shockingly, the forest that is expected to be species-rich recorded only two termite species of genera; Amitermes and Macrotermes. The two species were M. subhyalinus and A. conformis. Overall, shrub savanna vegetation type was the richest termite species diversity (M. subhyalinus, Odontotermes ceylonicus and M. michaelseni) with a record of two different termite genera (Microtermes and Odontotermes). DL and RFT land systems were the lowest in species diversity. It was noted that termites of the genus Macrotermes had a high frequency of occurrence in all the land systems in the study region while on the contrary, the other two genera Odontotermes and Amitermes were limited to SV and FS accordingly.



The mound structures variations observed, points to the existence of different termite species inhabiting the studied area. Although we did not classify the mounds on the basis of how different termite species feed, it is general knowledge that most soil feeders and fungus-growers are capable of constructing mounds (Makonde et al., 2015); thus suggesting their presence. We observed two types of mound structures; conoid mound with external openings being analogous to the species M. subhaylinus and closed dome shaped mounds being analogous to the species M. michaelseni. This was supported by the fact that we had only collected those species from the respective mounds. Studies by Vesala et al. (2017) and Ocko et al. (2017) had reported that a termite species of M. subhyalinus is capable of constructing vented termitaria and M. michaleseni to construct closed dome shaped mounds. Not long ago, a similar finding in the same region was reported on the mound structure of M. subhyalinus and M. jeanneli (Vesala et al., 2019). Another study reported a scenario where it was difficult to ascertain the true mound builder as termites of two different termite genera both inhabited the mound (Makonde et al., 2013). More recently, Paejaroen et al. (2021) found other species in the genus Globitermes and Microcerotermes to build large epigeal mounds in the tropics of Thailand which suggests more species could be involved in the ultimate morphology of a mound. Termite samples identified from the study by morphology were ultimately placed in three genera; Macrotermes, Odontotermes and Amitermes all of the family Termitidae. Coincidentally, the termite colonies that were identified to genus level through morphological characterization using soldiers were phylogenetically affiliated with the three genera indicating that morphological differences among genera are great enough for phylogenetic analysis of termites at the genus level. However, for species-level classification, this approach might take time and definitely need specialized knowledge of character differences (Janowiecki, 2015).


Molecular characterization results indicated that COII sequences guaranteed the ability to allocate the termites into molecularly well-defined species. There was accurate disparity of most sequences into different clusters. In comparison to their predetermined morphospecies, there was a high level of consistency. The nodes directly defining the clusters presented 99% nodal support. Furthermore, the three clusters remained definite to the fact that such groups comprised well-defined COII lineages instead of dissipated sequence dissimilarity (Hajibabaei et al., 2006). For our study, sequence separation in COX II mtDNA within distinct groups (intraspecific) were smaller than separation among groups (interspecific), although they prevailed within COII sequence divergences from Genbank. This portion of the mtDNA has a general characteristic of showing wide interspecies, yet little intraspecies separation, elucidating that species regularly create well distinct groups on phylogenetic tree. The empirical levels of sequence separation associated with species delimitation studies and in particular bar-coding (Hebert et al., 2003) support the finding of this study.


High Adenine Thiamine content of 63% in the studied sequences did not deviate from what mtDNA usually presents in insects (Crozier and Crozier, 1993) because it is a common characteristic of Cytochrome Oxidase II mtDNA region in arthropods. A+T partiality in such sequences is harmonious with information related to COII mitochondrial genes termite in the family Termitidae as well as some genera, that is, Coptotermes of the family Rhinotermitidae (Singla et al., 2016; Yeap et al., 2007). A report by Wahlberg et al. (2003) studying Phyciodes, family Nymphalidae, and order Lepidoptera also found A+T content from three varied gene portions in Nasuta subgroup of Drosophila in the range of 65-78%. In a different study by Hussin and Majid (2020), a lower value than that in our study was found (58.4%) and cited the use of incomplete sequences rather than complete COII gene region (which varies between insects in the range of 673 - 690 bp) as most probable reason for their findings. A study of genus Coptotermes in one of the Japanese forests (Tokuda et al., 2012) also reported A+T bias and correlated it to the genetic structures and arrangement of cytochrome oxidase gene sequences. This propounds that AT alteration coercion as a consequence of preselection in order to incorporate amino acids encrypted by AT loaded lines accompanies sequence separation. The phylogenesis research using mtDNA COX II gene fragments therefore depicted the similarity between different termite species. The level of dissimilarity among them is very important in understanding their relatedness. There have however been congruous phylogenetic reconstructions in termites with single or combined mitochondrial markers and nuclear markers having no overlaps of intra- and intergroup sequence variability which avoids ambiguous definition of clusters (Hausberger et al., 2011).


Our sequences from the DNA based characterisation are grouped into one major family Termitidae under the subfamilies Macrotermitinae and Termitinae and further to three main genera as in the morphological characterisation. From the genus Macrotermes, there was a clear separation into well supported two sub groups; M. subhylinus and M. michaelseni species. The grouping of these species was congruent with using conceptual approach for divergences occurring within the species and between species for termite delineation in Termitidae, going by the report of Osiemo et al. (2010). We did not however discover any obscure species as all the sequences clustering conjoined perfectly with morphospecies. The low sequence similarity between sample 28FS and Genebank samples may suggest a possibility of a new taxa, which still needs further detailed description using morphological, chemical and molecular properties. This is to say there was no sequences within the same morphospecies extensive branching nor clustering of distinct morphospecies. Other studies have analyzed morphospecies and found comparatively greater diversity than in our studies e. g (Cassalla and Korb, 2019), hence it is completely hard to compare them with our findings. Osiemo et al. (2010), in their research on termite diversity assessment in Kakamega Forest (found in Kenya), had more or less similar findings. Their study however found that there was deep sequence separation in the phonologically unrecognized as well as identified morphologically different species in termites feeding on soil. It is for this reason that these were considered cryptical species and therefore placed together separately as groups of organisms with similar gene characteristics. The forest in question, unlike our study area, is large and diverse and is within the tropics hence their observation. Other authors who have successfully uncovered ambiguous convergence in overactive divergent groups like termites and nematodes have been reported (Bourguigno et al., 2015). 


Morphological characterisation using soldier morphology of termites in the current study was possible only to the genus level whereby three genera were identified. Identification based on DNA approach was possible to the species level and termite samples into three genera and four different species. Notably, before construction of the phylogenetic tree, there was a specimen which was presumably placed as Nasutititermes species, because of the external features it conferred. However, after phylogenetic analysis, it became apparent that the sample had closer genetic association with A. conformis and very distinct from Nasutitermes genus and thus was classified as a species of the genus Amitermes. Termite characters sometimes develop in response to their immediate environment for the sake of adaptation and so would present as a certain species. This is what leads to such morphological ambivalence (Hausberger et al., 2011). Using both techniques, we were able to resolve the termite identification problem. Overall, molecular genetics have nearly endless possibilities for their applications to the study of termites.


Although many mounds were recorded in the herbaceous grassland vegetation, only one termite species from the genus Macrotemes was identified. The only possible explanation we could give for this observation was that the same colony was reestablished in different parts of the vegetation cover due to increase in their population. The many numbers of inactive mounds in the disturbed land was attributed to increased human activities which trigger deliberate destruction often in sites where agricultural activities were ongoing and also for fear that termites would cause damage. Notably, in sites such as the Taita Taveta University and Kenyatta High School (areas without farming practices), a few active mounds were recorded. Reduced species abundance in relation to land fragmentation is not a unique occurrence as it has been widely reported in Africa and other parts of the world (e.g. Ivory Coast: Coulibaly et al., 2016; Vietnam: Neoh et al., 2015; Panama: Basset et al., 2017). This supports the speculation that it is a worldwide trend. For instance, recording only a single termite species in the disturbed sites was a clear indication that agricultural activities were rampant hence termite migration and consequently inactive mounds. Egan et al. (2021) also reported similar findings. The shrub savanna vegetation type had the highest termite diversity, demonstrating high termite activities in the vegetation type. The possible explanation for this is that most members from the genera Macrotermes and Odontotermes are general wood and litter feeders, hence the high activities in such environment. Hypothetically, dryland savannah has lower diversity in comparison to wet savannah (Luke et al., 2014). Two termite species belonging to different genera were recorded in the forested vegetation. These were Amitermes and Macrotermes. The former, (though it can be found in a range of habitats) was only restricted to the forest. The explanation to this may be the forests or undisturbed areas support even the species that are or could be endangered as there are limited human disturbance. This however is dependent on whether the forest is protected or not as unprotected ones are subject to disturbance (Schyra et al., 2018). Moreover, it has been reported that Amitermes species has been reported to survive in cool and wet environments (Davies et al., 2014). In a West African study of termite species richness in protected and disturbed land of both savannah and forest ecosystems using COII gene, species richness decline was found in forest than in disturbed land whereas it was not the case for the savannah. However, collectively, the forest had more termite encounter rate as well as species number (Schyra et al., 2019). In both ecosystems, there was low composition of termites in the disturbed sites. Another study (Jamil et al., 2017) on the contrary, reported low species composition in the protected forests areas of Samusan Wildlife Sanctuary when compared to other forest studies elsewhere and they accredited their findings to intensive and extensive logging activities and loss of primary forest. In general, higher altitudes decrease heterogeneity of species of termites. This possibly is because at high altitudes there are reduced temperatures associated with rates of metabolism.



The results have shown that mound building termites of the three genera reported above have dominance over the studied regions of Taita Taveta county. The study demonstrates that agricultural activities affect termite activities and consequently the species abundances. Further, the study confirms that termite species identification based on polyphasic approach is more concise and informative than using the morphological characters alone. The knowledge gathered from this study using the dataset will be essential in other studies aiming to identify Kenyan species of termites in terms of ubiquity as well as their organization. Such identifications have been proven to be hard when done using only the morphology based approach. Thus, the study underlines the value of employing the two methods; morphology based and DNA-based characterization techniques. These contributions provide a strong foundation for future work that is needed to better understand this important economic pest.


The authors have not declared any conflict of interests.


Adaobi P, Abiye TD (2019). Studies on termite infestation of buildings in Ase, a rural community in the Niger Delta of Nigeria. The Journal of Basic and Applied Zoology 80(27):1-7.


Ahmed BM, Nkunika POY, Sileshi GW, French JRJ, Nyeko P, Jain S (2011). Potential Impact of Climate Change on Termite Distribution in Africa. British Journal of Environment and Climate Change 1(4):172-189.


Arif A, Muin M, Larekeng SH, Lestari PI (2019). Survey and morphological identification of termites (Insecta: Isoptera) in Teaching Forest of Hasanuddin University, Indonesia. Earth and Environmental Science 270:012001.


Ashelford KE, Chuzhanova NA, Fry JC, Jones AJ, Weightman AJ (2006). New screening software shows that most recent large 16S rRNA gene clone libraries contain chimeras. Applied and Environmental Microbiology 72(9):5734-5741.


Basset Y, Barrios H, Ramirez JA, Lopez Y, Coronado J, Perez F, Arizala S, Badilla RM, Leponce M (2017). Contrasting the distribution of butterflies and termites in plantations and tropical forests. Biodiversity Conservation 26:151-176.


Bourguignon T, Lo N, Cameron SL, Šobotník J, Hayashi Y, Shigenobu S, Watanabe D, Roisin Y, Miura T, Evans TA (2015). The evolutionary history of termites as inferred from 66 mitochondrial genomes. Molecular Biology of Evolution 32:406-421.


Casalla R, Korb J (2019). Termite diversity in Neotropical dry forests of Colombia and the potential role of rainfall in structuring termite diversity. Biotropica 1-13.


Coulibaly T, Akpesse AAM, Boga JP, Yapi A, Kouassi KP, Roisin Y (2016). Change in termite communities along a chronosequence of mango tree orchards in the north of Côte d'Ivoire. Journal of Insect Conservation 20:1011-1019.


Crozier R, Crozier Y (1993). The mitochondrial genome of the honeybee Apis mellifera: Complete sequence and genome organization. Genetics 133(1):97-117.


Davies AB, Levick SR, Asher GP, Robertson MP, Van Rensburg BJ, Pair CL (2014). Spatial variability and Abiotic determinants of termite mounds throughout a savanna catchment. Journal of Applied Ecology 49:422-430.


Effowe TQ, Boris DK, Abdoulaye BN, Bassan BS, Komina A, Isabella AG (2021). Temites Diversity in a protected park of the northern Sudanian Savanna of Togo. Nature conservation 43:79-91.


Egan B, Zwannda N, Barbara A (2021). Overview of the Genetic Diversity of African Macrotermes Termitidae: Macrotermitinae and Implications for Taxonomy, Ecology and Food Science. Insects 12(518):1-16.


Enagbonma BJ, Babalola OO (2019). Environmental Sustainability: A Review of Termite Mound Soil Material and its Bacteria. Sustainability 11(3847):1-10.


Engel MS (2011). Family group names for termites (Isoptera), Redux. pp. 171-184 in: Engel M.S (Eds) contributions celebrating Kumar Krishna zookeys 148.


Garrick RC, Collins BD, Yi RN, Dyer RJ, Hyseni C (2015). Identification of Eastern United States Reticulitermes Termite Species via PCR-RFLP Assessed using Training and Test Data. Insects 6:524-537.


Ghesini S, Muller G, Mario M (2020). First record of the subterranean termite Reticulitermes grassei in Switzerland. Bulletin of Insectology 73(1):149-151.


Hajibabaei M, Janzen DH, Burn JM, Hallwachs W, Hebert PD (2006). DNA barcodes distinguish species of tropical Lepidoptera. Proceedings of the National Academy of Sciences of the United States of America 103(4):968-971.


Hausberger B, Kimpel D, Van Neer A, Korb J (2011). Uncovering cryptic species diversity of a termite community in a West African savanna. Molecular Phylogenetics and Evolution 61(3):964-969.


Hebert PD, Ratnasingham S, de Waard JR (2003). Barcoding animal life: cytochrome oxidase subunit 1 divergences among closely related species. Proceedings of the Royal Society of London B: Biological Sciences 270(Suppl 1):S96-S99.


Himmi SK, Wikantyoso B, Fajar A, Tamardi D, Yusuf S (2020). Functional structures of termite assemblage in changing habitats and ecosystems. Series: Earth and Environmental Science 591:1-7.


Huda RKA, Alaa R, Batt MA (2019). The Differences in the Digestive System and Enzymes between Soldiers and Workers of the Subterranean Termite, Psammotermes hypostoma Desneux (Rhinotermitidae:Isoptera). Egyptian Academic Journal of Biological Sciences and Entomology 12(3):49-61.


Hussin NA, Majid AHA (2020). Genetic diversity and phylogenetic relationship of higher termite Globitermes sulphureus (Haviland) (Blattodea: Termitidae). Sociobiology 68(1):1-8.


Jamil N, Nurainie W, Wan Ismail W, Siti SA, Mazdan AA, Ratnawati H (2017). A preliminary survey of species composition of termites (Insecta: Isoptera) in Samunsam Wildlife Sanctuary, Sarawak. Tropical Life Sciences Research 28(2):201-213.


Janowiecki MA (2015). Genetic Variation and Species Distribution of Subterranean Termites in the Southeastern U.S. Focusing on Family Structure and the Invasive Formosan Subterranean Termite. MSC Thesis, University of Arkansas, p. 4.


Jouquet P, Ekta C, Amritha AVK (2018). Sustainable use of termite activity in agro-ecosystems with reference to earthworms. A review. Agronomy for Sustainable Development 38(3):1-11.


Kanwal HK, Acharya K, Ramesh G, Reddy M (2011). Molecular characterization of Morchella species from the Western Himalayan region of India. Current Microbiology 62(4):1245-1252.


Kapli P, Lutteropp S, Zhang J, Kobert K, Pavlidis P, Stamatakis A, Flouri T (2017). Mult-rate Poisson tree processes for single-locus species delimitation under maximum likelihood and markov chain Monte Carlo. Bioinformatics 33:1630-1638.


Korb J, Boris DK, Yvonne TC, Robin HCD, Jean NKBG, Mayouré EI, Guy J, N'golo A K, KM, Abdoulaye BN, SIO, Michael P, Yves R, Fernand S (2019). Termite Taxonomy, Challenges and Prospects: West Africa, A Case Example. Insects 10(32):1-8.


Kimura M (1980). A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16(2):111-120.


Krishna K, Grimaldi DA, Krishna V, Engel MS (2013). Treatise on the Isoptera of the World. Bulletin of the American Museum of Natural History 377(7):1-200.


Kumar S, Stecher G, Tamura K (2016). Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Molecular Evolutionary Biology 33(7):1870-1874.


Luke SH, Fayle TM, Eggleton P, Turner EC, Davies RG (2014). Functional Structure of Ant and Termite Assemblages in Old growth Forest, Logged Forest and Oil Palm Plantation in Malaysian Borneo. Biodiversity Conservation 23:2817-2832.


Makonde HM, Mwirichia R, Osiemo Z, Boga HI, Klenk HP (2015). 454 Pyrosequencing-based assessments of bacterial diversity and community structure in termite guts, mounds and surrounding soils. Springer Plus 4(471).


Makonde HM, Boga HI, Osiemo Z, Mwirichia R, Stielow JB, Göker M, Klenk HP (2013). Diversity of Termitomyces associated with fungus-farming termites assessed by cultural and culture-independent methods. PLoS ONE 8(2):e56464.


Mandal S, Chhakchhuak L, Gurusubramanian G, Kumar NS (2014). Mitochondrial Markers for Identification and Phylogenetic Studies in Insects, Review. Open Access DNA Barcodes 2:1-9.


Miura T, Maekawa K, Kitade O, Abe T, Matsumoto T (1998). Phylogenetic relationships among subfamilies in higher termites (Isoptera: Termitidae) based on mitochondrial COII gene Sequences. Annals of Entomological Society of America 91:515-523.


Mugerwa S, Mpairwe D, Zziwa E, Sawaans K, Peden D (2014). Integrated Termite Management for Improved Rainwater management. A synthesis of selected African Experiences. NBDC Technical Report-9.


Murthy KS (2020). Molecular identity of subterranean termites. International Journal of Applied Sciences and Biotechnology 8(4):410-416.


Murthy KSS, Banu Yl, Ramakrishna P (2017). Diversity of Subterranean Termites in South India Based on COI Gene. Journal of Biodiversity and Bio Prospective Development 4:161.


Namanda JM (2015). Food value and consumer preference of termites (isoptera: termitidae: pseudocanthotermes grandiceps) in western kenya: comparison with beef and fish. Master of Science Thesis, Eldoret University, p. 11.


Neoh KB, Bong LJ, Nguyen MT, Nguyen VT, Nguyen HQ, Itoh M, Kozan O, Yoshimura T (2015). Termite diversity and complexity in Vietnamese agroecosystems along a gradient of increasing disturbance. Journal of Insect Conservation 19:1129-1139.


Nkunika POY (1998). The biology and ecology of the damp wood termite, Porotermes adamsoni (Frogatt) (Isoptera: Termopsidae) in South Australia. PhD Thesis, University of Adelade, South Australia.


Ocko SA, King H, Andreen D, Bardunias P, Turner JS, Soar R, Mahadevan L (2017). Solar-powered ventilation of African termite mounds. Journal of Experimental Biology 220(18):3260-3269.


Osiemo Z, Marten A, Kaib M, Gitonga L, Boga H (2010). Open Relationships in The Castles of Clay. High Diversity and Low Specificity of Termitomyces Fungi Associated with Fungus Farming Termites in Africa. Journal of the International Union for the Study of Social Insects 57:351-363.


Patel P, Jadhav BL (2019). Genetic diversity and phylogeny of subterranean termites from Maharashtra using mitochondrial COI gene. Research Journal of life sciences, Bioinformatics, Pharmaceutical and Chemical Sciences 5(1):869-883.


Paejaroen P, Chetsada P, Supatra C, Weerachon S (2021). Morphological and Physicochemical properties of termite Mounds in the mixed Deciduous Forest of Thailand. Floresta e Ambiente 28(3):1-10.


Roonwal L (1977). Field Data on Intraspecific Variability in Mound Construction and Nesting Habits in Termites and its Ecological Relationships. Proceedings of the Indian National Science Academy 43(5):159-174.


Sambrook J, Fritsch EF, Maniatis T (1989). Molecular cloning: a laboratory manual (2nd Ed). Cold Spring Harbor, NY. Cold Spring Harbor Laboratory Press, Selection of the Formosan subterranean termite. Journal of Behavioral Ecology 19(4):764-773.


Schyra J, Gbenyedji JNBK, Meusemann K, Korb J (2019). Differences between termite assemblages in a West African savannah and forest ecosystem. PLoS ONE 2019. In revision.


Schyra J, Hausberger B, Korb J (2018). Phylogenetic Community Structure of Southern African Termites (Isoptera). Sociobiology 65(1):15-23.


Singla M, Goyal N, Sharma R, Singla N (2016). Reconstructing phylogenetic relationship among Indian termite species inferred from COII gene sequences (Blattodea: Isoptera : Termitidae ). Journal of Entomology and Zoology Studies 4:1-7.


Sneath PH, Sokal RR (1973). Numerical Taxonomy: The Principles and Practice of Numerical Classification (1st Ed). Freeman, San Francisco.


Sornuwat Y, Vongkaluong C, Tekamatsu Y (2004). A systematic key to termites of Thailand. Kasetsart Journal of Natural Sciences 38:349-368.


Tamura K, Kumar S (2002). Evolutionary distance estimation under heterogeneous substitution pattern among lineages. Molecular Biology and Evolution 19:1727-1736.


Tamura K, Nei M, Kumar S (2004). Prospects for inferring very large phylogenies by using the neighbor-joining method. Proceedings of the National Academy.of Sciences. U.S.A. 101:11030-11035.


Tokuda G, Isagawa H, Sugio K (2012). The complete mitogenome of the Formosan termite, Coptotermes formosanus Shiraki. Journal of the International Union for the study of Social Insects 59(1):17-24.


van Huis A (2017). Cultural significance of termites in sub-Saharan Africa. Journal of Ethnobiology Ethnomedicine 13(8):1-12.


Vesala R, Niskanen T, Liimatainen K, Boga H, Pellikka P, Rikkinen J (2017). Diversity of fungus-growing termites (Macrotermes) and their fungal symbionts (Termitomyces) in the semiarid Tsavo Ecosystem, Kenya. Biotropica 49(3):402-412.


Vesala R, Harjuntausta A, Hakkarainen A, Rönnholm P, Pellikka P, Rikkinen J (2019). Termite mound architecture regulates nest temperature and correlates with species identities of symbiotic fungi. Peer-Reviewed Journal 6:e6237.


Vidyashree AS, Kalleshwaraswamy CM, Mahadeva Swamy HM, Asokan R, Adarsha SK (2018). Morphological, molecular identification and phylogenetic analysis of termites from Western Ghats of Karnataka, India. Journal of Asian-Pacific Entomology 21:140-149.


Wahlberg N, Weingartner E, Nylin S (2003). Towards a better understanding of the higher systematics of Nymphalidae (Lepidoptera: papilionodae). Molecular phylogenetics and Evolution 28(3):473-485.


Wijerathna KHCA, Dias RKS (2012). A preliminary inventory of subterranean termites in the premises of faculty of science, University of Kelaniya and the potential of a ponerine ant species, Neemazal-F and Citronella oil in in the control of two termite species. Journal of science, University of Kelaniya 7:1-20.


Yamashina C (2013). Variation in savanna vegetation on termite mounds in north-eastern Namibia. Journal of Tropical Ecology 29:559-562.


Yeap Y, Othman S, Lee S, Lee Y (2007). Genetic relationship between Coptotermes gestroi and Coptotermes vastator (Isoptera: Rhinotermitidae). Journal of Economic Entomology 100(2):467-474.


Zhu Y, Jian L, Huhu L, Hui Y, Sheng X, Fei Z, Xuejia Z, Yun T, Xiangyang L (2012). Phylogenetic analysis of the gut bacterial microflora of the fungus-growing termite Macrotermes barneyi. African Journal of Microbiology 69:2071-2078.