TERMITE SAMPLES

Eight populations of R. dichrous and 12 populations of R. guangzhouensis were collected from 5 localities of Zhaoqing (23.1655833–23.1737778°N, 112.5394917°–112.5573944°E), Guangdong Province and 4 localities of Huizhou (23.2487639°–23.3840917°N, 114.0220361°–114.4395194°E), Guangdong Province, China, from 2011 to 2015 and were used in the morphological examinations. Some of the samples were chosen for molecular analyses according to the availability of worker termites and the success of DNA sequencing. A population of R. tricholabralis Ping and Li (Isoptera: Rhinotermitidae) collected from Wuhan (30.4795806°N, 114.3495361°E), Hubei Province, in Aug 2013 also was included in the molecular analysis. The specimens were preserved in 70% ethanol for morphological studies and in 100% ethanol for molecular analysis. Voucher specimens were deposited in Guangdong Institute of Applied Biological Resources, Guangzhou, China.

Table 1.

Termite species and populations included in the present study.

MORPHOLOGICAL EXAMINATIONS

Specimens were identified applying keys by Huang et al. (2000). The head and the pronotum of both species were photographed under a stereomicroscope (SMZ1000, Nikon, Chiyoda-ku, Japan) connected to a computer-assisted imaging camera. Forty soldiers from the 8 populations of R. dichrous and from the 12 populations of R. guangzhouensis were examined with the stereomicroscope. The following 11 morphological characteristics were measured: head length to upper base of condyle of mandibles (HL), head maximum width (HW), left mandible length (upper base of condyle to tip, LML), labrum length (LL), labrum maximum width (LW), postmentum medium length (POL), postmentum maximum width (POW), postmentum waist width (PWW), pronotum maximum length (PRL), pronotum maximum width (PRW), and hind tibia length (HTL). These measurement data were subjected to analysis of variance (ANOVA), and the significance of the differences was compared by the Duncan Multiple Range Test at the level of 5% probability using SAS^® software Version 9.0 (SAS 2002). The cluster dendrogram was constructed by using the minimum-variance method of Ward.

Fig. 1.

Photomicrographs of dorsal (A, D, E, F, I, and J), ventral (B and G), and lateral (C and H) views of Reticulitermes dichrous (A to E) and Reticulitermes guangzhouensis (F to J) soldier heads and pronotums.

DNA EXTRACTION, AMPLIFICATION, AND SEQUENCING

Total genomic DNA of 3 worker termites from populations of R. guangzhouensis and R. dichrous was extracted (Table 1). To avoid contamination with gut contents, including polysaccharides, which act as PCR inhibitors (Schrader et al. 2012), the abdomen was removed from specimens before extraction, and just the head and thorax were extracted. The DNA extraction was performed by TIANamp Genomic DNA Kit (TIANGEN, Beijing, China) according to the manufacturer's instructions. Two mitochondrial genes, COII and 16S rRNA, were amplified by polymerase chain reactions (PCRs) with primers TL2J3037 (alias AtLeu) (5′-ATGGCAGATTAGTGCAATGG-3′) (Liu & Beckenbach 1992) and TKN3785 (alias BtLys) (5′-GTTTAAGAGACCAGTACTTG-3′) (Simon et al. 1994) for COII, and with LR-J-13007 (5′-TTACGCTGTTATCCCTAA-3′) (Kambhampati & Smith 1995) and LR-N-13398 (5′-CGCCTGTTTATCAAAAACAT-3′) (Simon et al. 1994) for 16S rRNA. PCR amplification was performed in a 50 µL reaction mixture containing 2 µL DNA, 21 µL dH₂O, 25 µL Premix TaqTM (Ex Taq version, TaKaRa, Tokyo, Japan), 1 µL forward primer at 10 µM, and 1 µL reverse primer at 10 µM. Thermal cycling was conducted for 40 of the following cycles: predenaturation at 94 °C for 4 min, denaturation at 94 °C for 30 s, annealing at 54 °C for 30 s, extension at 72 °C for 40 s; followed by a final extension of 72 °C for 5 min after the cycles. Amplified products were checked on a 1% agarose gel, and sent to Generay Biotech Co., Ltd (Shanghai, China) for direct sequencing in both directions. GenBank accession numbers for termite DNA sequenced in this study are provided in Table 1, along with the accession numbers of some other species and populations.

MOLECULAR PHYLOGENETIC ANALYSIS

Molecular phylogenetic analyses were carried out on a dataset comprising COII gene sequences of 4 populations of R. dichrous and 3 populations of R. guangzhouensis, as well as 16S rRNA gene sequences of 4 populations of both species. DNA sequences of R. tricholabralis from this study and 10 other Reticulitermes species obtained from GenBank were included in the phylogenetic comparisons. Two species of Coptotermes, C. formosanus Shiraki (Isoptera: Rhinotermitidae), and C. gestroi were used as outgroup taxa.

Sequences were aligned with Clustal X (Larkin et al. 2007). Base compositional analyses were conducted with the computer program MEGA version 6.0 (Tamura et al. 2013). The distance estimation of MEGA 6.0 was used to calculate genetic distances according to the p-distance model. Molecular phylogenetic trees were constructed by the maximum likelihood method in MEGA 6.0. Consensus trees were determined using CONSENSE in the PHYLIP 3.6 software package (University of Washington, Seattle, Washington) with the majority rule (extended) model. Poisson model and 1,000 bootstrap repetitions were applied for the analyses.

Table 2.

Morphometric data (mm) of the soldier termites in different populations of Reticulitermes guangzhouensis and Reticulitermes dichrous.

Fig. 2.

Cluster dendrogram of 12 populations of Reticulitermes guangzhouensis and 8 populations of Reticulitermes dichrous.

MORPHOLOGICAL CHARACTERISTICS

Observations of R. dichrous and R. guangzhouensis soldiers from different populations indicated that the 2 species were similar in morphology. The head capsule of both species was elongated, broadly rounded at posterior margin, with a prominent hump at the front. In both species, the labrum was sharpened at the apex, the mandibles were stout with the tips slightly to moderately curved, the pronotum was covered by approximately 30 setae, and the anterior margin was concave with a shallow incised notch at the middle (Fig. 1). The morphological differences between the species mainly showed in the color of the head capsule and in the number of antennal segments. The head capsule color of R. dichrous was light brown, a little darker than that of R. guangzhouensis (Fig. 1A, F); the number of antennal segments of R. dichrous and R. guangzhouensis was 15 and 14 to 16, respectively. Furthermore, the maximum and minimum width of R. dichrous postmentum were generally narrower than those of R. guangzhouensis (Fig. 1B, G; Table 2), but the ratio of PWW to POW was similar in the 2 species. From the measurement data, large variations in significant differences in many characteristics were found among conspecific populations of both R. dichrous and R. guangzhouensi, but the differences were not significant between some populations of different species (Table 2). The cluster dendrogram constructed with the measurement data of the 2 species also showed that some populations of different species, instead of absolutely conspecific populations, clustered into 1 branch (Fig. 2).

MOLECULAR PHYLOGENETIC ANALYSIS

The average sizes of PCR products of COII and 16S rRNA genes for the 2 species examined were 763 and 432 base pairs (bp), respectively. For the COII gene, the multiple sequence alignment (including outgroups) resulted in 687 characters, of which 482 were constant and 148 were parsimony-informative. For the 16S rRNA gene, the multiple sequence alignment (including outgroups) resulted in 432 characters, of which 356 were constant and 56 were parsimony-informative. Average nucleotide composition among ingroup taxa was as follows: A = 39.3%, T = 23.5%, G = 13.9%, and C = 23.3% in the COII gene; A = 23.1%, T = 40.8%, G = 22.7%, and C = 13.4% in the 16S rRNA gene. Both mitochondrial genes were A and T rich.

The intraspecific genetic variations of R. dichrous ranged from 0.15 to 0.31% in the COII gene, with no variation in the 16S rRNA gene. The intraspecific genetic variations of R. guangzhouensis ranged from 0.00 to 2.31% in the COII gene, with no variation in the 16S rRNA gene. The genetic divergences between the 2 species ranged from 0.00 to 2.62% for COII and no variation was observed in the 16S rRNA gene. This was slightly higher than, or equal to, the intraspecific genetic variations of R. guangzhouensis. By comparison, the ranges between R. dichrous and R. guangzhouensis, and other analyzed Reticulitermes species were 5.08 to 8.94% in the COII gene and 2.24 to 4.47% in the 16S rRNA gene, which was higher than between R. dichrous and R. guangzhouensis.

Fig. 3.

Maximum likelihood phylogenetic tree of Reticulitermes guangzhouensis and Reticulitermes dichrous relative to other Reticulitermes species based on the COII gene. Bootstrap values are indicated at the nodes and represent 1,000 pseudoreplicates.

The molecular phylogenetic trees generated from the COII and 16S rRNA genes by the maximum likelihood method revealed similar phylogenetic relationships between R. dichrous and R. guangzhouensis (Figs. 3 and 4). In the tree inferred from the COII gene, all the populations of R. dichrous and R. guangzhouensis formed a clade with strong bootstrap support (99%). The clade comprised 2 subclades: 1 with R. dichrous and 2 populations of R. guangzhouensis from Xiangtou Mountain, and the other was R. guangzhouensis from Luofu Mountain. In the tree inferred from the 16S rRNA gene, populations of R. dichrous and R. guangzhouensis from this study and a population of R. guangzhouensis from GenBank also formed a common clade with a bootstrap value of 99%.