**4.2 Designing the experiment**

Appropriate sampling strategy requires samples to be taken from populations scattered in the entire research range, and for each population, sufficient individuals should be collected to represent it (S.Y. Chen & Y.P. Zhang, 2006). Based on this, seven populations representing *M. alternatus* from central, southern, southwestern, and northern Yunnan were selected, in which three populations were chosen from the southwestern Yunnan where the pine wood nematodes were reported. In an attempt to determine the source of the pine wood nematode, two reference populations from outside Yunnan were set, with one from Hubei Province representing central China and the other from Zhejiang Province representing eastern China (Fu et al., 2010; Fig. 2). Adult *M. alternatus* were collected by flight traps baited with barkborer lure (Chinese Academy of Forestry, Zhejiang, China).

Choosing the correct genetic marker can minimize problems in data analysis (Simon et al., 1994) and maximize the future potential of cross references with data published by other researchers. Population phylogeny requires gene markers with moderate evolutionary rate, free of recombination, which means that mtDNA, a maternal heritage genome without recombination, is the ideal choice. Considering that there has been no report on *Numts* in *M. alternatus*, which can also be detected and excluded by strict data examination when encountered. The prefunding research by Kawai et al. (2006) resolved a clear population phylogeny using a gene fragment of *cox2*; the research discussed here chose the same *cox2* marker.

Detecting Non-Local Japanese Pine Sawyers

southwestern Yunnan can be observed (Fig. 3).

Ruili 0 Wanding 0 1 Huaning 1 0 0 Stone Forest 2 0 0 2 Lianghe 2 1 1 1 2 Yongsheng 1 0 0 1 1 1 Zhejiang 0 1 1 0 0 1 0 Hubei 0 0 0 0 0 0 0 0

**5.2 Genetic distance** 

Pu'er

in Yunnan, Southwestern China via Modern Molecular Techniques 75

haplotypes being presented only in two of the populations from southwestern Yunnan (Table 1). When mapping haplotype frequencies into pie charts, a more obvious population boundary as well as the existence of non-local individuals in three populations from

**Population PE RL WD HN SF LH YS ZJ HB** 

**Unique haplotypes** 0 3 3 0 0 0 0 3 4 Table 1. Matrix of shared haplotypes of the nine *M. alternatus* populations (below diagonal) and the numbers of unique haplotypes (last row). Population codes correspond to those in Fig. 2.

Fig. 3. Mapping of haplotype frequencies of all nine populations of *M. alternatus*.

The Kimura two-parameter (K2P) distance varied from 0.0014 to 0.0132, with the distance between Pu'er (southern Yunnan) and Yongsheng (northern Yunnan) being the minimum and that between Stone Forest (central Yunnan) and Hubei being the maximum (Table 2). The K2P distances among populations from Pu'er, Huaning, Stone Forest, and Yongsheng and among populations from Ruili, Wanding, and Lianghe showed close genetic connection; but the K2P distances between any given population from Pu'er, Huanian, Stone Forest, or Yongsheng and that from Ruili, Wanding, or Lianghe suggested otherwise (Table 2). Hence,

The same PCR primers developed by Roehrdanz (1993) and Kawai et al. (2006) were applied, and the PCR reaction was performed by the protocols described by Kawai et al. (2006). Bioinformatic and statistic software like DAMBE 5.0.7 (Xia & Xie, 2001), MEGA 4.0 (Tamura et al., 2007), SAMOVA 1.0 (Dupanloup et al., 2002), AMOVA 3.1 (Excoffier et al., 2005), and SSPS 13.0 (SPSS Inc., Illinois, US) were applied to analyze haplotype assemblage, Kimura two-parameter (K2P) distance (Kimura, 1980), NJ phylogenetic reconstruction (Saitou & Nei, 1987), population grouping, and multidimensional scaling (MDS) (Lessa, 1990). The mapping of haplotype distribution was performed in AcrView 3.3 (ESRI, USA).

Fig. 2. The sampling sites of *M. alternatus* in Yunnan, Hubei, and Zhejiang. Dots on the map of China represent the capital cities of the three provinces. After Fu et al. (2010).
