Morphological characters, sampling, and DNA extraction
The three subspecies of Avicennia marina are marina, eucalyptifolia , and australasica . The subspecies marinais widely distributed from eastern Africa, through the Middle East, South Asia, Southeast Asia, and north to South China (Figure 1). It is also found in western Australia. The subspecies eucalyptifolia is mainly distributed in northern Australia and extends to southern Philippines, western Indonesia, and the Southwestern Pacific islands. There is a significant range overlap of the two subspecies in western Australia. Australasica is restricted to south-eastern Australia and northern New Zealand (Figure 1). Australasica can be morphologically distinguished from the other two by its fully pubescent calyx lobes and bracts (Duke, 1991, 2006). These structures are more glabrous in the other two subspecies. The bark of australasica is grey fissured, with short longitudinal fissures or reticulate lines, while the bark of the other two subspecies is smooth green or chalky white with flaky patches. Eucalyptifolia is mainly distinguished by its lanceolate leaves (as opposed to ovate to elliptic), as well as the style in open flowers which are positioned level with upper edges of anthers (instead of the lower edges of anthers) (Duke, 1991, 2006).Marina may also be distinguished by its larger flowers and thicker leaves. However, these distinctions in morphological characters may be inconclusive where two putative subspecies coexist (Duke, 2006). Typical for mangrove trees, propagules of A. marina are bouyant on sea water and disperse over sea to nearby locations with mangorve habitats (Duke, 2006).
We sampled 16 populations, 577 individuals (16 to 100 individuals per population) from East Africa, South China, Southeast Asia, Australia to New Zealand, covering A. marina ’s range (Table 1, Figure 1). To avoid sampling offspring from the same tree, sampled individuals were at least five meters apart. At each site, we sampled as many individuals as were available, but no more than 100. Leaves from each individual were dried, labeled, and stored for DNA extraction. DNA was extracted using the modified CTAB method (Doyle & Doyle, 1987). DNA content of each extraction was measured by NanoDrop 2000. For each population, we pooled 300 ng of DNA from each plant to make one DNA pool, ensuring that it contains the same proportion of DNA from each individual. Sixteen DNA pools were used in our experiments.