The present study was undertaken to examine the population structure of the Solanum pimpinellifolium collection maintained by AVRDC – The World Vegetable Center – and to construct a core set of this collection. Out of the entire collection of 322 accessions, a diverse subset of 190 accessions was chosen representing 14 countries of origin. Data on 32 qualitative and 22 quantitative phenotypic traits (IPGRI–AVRDC descriptor traits) and 48 simple sequence repeat markers evenly distributed over the genome were used to develop the core set. A total of 377 alleles were detected with 7.85 alleles per locus, on average. Of these, 52 alleles at 28 loci were extremely rare-frequency alleles. The 190 accessions clustered into two main populations and an admixture group. Population I (PopI) included 99 accessions, 93 of which originated from Peru. Population II (PopII) contained 49 accessions, the majority of which originated from Ecuador and Mexico. The remaining 42 accessions were classified as admixture group. The two main populations were further subdivided into five subgroups. Values of Fst among the five sub-populations were significant (average pairwise Fst of 0.296), suggesting a real difference between these populations. A clear differentiation was observed among and within populations based on geography. Peruvian accessions were genetically more diverse than accessions originating in Ecuador and Mexico. Within the Peruvian group, a gradual increase in genetic diversity was observed from southern to northern Peru. The constructed core collection consists of 75 accessions representing 23.4% of AVRDC's entire S. pimpinellifolium collection and 39.5% of the subset used in this study. It is a well-balanced core with a good representation of the different populations (31 accessions from PopI, 22 from PopII and 22 from the Admixture group) and geographic origins (40 accessions from Peru, 17 from Ecuador, 14 from Mexico and four from other countries).

Based on the distribution, molecular similarity and use of Morinda citrifolia L., and occurrence of a wild Morinda species, Southeast Asia and Micronesia have been suggested to be the places where noni originated. The present article discusses the indices used by Vavilov and subsequent authorities on the origin of crop plants to argue that South Asia (Southeast India) has a greater probability of being the centre of domestication/origin for noni than Southeast Asia or Micronesia. The basic reasoning is that economically important plant cannot originate without richness in biodiversity and ingenuity of local people. India with rich floristic diversity, one of the centres of origin of crop plants with a natural distribution of Morinda species, including M. citrifolia L. and its immediate ancestors, has the oldest reference of occurrence, use and cultivation (Vedic literature); therefore, it appears to be the more probable centre of noni's origin. The ancient history of the expansion of Indian culture, religion and trade to Southeast Asian countries corroborate the possible role of Indians in the introduction of noni or knowledge regarding its value to Southeast Asia, from which it was carried to Micronesia and Polynesia, which provided a more favourable environmental niche for perpetuation and use.