The U.S. germplasm collection for peanut, Arachis hypogaea L., consists of 7432 accessions and contains a great amount of genetic diversity. Information on economically important traits does not exist for most accessions due to the time and labor required for evaluation. The development of a core collection for peanut would provide a subset of accessions that are representative of the entire collection and that could be extensively examined. The objective of this research was to select a core collection for peanut. Data for the U.S. peanut germplasm collection were obtained from the Germplasm Resources Information Network (GRIN). The entire germplasm collection was then stratified by country of origin and by the amount of available morphological data. When information was available for at least four of the same morphological variables for at least 16 accessions from the same country of origin, then the data for these accessions were analyzed using multivariate statistical analysis. Results allowed the accessions to be clustered into groups which, theoretically, are genetically similar. Random sampling was then used to select almost-equal-to 10% from each group. Accessions with inadequate data for multivariate analysis were selected using a 10% random sample from each country of origin. Accessions from countries having few (less-than-or-equal-to 5) entries in the collection were pooled and a 10% random sample was selected. The resulting 831 accessions form a core collection for peanut. Examination of data for six phenotypic traits indicated that the genetic variation expressed for each trait in the entire collection has been preserved in this core collection. This peanut core collection has many potential uses and should increase the utilization of peanut germplasm resources.
