Experimental
Ziziphus mauritiana Lam., commonly known as ‘ber’ or Indian jujube (Supplementary Fig. S1A, available online), is one of the most important fruit crops grown in the Thar Desert of India (Supplementary Fig. S1D, available online). It is drought tolerant and widely distributed in the warmer parts of India, Pakistan, Bangladesh, Sri Lanka, Central to South Africa and the northern parts of Australia (Pareek, 2001). Its centre of origin is believed to be in the Indo-Malaysian region (Pareek, 2001; Jagwan and Singh, Reference Jagwan and Singh2009). The highly heterogeneous phenotype of this species suggests a long history of cross-pollination. More than 300 accessions (Supplementary Table S1, available online) have been maintained at the field genebank of the Central Institute for Arid Horticulture since 1996 (Awasti and More, Reference Awasti and More2009). Since the maintenance of such a large collection of germplasm is expensive, the best way to promote the utilization of the collection is to develop a core collection (Holden, Reference Holden, Holden and Williams1984) that captures as much genetic variation as possible and minimizes redundancy. Core collections have been developed for various crops (Singh, Reference Singh2009; Lin et al., Reference Lin, Guo, Bai, Yong, Jia, Li and Sheng2011; Socias i Company et al., Reference Socias I Company, Alonso, Espiau, Fernandez, Marti, Kodad, Avanzato, Bacchetta, Botta, Drogoudi, Duval, Metzidakis, Rovira, Silva and Spera2011; Wang et al., Reference Wang, Zhang, Sun, Ning and Yang2011) using a variety of criteria and sampling strategies (Zhao et al., Reference Zhao, Chung, Lee, Ma, Kim, Kim, Chung, Lee, Kim, Kim and Park2011). The core collection of Indian jujube described herein was assembled by considering 23 qualitative (Supplementary Table S2, available online) and 20 quantitative (Supplementary Table S3, available online) traits, following a list of descriptors given by Vashishtha and Pareek (Reference Vashishtha and Pareek1989) and Vashishtha (Reference Vashishtha2001). They were measured during the periods of vegetative growth, flowering, fruiting and fruit development by sampling three clones per accession (produced by budding). The advanced M strategy implemented in the program PowerCore 1.0 was used to develop the core collection (Kim et al., Reference Kim, Chung, Cho, Ma, Chandrabalan, Gwag, Kim, Cho and Park2007), and clustering was achieved using the Euclidean method for most of the accessions belonging to the core and whole collections. The diversity present in both the core and whole collections was estimated using SAS 9.3 software (SAS Institute Inc., 2011).
Of the 330 Z. mauritiana germplasm accessions, 52 (15.8 %) represented the core collection (Supplementary Table S1, available online), which had a polymorphic information content of 0.99 with an efficiency index of 0.83 and a coverage of 100 %. The mean difference (MD), variable difference (VD) and coincidence rate (CR) between the core and whole collections for continuous variables were 6.1, 34.9 and 98.1 %, respectively. There was no significant difference in the Shannon–Weaver and Nei diversity indices between the core and whole collections for qualitative traits. In total, 85 classes were identified for 23 qualitative traits in the whole collection, which were found to be equally represented in each class in the core collection (Supplementary Table S2, available online). Mean values and ranges between the core and whole collections for each quantitative trait were not significantly different (Supplementary Table S3, available online). Phylogenetic trees generated for the core and whole collections were easily comparable. Overall, four major clusters were identified for the whole collection. The majority of the accessions were in cluster II and were closely related, whereas the accessions in clusters III and IV were more diverse (Supplementary Fig. S2, available online), which was similar to the phylogenetic tree of the core accession (Supplementary Fig. S3, available online).
Discussion
The core collection of Z. mauritiana developed herein is for its conservation and evaluation for various traits as the collection involves a large number of accessions (Brown, Reference Brown1989). Many methods are available to develop core collections that represent the entire diversity of the whole collection (Gouesnard et al., Reference Gouesnard, Bataillon, Decoux, Rozale, Schoen and David2001; Jansen and van Hintum, Reference Jansen and van Hintum2007). In this study, PowerCore software was used to develop a core collection that represented 100 % diversity of the whole collection with high efficiency (Kim et al., Reference Kim, Chung, Cho, Ma, Chandrabalan, Gwag, Kim, Cho and Park2007; Zhao et al., Reference Zhao, Chung, Lee, Ma, Kim, Kim, Chung, Lee, Kim, Kim and Park2011). A core collection comprising 52 accessions is a suitable size for this purpose (Brown, Reference Brown1989; Singh, Reference Singh2009). Statistical consistency parameters such as MD, VD and CR indicated that there is enough consistency between the core and whole collections (Kim et al., Reference Kim, Chung, Cho, Ma, Chandrabalan, Gwag, Kim, Cho and Park2007). The distribution of the ranges of each variable in the core collection is also well represented with respect to the whole collection as the CR of the core collection set is more than 80 % (Hu et al., Reference Hu, Zhu and Xu2000). Various diversity indices, mean values, ranges and distributions of the core accession within the clusters of the phylogenetic tree of the whole collection suggested that core accessions represented the entire range of diversity of the whole collection. Phylogenetic clustering of the accessions was not dependent on their geographical locations of origin. This is the first report of a core collection developed for Z. mauritiana, and this will be a working collection facilitating the breeder and biotechnologist in the identification of economically important traits for the conservation and improvement of this species.
Supplementary material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S1479262113000257
Acknowledgements
The authors thank the World Bank-funded National Agricultural Innovation Project (NAIP), ICAR for providing financial assistance under the subproject ‘Bioprospecting of genes and allele mining for abiotic stress tolerance’. They are also highly grateful to Dr K.V. Bhat, NBPGR, New Delhi for his suggestions during the investigation and Dr Robert Koebner, CropGen International, UK for his critical remarks and corrections on the manuscript.
