Validating a core collection of Peruvian quinoa germplasm

Genetic Resources and Crop Evolution 46: 285–290, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands. 285 Validating a core collection of Peruvian quinoa germplasm Rodomiro Ortiz1,∗ , Sten Madsen1 , Enrique N. Ruiz-Tapia1,2 , Sven-Erik Jacobsen3 , Angel Mujica-Sánchez2 , Jørgen L. Christiansen1 & Olav Stølen1 1 The Royal Veterinary and Agricultural University, Department of Agricultural Sciences, 40 Thorvaldsensvej, DK-1871 Frederiksberg C, Copenhagen, Denmark; 2 Universidad Nacional del Altiplano-Puno, Av. del Ejército 329, Puno, Perú; 3 Centro Internacional de la Papa, Apartado 1558, Lima, Perú (∗ corresponding author. E-mail: [email protected]) Received 17 March 1998; accepted in revised form 5 October 1998 Key words: average linkage cluster analysis, Chenopodium quinoa, geographical variation, salt tolerance Abstract A Peruvian quinoa core collection, containing 103 chosen ecotypes or landraces, was defined on a geographically stratified non-overlapping sampling procedure. The objective was to determine whether this protocol was correct. Hence, a phenotypic distance matrix among 76 accessions from this core collection was created by calculating the difference between each pair of accessions for each characteristic. These 76 accessions were chosen because they had complete data for eight morphological descriptors. The diversity index was calculated by averaging all the differences in the phenotypic value for each quantitative descriptor divided by the respective range, and considering a matching coefficient for the qualitative descriptors, i.e., 0 for characteristics that matched, and 1 if they did not. The total sum of squares from the analyses of variance for the phenotypic diversity index, based on morphological descriptors, was partitioned between and within Peruvian Departments. The between Department cluster variance was smaller (0.0022) than the within Department cluster variance (0.0144), because accessions with the same stem colour or similar inflorescence type and colour were grouped together irrespective of their Department. This was further corroborated by the average linkage cluster analysis, which was performed on the phenotypic diversity matrix to study the pattern of variation and the relationship between the quinoa landraces of this core subset according to their geographical origin. Within each sub-cluster (based on above qualitative descriptors) accessions were ranked by their Department of collection, which indicated that the sampling method was appropriate for choosing accessions for the Peruvian quinoa core collection. Introduction Quinoa (Chenopodium quinoa Willd.) was domesticated by pre-Columbian farmers in the Andean plateau (or ‘Altiplano’ in Spanish) between Perú and Bolivia. Although quinoa was developed as a short-day tropical pseudo-cereal crop, some adapted genotypes have shown their potential in long-day northern European environments in recent years (Darwinkel & Stølen, 1997; Jacobsen, 1997; Ortiz & Stølen, 1997; Stølen et al., 1997; Jacobsen et al., 1994; Jacobsen & Stølen, 1993). Quinoa grains, which are the fruit of a herbaceous plant, are cooked before being eaten. They have a high protein content (13–16%) and high energy value (51– 61% starch), and are rich in essential amino acids such as lysine, methionine and cystine (Mastebroek & van Soest, 1994). Furthermore, quinoa leaves may be also eaten. Hence, quinoa has a great value for both human and animal nutrition. However, the quinoa seed coat (pericarp) contains saponins that may reduce the palatability owing to their bitterness and they could also be toxic (Darwinkel & Stølen, 1997). Saponins may be removed by washing or abrasion of seeds, or through the selection of saponin-free (ca. 0% content) quinoa cultivars. Ex-situ germplasm collections are becoming large and diffuse, which may lead to ineffective man- 286 agement and non-rational utilization of genetic resources by plant breeders. A core collection is a subset of a large germplasm collection that contains chosen accessions representing the genetic variability of the whole collection (Brown, 1989). Core collections must be assembled by grouping accessions and sampling within these groups (Hodgkin et al., 1995). A core collection of Peruvian quinoa germplasm has been defined recently (Ortiz et al., 1998) from the whole collection (1029 accessions) held at the genebank of the Universidad Nacional del Altiplano-Puno (UNAP). This core collection contains 103 chosen ecotypes or landraces that may capture most of the genetic variability available in this Peruvian germplasm. The accessions of this core collection were chosen based on a geographically stratified non-overlapping sampling procedure. This quinoa core collection has already been used for screening for salt tolerance during seed germination (Ruiz-Tapia et al., 1997). The most tolerant accessions (> 75% seed germination at 0.6 M NaCl) were originally collected in different districts of Puno and Cusco (southern Perú), and show distinct morphological characteristics. Furthermore, four of these accessions were selected as promising genotypes because of their speed of seed germination under salt stress and further development of cotyledons in their seedlings. Nevertheless, the efficacy of the strategy considered for the development of the Peruvian core collection should be tested. Hence, the aim of this research was to investigate whether the core collection of Peruvian quinoa germplasm was well defined. Materials and methods A total of 76 out of 103 accessions of the core collection were chosen for this research. These accessions, collected in six of the Peruvian Departments that grow quinoa, have location and altitude descriptors and a complete description for eight morphological descriptors, which were recorded in either a categorical scale (1-N), or in SI units (Mujica et al., 1991). The qualitative descriptors were stem and inflorescence colour, and inflorescence type, whereas the quantitative descriptors were days to flowering, inflorescence length, plant height, plant biomass, and grain yield. Characterization of this germplasm was based on data recorded at the same experimental station of the UNAP at Camacani (3825 m). A phenotypic distance matrix was created by calculating the difference between each pair of accessions for each characteristic (Hill et al., 1998). For qualitative descriptors recorded in well defined phenotypes using a 1 to N scale, the distance between two quinoa accessions was scored as zero if their phenotypes matched, and one if they did not. For each quantitative descriptor, the distance between two accessions was calculated by averaging all the differences in the phenotypic value for each descriptor divided by the respective range (Johns et al., 1997), thereby transforming these distances onto a 0 to 1 scale. Thus the phenotypic distance between two accessions was determined by summing the individual descriptor distances between them, and dividing by the total number of descriptors recorded in both accessions. The analyses of variance for the phenotypic diversity index, based on morphological descriptors, partitioned the variation into between and within distinct geographical departments. Phenotypic distances were treated as deviations from the Department mean, and the analyses considered the squared deviations as variances. The ratio φ F S (Wright, 1951) was calculated by dividing the between-cluster mean square (σ 2 C ) and the total-mean square (σ 2 W + σ 2 C ), where σ 2 W was the within-cluster mean square. Wright’s φ F S represents the correlation between random genetic accessions within a group relative to random accessions from the population at large. A value close to 1 (maximum) indicates greater partitioning of the population into sub-groups. Average linkage cluster analysis was performed on the phenotypic diversity matrix to study the pattern of variation and relationship between quinoa accessions according to the Department from which they were collected. All statistical analyses were calculated using SAS (Anonymous, 1990). Results and discussion The lowest phenotypic diversity index (Table 1) was calculated as 0.0207 between two accessions from the southern Department of Puno: 280 (collected at 3850 m in the district of Ayaviri) and 870 (collected at 3830 m in the district of Chucuito). Similarly, the largest phenotypic distance (0.8068) was also recorded in this department between accessions 19 (collected at 3885 m in the district of Pomata) and 371 (collected at 3870 m in the district of Ilave). It was not surprising to observe this wide diversity in Puno, because this Department accounts for most of the accessions in both 287 Table 1. Range of the phenotypic diversity index based on morphological descriptors between (above diagonal) and within (in italics in diagonal) quinoa accessions according to Department of collection Department L5 L6 L8 L9 L12 L21 Ayacucho (L5) 0.1085 0.4857 0.2778 0.5609 0.0587 0.5729 0.1407 0.4804 0.4716 0.5904 0.2156 0.6994 Cajamarca (L6) 0.3935 0.6229 0.4355 0.5710 0.4491 0.2465 0.7973 Cusco (L8) 0.0614 0.5627 0.1982 0.5622 0.4514 0.6274 0.1992 0.7365 0.1976 0.2346 0.5032 0.5898 0.2608 0.6964 Huancavelica (L9) Junin (L12) 0.1350 0.6770 Puno (L21) 0.0207 0.8068 the whole (46%) and the core (65%) collection, and grows 85% of the Peruvian quinoa acreage (Ortiz et al., 1998). One of the most distinct accessions from the core subset was collected at Junin (876), a Department in central Perú accounting for 0.8% of the accessions in UNAP’s genebank, where local farmers grow quinoa on a small scale (1% of the total national acreage). This finding confirmed that the method chosen to allocate a random accession from this Department to the core collection was correct, because this potential unique genetic diversity was not missed during the sampling. Likewise, the only accession in the core collection from the northern Department of Cajamarca (585), which accounts for 0.5% of the accessions in UNAP’s genebank and 0.4% of the national acreage, also showed a distinct morphology in this core subset, as shown by its phenotypic diversity index from other accessions collected in other Peruvian Departments. Although the source of variation between Departments was highly significant (P < 0.001) (Table 2), the between geographical cluster variance (0.0022) was smaller than the within cluster variance (0.0144). Thus, Wright’s φ F S was calculated as 0.1325, which confirmed that much of the variation between accessions was due to differences within rather than among Departments. This low differentiation between De- partments reflects the nature of the sampling method chosen to select accessions for the Peruvian quinoa core collection. Accessions were allocated to the core subset from each Department of collection (Ortiz et al., 1998). For example, the multivariate pattern of morphological variation within Puno, Cusco and Ayacucho (> 100 accessions from each in the whole collection) determined the selection of accessions for the core, while the single accessions from Cajamarca and Junin (≤ 10 accessions) were chosen at random, and those from Huancavelica (20 accessions) were allocated according to their altitude for this Department. Further analysis revealed that accessions were ranked together in the dendrogram by the average linkage cluster method (Figure 1) according to their stem colour or similar inflorescence type and colour, irrespective of their geographical origin. Yet, accessions were ranked by their geographical origin within each sub-cluster early defined by the above qualitative descriptors. For example, all landraces from accession 955 (top of dendogram) to 18 possessed yellow stems and white inflorescences. In this sub-cluster, accessions were gathered together by their respective Department of origin, except those from Huancavelica (972, 974 and 966). Two accessions collected at lower altitudes (< 3300 m) were clustered in between landraces from Ayacucho, while accession 966, collected 288 Figure 1. Dendrogram of clusters resulting from the phenotypic diversity index on quinoa accessions. Department codes: L5 = Ayacucho, L6= Cajamarca, L8 = Cusco, L9 = Huancavelica, L12= Junin, L21 = Puno. 289 Table 2. Analysis of variance between and within Department clusters for the phenotypic diversity index based on agronomic characteristics Source of variation Mean square Expected mean square Between Department clusters Within Department clusters 0.29764100∗∗∗ 0.01440785 σ 2W + n σ 2C σ 2W ∗∗∗ indicates that the source of variation was significant according to respective F-test at P < 0.001, while n is the weighted mean for the total number of comparisons within each cluster (i.e., 129.71). Table 3. Phenotypic diversity index between most promising accessions of the Department of Puno whose seeds germinated at 0.6 M NaCl Accession 197 880 864 346 856 Collection site Altitude District Germination (%)1 3825 3825 3850 3885 3870 82 82 80 77 76 Puno Juliaca Ayaviri Yunguyo Checachall. 880 Phenotypic diversity index 864 346 856 0.2593 0.3403 0.4263 0.0767 0.2910 0.3013 0.6030 0.4854 0.5943 0.5800 1 After Ruiz-Tapia et al. (1997). at a higher altitude (3670 m), was in between these accessions collected in Cusco. This clustering confirmed a previous observation by Ortega (1997), who reported that there were two major diversity groups of Peruvian quinoa: one grown by farmers in the valleys and characterized by a branched, lax inflorescence and exhuberant plant growth, and another grown by farmers in the highlands that showed a terminal panicle and shorter plants. For example, some landraces from Puno (collected at ≥ 3800 m) are, on average, early flowering and short, with large inflorescence and high biomass (Ortiz et al., 1998; Mujica et al., 1991), and hence have a high yield potential on the Altiplano. The most promising accessions from the Department of Puno selected due to their seed germination under salt stress (Ruiz-Tapia et al., 1997) were collected in different districts (Table 3), and showed very distinct morphologies (except for 197 and 346). Although accessions 197 and 346 exhibited similar phenotypes, their collection site and seed germination result were different, which suggest that they may be distinct accessions. This screening for salt tolerance also shows how the core collection could be a starting point for further exploitation of the genetic resources available in the quinoa genebank of UNAP. Although the phenotypic diversity index and the average linkage cluster analysis provides support for the method chosen by Ortiz et al. (1998) in the de- velopment of a core collection of Peruvian quinoa germplasm, further research will be needed to confirm whether these accessions showing a small phenotypic diversity index, either within or between Departments, are indeed distinct. DNA markers may provide a means to determine whether unique genetic material form this core collection and to discard any potential duplicated accession. As yet, however, DNA marker systems are not available for this crop. Acknowledgements We thank Dr John Hill, visiting lecturer in the Department of Agricultural Sciences at The Royal Veterinary and Agricultural University, for his critical review of this paper, and the partial financial support of the Danish International Development Assistance (Danida) for this research. References Anonymous, 1990. SAS procedure guide, version 6, 3rd ed. SAS Institute Inc., Cary, North Carolina. Brown, A.H.D., 1989. Core collections: a practical approach to genetic resources management. Genome 31: 818–824. Darwinkel, A. & O. Stølen, 1997. Understanding the quinoa crop: guidelines for growing in temperate regions of N.W. Europe. European Commission (DG VI F.II.3 – Coordination of Agricultural Research), Brussels, 23 pp. 290 Hill, J., H.C. Becker & P.M.A. Tigerstedt, 1998. Quantitative and Ecological Aspects of Plant Breeding. Chapman & Hall, London, 275 pp. Hodgkin, T., A.H.D. Brown, Th.J.L. van Hintum & E.A.V. Morales (Eds), 1995. Core collections of plant genetic resources. John Wiley & Sons, Chichester, 269 pp. Jacobsen, S.E., 1997. Adaptation of quinoa (Chenopodium quinoa) to Northern European agriculture: studies on developmental pattern. Euphytica 96: 41–48. Jacobsen, S.E. & O. Stølen, 1993. Quinoa – Morphology, phenology and prospects for its production as a new crop in Europe. Eur. J. Agron. 2: 19–29. Jacobsen, S.E., I. Jørgensen & O. Stølen, 1994. Cultivation of quinoa (Chenopodium quinoa) under temperate climatic conditions in Denmark. J. Agric. Sci. (Cambridge) 122: 47–52. Johns, M.A., P.W. Skroch, J. Nienhuis, P. Hinrichsen, G. Bascur & C. Muñoz-Schick, 1997. Gene pool classification of common bean landraces from Chile based on RAPD and morphological data. Crop Sci. 37: 605–613. Mastebroek, H.D. & L.J.M. van Soest, 1994. Gierstmelde blijkt multi-purpose-gewas. Prophyta 1: 15–17. Mujica, A., M. Rodriguez, S. Pino, V. Apaza & F. Huapaya, 1991. Catálogo de cultivos andinos: quinua (Chenopodium quinoa Willd.). Universidad Nacional del Altiplano Puno, Perú. Ortega, R., 1997. Peruvian in situ conservation of Andean crops. In: N. Maxted, B.V. Ford-Lloyd & J.G. Hawkes (Eds), Plant genetic conservation: the in situ approach. Chapman & Hall, London, pp. 302–314. Ortiz, R., E.N. Ruiz-Tapia & A. Mujica-Sánchez, 1998. Sampling strategy for a core collection of Peruvian quinoa germplasm. Theor. Appl. Genet. 96: 475–483. Ortiz, R. & O. Stølen (Eds), 1997. Spelt and quinoa. Crop Development for the Cool and Wet Regions of Europe. European Commission, Brussels, 129 pp. Ruiz-Tapia, E.N., O. Stølen, J.L. Christiansen & R. Ortiz, 1997. Assessment of quinoa accessions (Chenopodium quinoa Willdenow) for tolerance to salinity. In Ortiz, R. & O. Stølen (Eds), Spelt and quinoa. Crop Development for the Cool and Wet Regions of Europe. European Commission, Brussels, pp. 69–78. Stølen, O., K. Bruhn, K. Pitham & J. Hill (Eds), 1997. Small grains cereals and pseudocereals. Crop Development for the Cool and Wet Regions of Europe. European Commission, European Cooperation in the Field of Scientific and Technical Research, Brussels, 183 pp. Wright, S., 1951. The genetical structure of populations. Ann. Eugen. 15: 323–324.