Introduction
Spelt (Triticum spelta L.) is the 6 × alloploid produced by the hybrid between tetraploid cultivated emmer wheat (T. dicoccon Schrank) and the goatgrass Aegilops tauschii ssp. strangulata Coss. (McFadden and Sears, Reference McFadden and Sears1946; Kerber and Rowland, 1974). It is among the oldest of the European cereals (Harlan, Reference Harlan, Evans and Peacock1981), and was widely cultivated in Spain until the end of the 1960 s, when increasing mechanization led to its substitution by bread wheats of short stature. Today, spelt only survives in marginal farming areas of Asturias, where traditional farming systems still continue. In this region, the crop is called escanda mayor or Asturian fisga and, in most cases, is grown in mixtures with emmer. Fortunately, part of the variability of this crop has been preserved in ex situ collections.
Our collection includes 405 accessions from the ex situ collections obtained from Centro de Recursos Fitogenéticos INIA (Alcalá de Henares, Spain) and National Small Grain Collections (Aberdeen, Idaho, USA). Many of these accessions date to an expedition carried out in Asturias (northern Spain) by the Swiss Federal Research Station for Agroecology and Agriculture during the 1930 s, and have been conserved in the Germplasm Bank of this Swiss Institution since 1939 (Dr G. Kleijer, pers. commun.). All accessions were typed for their high molecular weight glutenin subunit composition (Caballero et al., Reference Caballero, Martín and Alvarez2001), and 30 were selected for multiplication and further characterization. The aim of the present work was to study the variability for some key quantitative traits in the spelt accessions from the ex situ collection, as a source of future crop improvement.
Experimental
Twenty-five spelt lines were each multiplied from a single individual of each selected accession, and grouped by awnedness, awn colour, glume colour and glume hairiness (Szabó and Hammer, Reference Szabó, Hammer, Padulosi, Hammer and Heller1996) (Table 1). Several spikes of each plant were bagged to avoid out-crossing. Progeny harvested from these spikes were grown in a plot of length 1 m in the Guadalquivir River Valley (Córdoba, Spain) using the standard agronomic practices for the region (175 kg/ha N, 90 kg/ha P and 90 kg/ha K). Ten morphological traits were measured: plant height, spike length, spike weight, distance between spikelets, number of spikelets per spike, number of grains per spike, number of grains per spikelet, grain length, thousand-grain weight, and weight of grain per spike. Mean values and coefficient of variation (CV) of each trait were computed.
High molecular weight glutenin subunit composition of the spelt lines grouped by awnedness and glume colour
N° UCO, accession number of the Germplasm Bank of the Department of Genetics in University of Córdoba; HMW, high molecular weight.
Discussion
On the basis of the chosen morphological discriminators, six botanical varieties could be identified. Only one of these is unawned, having red and glabrous glumes [var. duhamelianum (Mazz.) Koern.]. The other five are all awned: one has blue pubescent awns [var. caeruleum (Alef.) Körn.]; one has white glabrous glumes [var. arduini (Mazz.) Körn.] and one white pubescent glumes (var. albivelutinum Körn.); one has red glabrous glumes [var. vulpinum (Alef.) Körn.] and the last has red pubescent glumes (var. rubrivelutinum Körn.). A representative sample of this variability found is illustrated in Fig. 1.
Variation for awn and glume colour among spelt lines. 1, Awnless spike (ESP-019); 2, white spike (ESP-023); 3, red spike (ESP-020); and 4, blue spike (ESP-012).
Most traits were highly variable between accessions, except for plant height, number of spikelets per spike and kernel length, for which the CV was < 10% (Table 2). Mean plant height was high compared to that of modern wheats cultivated in Spain. The number of spikelets per spike varied little, so that the variation in spike length was due to differences in the distance between spikelets rather than to the number of spikelets. The number of grains per spike was significantly lower in the awnless than in the awned types, but kernel weight varied little. Awnless lines had the lowest values for all the traits measured, with the exception of the thousand-grain weight. Red glumed types had significantly lighter kernels than the awnless lines. Lines with white and blue glumes were superior than the other two types, in terms of the number of spikelets per spike, the number of grains per spike and the number of grains per spikelet, and, consequently, grain weight per spike (Table 2). They also produced the tallest plants, apart from line ESP-022, and thus showed a higher tendency to lodge.
Mean values for quantitative traits
Our data confirm that spelt lines show a wide genetic variability, similar to the situation with English emmer (Van der Veen and Palmer, Reference Van der Veen and Palmer1997) and Spanish emmer (Alvarez et al., Reference Alvarez, Caballero, Ureña, Vacas and Martín2006). The presence of awnless types showing a slightly lower plant height than those currently grown in Asturias is of interest. In the collection mission of Caballero et al. (Reference Caballero, Martín and Alvarez2007), no awnless autochthon materials were found, which suggests that genetic erosion has been occurring (Tellez-Molina and Alonso-Peña, Reference Tellez-Molina and Alonso-Peña1952). In conclusion, these spelt materials display a wide range of quantitative variability, some of which could be useful to widen the genetic background of the crop cultivated in Asturias.
Acknowledgements
This research was supported by grant AGL2004-03 361-C02-01 from the Spanish Ministry of Education and Science and the European Regional Development Fund (FEDER) from the European Union.
