INTRODUCTION: One of possible ways how to affect assimilate distribution is to increase grain weight through increasing grain number (1). For this purpose, wide genetic variability in distinctions of spike morphological structure in wheat can be used. While particular elements of spike productivity are of a multigenic character of inheritance, spike morphological structure is a character which is easily measured and, at the same time, controlled by a less number of genes. The objective of this study was to analyse spike productivity in selected groups of winter wheat genotypes (Triticum aestivum L.) differing in spike morphological structure.

MATERIAL:A collection of wheat gene resources exhibiting different morphological structure of the spike which is maintained at the Agricultural Research Institute Kromeriz, Ltd. (ARI) was used in the presented study. The greatest part of original genetic materials was provided by the research institutes in Zagreb (2) and Novi Sad (3) (Yugoslavia), VIR in Sankt Petersburg (Russia), and CIMMYT (Mexico). These materials were mostly derived from wide crosses of T. aestivum L. with tetraploid species. At the ARI Kromeriz, this material was crossed with selected Czech and foreign cultivars. Based on the spike morphology the developed lines (F₇-F₁₂ generations) were classified into the groups as follows (Fig. 1):

A. normal spike structure (NS), a single spikelet is situated in one node of spike rachis. "Gigas" forms have been described which exhibit an increased spikelet number per spike rachis and considerably greater length of spike rachis (4). In some NS forms there is a higher number of florets in spikelets, "multifloret", which is associated with fan-like arrangements of florets, "flabellum" (5).

B. vertical sessile spikelets (VSS), sometimes designated "banana twin spikelets", when two or three spikelets grow up vertically in a spike rachis node. The VSS expression is controlled dominantly (6, 7, 8) or recessively.

C. tetrastichon (TSS – tetrastichon sessile spikelets) when three or mostly only two spikelets are sessile close to each other in a horizontal position in a spike rachis node.

D. floribunda – a high number of spikelets grow up in a common spike rachis node close to each other and, at the same time, above each other (7). This morphotype is characteristic of little space among spikelets, which often hampers the development of inflorescence organs. That may result in reduced spike fertility (9). Fertile forms are named "multirow spike" (MRS) where spikelets are distributed in the spike circumference (10).

E. spike branching (TFS – transitional forms spikelets). Branchiness of the turgidum type was incorporated in T. aestivum L. using wide crosses between T. aestivum L. and branching forms of T. turgidum L. (2, 3). It displays a wide range of different levels which are usually less stable in the phenotype in various environments. In some cases, a branching level also depends on the order and a differentiation degree of apices in one plant, which can be expressed by the presence of both branching and standard spike morphotypes in a plant, so-called "hetero-branching" (11). Spike branching has been studied by several authors (2, 3, 8, 9, 11, 12, 13). A common feature of these studies is that one or more recessive genes depending on the genotype used control spike branching.

F. larger size of glumes and lemmas (LG). There are some wheat forms (T. aestivum L.) in which a larger size of glumes was transferred by wide crosses with T. polonicum L. A considerably larger size of cover parts of florets is a typical character of T. polonicum L. and is conditioned by P gene on long arm of chromosome 7A (14).

G. screwedness of spike rachis (SCR – screwed spike). There are usually both directions of screwedness. It is a mutation, which is controlled by a recessive gene (15).

METHODS: The samples were cultivated at the ARI Kromeriz in the 1995/96 growing season. Progenies of selected plants were grown in hand-sown plots of 0.8 m² without replications. There were 10 rows, 15 seeds per row, i.e. 150 seeds in total in each plot. Check cultivars were Astella, Hana and Siria. Thirty spikes of each plot were used to analyse elements of spike productivity and sheaves were used to assess harvest index.

RESULTS AND DISCUSSION: Table 1 shows that the highest values of grain weight per spike were recorded in SCR, VSS, and NS morphotypes (2.39, 2.21 and 2.14 g, respectively). Grain weight in check cultivars did not surpass 2 g. Increasing grain weight per spike did not lead towards adequate increase in harvest index. Grain weight per spike correlated more with stem weight (r=0.86**) than with stem length (r=0.39**). That indicates a possibility of selecting short-stem genotypes with high spike productivity. A number of days to maturity showed a higher correlation with a node number per spike rachis (r=0.61**) than with a spikelet number per spike (r=0.27*). It suggests that it would be possible to partly limit undesired lateness by selecting forms with a high grain number in spikelets or those with an increased number of spikelets growing up in common nodes of spike rachis. Significant correlation between a grain number per spike and stem diameter (r=0.38**) show the importance of a size and number of vascular bundles in stem for spike productivity (16). Yielding ability of selected genotypes, which were sown in yield plots in a successive year, did not usually reach the yields given by check cultivars due to lower tillering ability. The investigated genetic material can be used for genetic and physiological studies. A limited number of gene resources are important for breeding such as those which have a higher grain number per spike rachis node (i.e. multifloret forms of NS and SCR or forms with a higher spikelet number VSS, TSS, and MRS). They are not widely used because of some adverse characters transferred from other less cultivated wheat species. Seventy-four selected gene resources have been provided to breeders and the Gene Bank at RICP Prague for conservation.

Table 1: Characteristics of wheat gene resources groups with different spike morphology (mean values of the sets)

REFERENCES:

1. Siddique, K.H.M., Whan, B.R.: Euphytica, 73, 1994: 241-254

2. Koric, S.: J. Sci. Agric. Res., Zagreb, 41, 1980: 271-282

3. Dencic, S.: Proc. 7^th IWGS, Cambridge, 1988: 499-502

4. Atsmon, D., Busch, M.G., Evans, L.T.: Aust. J. Plant Physiol., 13(3), 1986: 365-379

5. Foltýn, J.: Genet. a Šlecht., 20(1), 1984: 9-14

6. Foltýn, J.: Genet. a Šlecht., 25(3), 1989: 117-123

7. Foltýn, J.: Genet. a Šlecht., 27(2-3), 1991: 131-137

8. Foltýn, J.: Genet. a Šlecht., 26(1), 1990: 13-22

9. Martinek, P.: Genet. a Šlecht., 30(1), 1994: 61-74

10. Martinek, P., Bednár, j., Provazníková, E., Krulícková, K.: Proc. "Agronomická fakulta a vývoj polnohospodárstva na Slovensku", Nitra, Slovakia, 1996: 47-53

11. Huang, G. Yen, C.: Proc. 7^th IWGS, Cambridge, 1988: 527-532

12. Pennell, A. L., Halloran, G. M.: Euphytica, 32, 1983: 763-776

13. Košner, J., Foltýn, J.: Genet. a Šlecht. , 25(1), 1989: 11-17

14. Watanabe, N., Yotani, Z., Fruta, Y.: Euphytica, 91(2), 1996: 235-239

15. Smocek, J.: Rostl. Výr., 37(6-7), 1991: 507-514

16. Nátrová, Z., Nátr, L.: Field Crops Research. 31(1-2), 1993: 121-130

ACKNOWLEDGEMENT: The present work was supported by the Grant Agency of the Czech Republic, project no. 503/95/1290.