Microsatellite diversity at 18 loci was analysed in 94 individual plants of 10 wild barley, Hordeum spontaneum (C. Koch) Thell., populations sampled from Israel across a southward transect of increasing aridity. Allelic distribution in populations was not distributed randomly. Estimates of mean gene diversity were highest in stressful arid-hot environments. Sixty-four per cent of the genetic variation was partitioned within populations and 36% between populations. Associations between ecogeographical variables and gene diversity, H(e), were established in nine microsatellite loci. By employing principle component analysis we reduced the number of ecogeographical variables to three principal components including water factors, temperature and geography. At three loci, stepwise multiple regression analysis explained significantly the gene diversity by a single principal component (water factors). Based on these observations it is suggested that simple sequence repeats are not necessarily biologically neutral.

Abstract. The HVAI protein belongs to the LEA3 group, which is expressed during the late stage of seed maturation. It is also induced by exogenous abscisic acid (ABA) and a variety of environmental stresses in germinating barley (Hordeum vulgare L.). In the present work, the potential role of HVA! was investigated by studying its tissue distribution and subcellular localization in mature and stressed seeds by immuno-microscopic methods. In the mature seed, HVA1 protein was detected in all tissues except the non-living starchy endosperm. During germination the amount of HVA 1 protein decreased but did not totally disappear. Incubation with 100/aM ABA, cold treatment or drought stress dramatically increased HVA1 expression in the germinated seed. In this work, the distribution of a LEA3 group protein was studied in a cereal seed for the first time by immuno-electron microscopy. In the scutellum and aleurone layer, HVA1 was localized both in the cytoplasm and protein storage vacuoles (PSVs). HVA1 protein was found to be threefold more abundant in PSVs than in the cytoplasm of an unstressed seed tissue. The ratio increased with ABA or stress treatments to at least ninefold. The role of HVA1 in PSVs remains unclear: a previously suggested possibility is ion sequestration to prevent precipitation during stress. On the other hand, HVA1 protein could also be degraded in PSVs. HVA1 protein does not have the signal peptide typical of proteins which are glycosylated and targeted into the vacuole via the Golgi complex. Because HVA1 is not glycosylated, it may use an alternative, ER-independent vacuolar pathway, also found in yeast cells.