Variation existed between plants of the lucerne (Medicago sativa L.) cultivar CUF 101 for dry matter production, shoot number and length, and leaf damage when grown for 70 days under 250 mM NaCl (15 h photoperiod, 20�C day, 10�C night). Salt tolerance evaluation using the criteria percentage leaf damage (percentage of total number of leaves with complete or partial necrosis) and length of the main shoot, isolated plants which showed salt tolerance of reasonably high heritability (h2=0.41). Two generations of recurrent selection for tolerance significantly increased the mean population tolerance without decreasing production under non-saline conditions. While both sodium and chloride concentrations of the shoot were lower in the tolerant than in less tolerant plants, chloride was more closely associated with salt tolerance than sodium. Sodium and chloride concentrations in the roots did not vary with the level of salt tolerance. No association of shoot and root potassium concentration with tolerance was evident. Selection for salt tolerance in lucerne plants using percentage leaf damage of less than 10% as the main criterion should give a rapid response to selection. The efficiency of selection may be increased if selection is based on the efficiency of chloride exclusion from the shoots and/or the level of chloride tolerated by the shoots prior to leaf damage becoming evident.

Efforts to develop new crop varieties with improved salt tolerance have been intensified over the past 15-20 years. Despite the existence of genetic variation for salt tolerance within species, and many methods available for expanding the source of genetic variation, there is only a limited number of varieties that have been developed with improved tolerance. These new varieties have all been based upon selection for agronomic characters such as yield or survival in saline conditions. That is, based upon characters that integrate the various physiological mechanisms responsible for tolerance. Yet over the same time period, knowledge of physiological salt responses has increased substantially.Selection and breeding to increase salt tolerance might be more successful if selection is based directly on the physiological mechanisms or characters conferring tolerance. Basic questions associated with using physiological selection criteria are discussed in the paper. These are centred around the need for genetic variation, the importance of the targeted mechanism, the ease of detection of the physiological mechanism (including the analytical requirements) and the breeding strategy. Many mechanisms, including ion exclusion, ion accumulation, compatible solute production and osmotic adjustment have been associated with genetic variation in salt tolerance. Yet their successful use in improving salt tolerance, via physiological selection criteria, is largely non-existent. Consideration is given to the role of physiological criteria in the short and long term in improving salt tolerance. In several glycophytic ~ species, particularly legumes, physiological selection based on ion exclusion from the shoots shows promise. Recent results for white clover indicate the potential for using a broad physiological selection criterion of restricted C1 accumulation in the shoots, with scope for future refinement based upon the specific physiological characters that combined result in ion exclusion.

Summary. The salt tolerance of 29 lines of annual and perennial forage legume species was evaluated in 4 separate experiments over 0–100 mol NaCl/m3 in the greenhouse with the aim of identifying genetic material that is more salt tolerant than the more traditionally grown forage legume species. Several species or lines showed potential as salt-tolerant germplasm including Trifolium tomentosum, 2 lines of T. squamosum and T. alexandrinum cvv. Mescani and Wardan which were all more salt tolerant than T. subterraneum. Two lines of Lotus tenuis and 1 line of L. corniculatus were also relatively salt tolerant. Some of this material had never been assessed before under saline conditions. In contrast, several other species (T. arvense, T. vesiculosum, T. angustifolium and T. pratense) were found to be extremely salt sensitive and/or produced very small amounts of dry matter over all NaCl concentrations. We believe that further selection and field evaluation (including selection for increased productivity and salt tolerance over a range of growth stages) is required for the material that showed potential in order to fully assess its performance under saline soil conditions.

The potential to improve the salt tolerance of white clover (Trifolium repens L.) v^^as evaluated in populations developed by selecting plants within the cv. Haifa which had high and low concentrations of Cl~ in the shoots. Under saline conditions (40 mM NaCl), the low-Cl populations consistently maintained lower concentrations of both Na and Cl in all plant parts and produced more dry matter than the original parent cultivar (Haifa), or the high-Cl populations. Under non-saline conditions, there were no differences between populations for these characters. In young seedlings, concentrations of Cl" in the shoots of the low-Cl and high-Cl populations started to diverge after 4 d exposure to saline conditions and there were significant differences between populations after 6 d growth at 40 mM NaCl. Results using ^®C1 suggested that the salt-tolerant low Cl population was better able to retain ^^Cl in the roots and to limit its translocation to the leaves and petioles than the high-Cl population.Realized heritability values for shoot Cl~ concentration were found to be moderate in the first cycle of selection (0-24 and 0-37 for high and low Cr concentrations respectively), but were substantially lower (0-10 and 0-09 respectively) in the second selection cycle. We propose that it is possible to increase levels of salt tolerance in white clover by selecting for low shoot Cl' concentrations under saline conditions and that this criterion could be used at very early stages of exposure to NaCl (i.e. day 4 or 6).

No abstract