Producing native seed of Stipa lagascae R. and Sch. inPresaharian Tunisia: high yields, high genetic variation orboth?

Genetic shifts are predicted to be greatest when highly diversematerial originating from an arid environment is bulk multiplied inirrigated and fertilised conditions. This is a key problem if largequantities of quality native seed are needed for restorationpurposes. In Presaharian Tunisia, we investigated two intermingledyield aspects of irrigated seed multiplication of Stipalagascae R. and Sch.: (1) nitrogen fertilisationand (2) selection for vegetative vigour. We set up twocontiguous nitrogen fertilisation trials, one withnon-selected seeds (from a nearby locality), and onewith selected seeds (48 clonal lines seeded in replicatedrows), and compared straw yields. Both selected and unselectedmaterial did not respond to more than 50 kg of nitrogenper ha. Unfertilised yields of both trials were similar,but the straw yield of fertilised selected material was more than100% higher than the straw yield of fertilised unselectedmaterial. In the trial with selected material, we collected straw andleaf yields per row, and scored each row for vigour and for phenologyduring three successive growing seasons. Genotypic differences inseed output and in mortality were large and mainly driven byvariation in potential straw for leaf ratios.The compromise between high yields and high genetic variation couldconsist of the following principles. (1) Create driergrowing conditions so that the actual straw forleaf ratio stays below the genotypic potential. (2) Useinstead fertilisation and selection to enhance seed yields.(3) Avoid bulk multiplication.