Notes: Research conducted in El Salvador, Central America, demonstrated that an interval of 3 days between application of glyphosate and tillage was sufficient to cause 90% reduction in purple nutsedge (Cyperus rotundus L.) plants, while delays of 11–23 days generally gave slightly less reduction. à second application to the same plots 35 days following tillage resulted in more than 90% reduction with all intervals. Approximately 3 months after the initial treatment, tuber numbers had been reduced to half the original population. Germination of the remaining tubers was reduced by more than 50%.Glyphosate applied during the dry season caused an average of 79% reduction in plant numbers compared with 88% in the rainy season. However, in the dry season, the remaining plants had no competition from other weeds and after 5 months there was only à 40% reduction in nutsedge population. During the rainy season, 1, 2 and 3 kg/ha were equally effective, but 1 kg/ha was not sufficient in the dry season.

Notes: Glyphosate at 2 kg/ha was more effective in reducing regrowth of purple nutsedge (Cyperus rotundus L.) scapes at 90% than at 50% relative humidity (r.h.), and more effective at −2 bars than at −11 bars of plant water potential. Regrowth of treated plants subjected to water potentials of −1 to −8 bars was reduced 54–60% while at −11 bars growth inhibition was only 34%. A time interval of as little as 8 h between application and excision was sufficient to give 47% reduction in regrowth at 90% r.h. None of the treated plants, except those clipped immediately after application, produced new shoots from the basal bulb, while all the untreated control plants produced one or more new shoots. Experiments using 14C-glyphosate substantiated these results. Three times more 14C-label was translocated into the underground parts of nutsedge at 90% than at 50% r.h. Twice as much translocated at −2 bars than at −11 bars of water potential.

Notes: Abstract MORB suites display variations in their chemical differentiation trends which are closely related to the incompatible element enrichment of the basalts. We examine suites of primitive to evolved basalts from the Pacific-Nazca Ridge at 28° S (mostly depleted); from the Juan Fernandez microplate region (depleted) and from the Explorer Ridge, northeast Pacific (mostly enriched). Trends for incompatible element enriched MORBs consistently show less depletion of Al2O3 and less enrichment of FeO when plotted on MgO variation diagrams. Least squares modeling indicates that enriched basalts have undergone less plagioclase crystallization than depleted basalts especially in the early stages of differentiation. Using thermodynamic modelling, we show that variations between MORB differentiation trends result largely from differences in the major element chemistry and H2O content of primary magmas. Our chosen enriched and depleted near-primary magmas are similar in major element chemistry but the enriched near-primary magma has higher H2O and lower Al2O3 than the depleted near-primary magma. The MORB crystallization sequence is: olivine→olivine+plagioclase → olivine+plagioclase+high-Ca pyroxene; and the separate and combined effects of lower Al2O3 and higher H2O are to cause plagioclase to crystallize later (lower temperature), and to make the interval of olivine+plagioclase crystallization shorter. As a result, enriched differentiates have higher Al2O3 and lower FeO than depleted MORBs at a given MgO content, even though their parents' Al2O3 is lower. Crystallization of enriched basalts at higher pressure than depleted basalts is not able to account for differences between the differentiation trends because the proportion of plagioclase is higher during three-phase crystallization at high pressure. The variations in trends do not depend on geographic location and thus are superimposed on any regional variations in MORB chemistry or mantle source. Nor are they related to spreading rate. Depleted basalts from the fast-spreading 28° S and Juan Fernandez ridges have differentiation trends similar to depleted basalts from the medium-spreading Galapagos Spreading Center, whereas differentiation trends for enriched basalts from the medium-spreading Explorer Ridge are quite different. Fe3+/Fetotal is similar (and quite low) for enriched and depleted basalts, indicating that neither oxidation state nor early magnetite crystallization are important.