The mechanisms that allow introduced plants to become invasive are poorly understood. Here, we present a test of the evolution of increased competitive ability hypothesis, which holds that because specialized natural enemies may be absent from the introduced range, exotic plants may evolve to invest less in anti-herbivore defenses and thereby gain a competitive advantage over native plants. We grew Solidago gigantea plants derived from both the native range (North America) and the invasive range (Europe) in a common garden in the native range for 2 years. Half the plants were treated with insecticide to protect them from insect herbivores and the other half were exposed to insects that colonized the garden from nearby fields. Insect herbivore biomass was significantly higher on European plants than US plants in the first year but not the second. European plants were more heavily attacked by pathogens in both years of the study. When exposed to insect herbivores, US plants produced more seed than European plants, but when plants were protected from herbivores, seed production was equivalent between US plants and European plants. The presence of insect herbivores suppressed seed production of European plants much more than that of US plants, even though the level of herbivory experienced by European and US plants was similar in the second year, suggesting that the ability to tolerate herbivory was diminished in European plants. These results partially support the EICA hypothesis: plants from the introduced range were more susceptible to some natural enemies and benefited more from insect removal than plants from the native range. The prediction that European plants would perform better than US plants in the absence of insect herbivores was not supported.

The evolution of increased competitive ability (EICA) hypothesis predicts that invasive plant species may escape their specialized natural enemies in their introduced range and subsequently evolve with a decrease in investment in anti-herbivore chemical defenses relative to native conspecifics. We compared the chemical profile of 10 populations of US native and 20 populations of European invasive Solidago gigantea. To test for differences in inducibility between native and invasive populations, we measured secondary chemistry in both damaged and undamaged plants. We also performed bioassays with three specialist and two generalist insect herbivores from four different feeding guilds. There was no evidence that invasive populations had reduced concentrations of sesquiterpenes, diterpenes, or short-chain hydrocarbons (SCH), although significant variation among populations was detected. Sesquiterpene and diterpene concentrations were not influenced by damage to the host plant, whereas SCH concentrations were decreased by damage for both native and invasive plants. Performance of the three specialist insects was not affected by the continental origin of the host plant. However, larval mass of the generalist caterpillar Spodoptera exigua was 37% lower on native plants compared to invasive plants. The other generalist insect, a xylem-tapping spittlebug that occurs on both continents, performed equally well on native and invasive plants. These results offer partial support for the defense predictions of the EICA hypothesis: the better growth of Spodoptera caterpillars on European plants shows that some defenses have been lost in the introduced range, even though our measures of secondary chemistry did not detect differences between continents. Our results show significant variation in chemical defenses and herbivore performance across populations on both continents and emphasize the need for analysis across a broad spatial scale and the use of multiple herbivores.

Legumes oversown onto hill country in Marlborough were compared in six trials. At low altitude sites with northerly aspects, legume establishment was negligible unless resident vegetation was treated with herbicide prior to oversowing. On a southerly aspect, legumes established without herbicide although yields were higher if it was applied. 'Grasslands Pawera' red clover outyielded other legumes at 350m. producing 4.6 and 10.1 t DM/ha on a sunny and shady face respectively. Pawera also produced more than lucerne, 'Grasslands Maku' lotus, 'Grasslands Huia' white clover, alsike clover and subterranean clover at altitudes of 780 and lOOOm, but yields were highest from lucerne at 600m and lotus at 1300m. Factors which may have influenced legume establishment and yields are discussed. Keywords: clovers, lotus, lucerne, herbicide, hill country, aspect, altitude.

A summary of results is given from trials investigating the effect of sowing date, lime, altitude, aspect and seed treatment on the establishment and production of oversown lucerne at various sites in the semi-arid to sub-humid zones of North and Central Otago. , Spring sowings were generally more successful in establishing lucerne than autumn sowings. Early to mid-August sowings appeared more satisfactory on low altitude sunny country, while colder, shady aspect country responded to slightly later sowings. Higher numbers of plants established at the higher altitudes. Lime pelleting and heavy inoculation rates increased lucerne establishment when favourable environmental conditions followed sowing, but did not ensure establishment under adverse conditions. At pH 6.2 to 6.6, there was no consistent response in lucerne establishment to 1200 kg/ha lime, although up to 60% increases in establishment were obtained on some sites. Herbage yields of over 5 t/ha dry matter were recorded on the best plots, at lucerne plant densities in excess of 10 to 20/ms. With present techniques, oversowing of lucerne in this environment is not recommended, except on damper sites, as satisfactory lucerne establishment has not been achieved consistently in trial work.

Eleven years of experimentation ha've resulted in certain establish ment procedures being recommended and it has been shown that direct drilling of lucerne into chemically killed turf has given yields comparable with notmal cultivation methods. Maintenance topdressing trials have shown responses to increased quantities of potash, intermittent responses to sulphur, conststent responses to lime, and fleeting responses to boron. A grazing experiment with four lucerne varieties is described where increased yields have been obtained from a comparison of haying and grazing over production from a pure hay stand.