Summary 1.Comparisons of adult and juvenile plant stages are often confounded by other factors such as differences in height between plants. However, these factors can be teased apart using common garden experiments and appropriate sampling designs. 2. Using paired sampling of upper and lower canopies of Eucalyptus nitens trees, which were growing in a common environment trial and only one of which had undergone the transition to adult foliage, this study assessed the effects of ontogeny and canopy position on the feeding preferences of Common Brushtail Possums ( Trichosurus vulpecula ) in captive feeding trials under choice conditions. 3. Possums preferred juvenile foliage over adult foliage. In contrast, canopy position did not influence preferences. 4 Adult foliage contained significantly less sideroxylonals (a compound known to deter herbivory by Brushtail Possums), but was significantly tougher (with thicker cuticle, more fibre, more lignin and a higher percentage dry mass) than juvenile foliage, suggesting that leaf toughness was more significant than defensive chemistry in influencing intake by possums. Adult foliage did not differ significantly from juvenile foliage in nitrogen, total phenolics or essential oil concentration. 5. Greater resistance to herbivory of adult foliage compared with juvenile foliage by a significant browser of eucalypt tree foliage could contribute to selection for an earlier change from juvenile to adult foliage in areas that are heavily browsed by possums.

This study assessed how the palatability of leaves of different age classes (young, intermediate and older) of Eucalyptus nitens seedlings varied with plant nutrient status, based on captive feeding trials with two mammalian herbivores, Red-bellied Pademelons (Thylogale billardierii), and Common Brushtail Possums (Trichosurus vulpecula). Seedlings were grown under three nutrient treatments (low, medium and high) and we determined how palatability was related to chemical and physical characteristics of the leaves. Pademelons ate more older leaves than young and intermediate leaves for all treatments. This pattern was best explained by sideroxylonals (formylated pholoroglucinol compounds known to deter herbivory by other marsupials), and/or essential oil compounds that were in lower concentrations in older leaves. In the low nutrient treatment, possums also ate more of the older leaves. However, in the medium and high nutrient treatments, possums ate more intermediate leaves than older leaves and showed a behavioural preference for young leaves (consuming younger leaves first) over intermediate and older leaves, in spite of high levels of sideroxylonals and essential oils. The young leaves did, however, have the highest nitrogen concentration of the leaf age classes. Thus, either sideroxylonals and essential oils provided little or no deterrent to possums, or the deterrent was outweighed by other factors such as high nitrogen. This study indicates that mammalian herbivores show different levels of relative use and damage to leaf age classes at varying levels of plant nutrient status and, therefore, their impact on plant fitness may vary with environment.

The Tasmanian montane and rainforest conifer genus Athrotaxis provides a system for investigating the relationship between leaf form and function and its adaptive significance. The two species differ markedly in leaf size, shape, degree of imbricacy and stomatal distribution, whereas natural hybrid swarms and glasshouse-grown hybrid progeny are highly variable for these traits. In glasshouse-grown plants of the true species and a diverse hybrid progeny, stomatal conductance and density were strongly correlated, and varied by approximately 400% among individuals. Hybrids displayed lower stomatal densities and less discrimination of 13C than the true species, leading to a negative relationship between stomatal density and δ13C. In contrast with the highly variable stomatal densities and δ13C in glasshouse plants, field-grown plants were highly conservative in both characters. This, combined with relatively low stomatal density and high water-use efficiency in field-grown plants suggests optimisation of the trade-off between assimilation and water loss. Foliar conductance in the light for the hybrids and A.selaginoides was only 4–6 times as great as, and was strongly correlated with, conductance in the dark, suggesting incomplete stomatal closure or high cuticular conductance. Athrotaxis cupressoides was less ‘leaky’. This may reflect adaptation to its more exposed habitat.

Ontogenetic changes in leaf chemistry can affect plant-herbivore interactions profoundly. Various theoretical models predict different ontogenetic trajectories of defence chemicals. Empirical tests do not consistently support one model. In Eucalyptus nitens, a fast-growing tree, we assessed early developmental changes to seedlings, in foliage concentrations of nitrogen and the full suite of known secondary (defence) chemicals. This included the terpene, a-pinene, whose impact on marsupial herbivory is unknown. To test for the influence of abiotic conditions on the ontogenetic trajectories we overlaid a nutrient treatment. Ontogenetic trajectories varied among compounds. Sideroxylonals and cineole were barely detected in very young seedlings, but increased substantially over the first 200 days. Total phenolic concentration increased fourfold over this time. In contrast, a-pinene concentration peaked within the first 60 days and again between 150 and 200 days. Nutrients altered the degree but not the direction of change of most chemicals. A shorter trial run at a different season showed qualitatively similar patterns, although a-pinene concentration started very high. We investigated the effect of detected levels of a-pinene and cineole on food intake by two mammalian herbivores, common brushtail possums (Trichosurus vulpecula) and red-bellied pademelons (Thylogale billardierii). Under no-choice conditions neither terpene reduced intake; but with a choice, possums preferred a-pinene to cineole. The ontogenetic trajectories of most compounds were therefore consistent with models that predict an increase as plants develop. Published data from later developmental stages in E. nitens also confirm this pattern. a-Pinene, however, was the only secondary compound found at significant levels in very young seedlings; but it did not constrain feeding by marsupial herbivores. Models must allow for different roles of defensive secondary chemicals, presumably associated with different selective pressures as plants age, which result in different ontogenetic trajectories.