By affecting plant growth and phytochemistry elevated CO 2 may have indirect effects on the performance of herbivores. These effects show considerable variability across studies and may depend on nutrient availability, the carbon/nutrient-balance in plant tissues and the secondary metabolism of plants. We studied the responses to elevated CO 2 and different nutrient availability of 12 herbaceous plant species differing in their investment into secondary compounds. Caterpillars of the generalist herbivore Spodoptera littoralis were reared on the leaves produced and their consumption and growth rates analysed. Elevated CO 2 resulted in a similar increase of biomass in all plant species, whereas the positive effect of fertilization varied among plant species. Specific leaf weight was influenced by elevated CO 2 , but the effect depended on nutrient level and identity of plant species. Elevated CO 2 increased the C/N ratio of the leaves of most species. Caterpillars consumed more leaf material when plants were grown under elevated CO 2 and low nutrients. This indicates compensatory feeding due to lower tissue quality. However, the effects of elevated CO 2 , nutrient availability and plant species identity on leaf consumption interacted. Both the effects of CO 2 and nutrient availability on the relative growth rate of the herbivore depended on the plant species. The feeding rate of S. littoralis on plant species that do not produce nitrogen-containing secondary compounds (NCSC) was higher under low nutrient availability. In contrast, in plants producing NCSC nutrient availability had no effect on the feeding rate. This suggests that compensatory feeding in response to low nutrient contents may not be possible if plants produce NCSC. We conclude that elevated CO 2 causes species-specific changes in the quality of plant tissues and consequently in changes in the preferences of herbivores for plant species. This could result in changes in plant community composition.

No abstract

Palatability of parasitic plants may be influenced by their host species, because the parasites take up nutrients and secondary compounds from the hosts. If parasitic plants acquired the full spectrum of secondary compounds from their host, one would expect a correlation between host and parasite palatability. We examined the palatability of leaves of the root-hemiparasite Melampyrum arvense grown with different host plants and the palatability of these host plants for two generalist herbivores, the caterpillar of Spodoptera littoralis and the slug Arion lusitanicus. We used 19 species of host plants from 11 families that are known to contain a wide spectrum of anti-herbivore compounds. Growth of M. arvense was strongly influenced by the host species. The palatability of the individual host species for the two herbivores differed strongly. Both A. lusitanicus and S. littoralis discriminated also between hemiparasites grown with different host plants. There was no correlation between the palatability of a host species and that of the parasites grown on that host, i.e., hemiparasites grown on palatable host species were not more palatable than those grown on unpalatable hosts. We suggest an interacting pattern of specific effects of chemical anti-herbivore defences and indirect effects of the hosts on herbivores through effects on growth and tissue quality of the parasites.

Our understanding of the patterns of plant diversity in tropical forests and their responses to fragmentation are mostly based on tree surveys. But are these patterns and responses representative of other plant life-forms? We sampled trees, lianas, herbs, and ferns in a fragmented tropical forest landscape in South-west China. We compared community types generated by clustering presence-absence data for the non-tree life-forms with those generated for trees. We tested how well measures of tree diversity, density and composition, predicted cognate indices in other life-forms. We compared fragmentation responses, with respect to the three measures, of all four life-forms. Presence-absence data from all life-forms generated three community clusters, with only small differences between classifications, suggesting that tree data identified community types representative of all vascular plant life-forms. Tree species diversity and density indices poorly predicted cognate indices of lianas and ferns, but represented herbs well. However, the slopes of these relationships differed substantially between community types. All life-forms responded to fragmentation variables but their responses did not consistently match with responses of trees. Plot-level tree data can identify vegetation community types, but is poorly representative of the richness and density of other life-forms, and poorly represents forest fragmentation responses for the entire plant community.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.